Sep 08, 2023     gap tape       Category: Z-Tools+Tips

after Phase 1 tests are done, I can add the gap tape commonly used on gliders. See https://www.craggyaero.com/misc1.htm

COrrection - DON'T use gap tape on the canard. See Zeitlin comment on the early wing shape:

Marc J. Zeitlin
Tehachapi, CA 93561
Posted September 29, 2008
coyote said: ... the gap may be used to blow the upper surface under high angle of attack to reattach the flow, then it may be detrimental to use some.

Bingo. The canard/elevator (GU or Roncz) is actually a slotted flap airfoil. The airflow through the slot/gap is critical to it's operation.


coyote said: If it is just a "mechanical" gap to allow the elevator movement, then it may be positive to seal ...

This is NOT the case.


coyote said: I would clearly clasify the GU canard in the second category ...

You'd be wrong.


coyote said: but i'm not 100% sure, and i feel lot of you guys may had the idea and have made the test already !

It is unlikely that anyone who gap sealed their canard (GU or Roncz) would be able to get off the ground for any flight testing. Removing attached flow over the top of the elevator would be a very bad thing.

  Feb 23, 2024     Schematics - (2 hours)       Category: C22 Electrical

Gathering into one entry the schematics that are distributed into project entries, both pdf's and dwg's.

  Sep 12, 2021     Nose Lift - (100 hours)       Category: C13 Nose+Gea

This is a 2-part entry. First, the work done in 2009 by the original builder, then the 2021 work I have done.

2009-07-02 (5.00 hours):

I know it doesn't seem like a big step, or something that should have taken this long. But today I installed the switch for the Wilhelmson nose gear lift. Several years back, when I purchased this unit, I also bought the AEX1 auto-extension module. I quickly came to hate it - every time I turned on the power, it wanted to extend the gear, and nearly caused several accidents in the build process because of that. And I didn't like the standard answer of adding a breaker or switch - what good is a safety device that you can disable? This also came with a big rats-nest of wires, some of which didn't even match the supplied wiring diagram. For instance, "white" for me was actually grey. Blue was purple. It took a while to wire, and it looked like a mess. I never did get it neatened up to my satisfaction.

Today I cut all the wires out of the rats nest, removed the AEX1, and rewired the unit with the bare minimum of wires. I kept everything neat, bundled all wire bundles together with heat shrink tubing, and now I have a clean and simple installation that I'm finally happy with! The only thing I have left to do is replace the LED grommets - they're just sort of sitting there right now. I need deeper grommets for my instrument panel. I'll address all the LEDs at once when I do the full panel setup later. Besides, I'd just have to remove them anyway when I do my panel overlay (which I may do with carbon fiber trim sheets off eBay.

9/12/2021.

I also found a rat's nest of wires, and no functioning auto extension. I built an auto extension unit following the schematic posted by Marc Zeitlin (attached) and removed the wiring of the prior install. I constructed the unit as a panel-mounted module, and located it below the SDS ignition controller, then remade it to have the switches and transit lights in my switch panel, assorted relays on a remote mounted tray. Wilhelmson guidance says the lift is fused at 10 a, and the wire sizing guides say that the installed power leads (18 AWG) are wildly redundant: good to 30 amps over the 3 foot distance. But other builders have found that a fully fueled Cozy with 2 large passengers puts a significant load on the Wilhelmson nose lift. One builder went with a 15a breaker and 14 AWG wire, and believes this is beneficial. I will move ahead with the current installation and then revise if appropriate.

The Zeitling Auto Landing Gear Extension uses 2 relays of 2 circuits each and one relay of 3 circuits. In addition there are 2 pitot pressure activated switches, a throttle position switch and a LIDAR (light detection and ranging) unit to sense AGL altitude. In operation, the unit will auto extend when there are 3 simultaneous conditions: air speed of 40-90 knots, throttle less than 10% and less than 280 feet (85m) AGL. The unit has a disarm switch protected by a safety switch cover. Attached are the LIDAR install manual and the settings I programmed in.

This was a challenging bit of kit to assemble and then install. I started with some basics - I learned to label each of the circuits on the several relays, noting common, normally open and normally closed. I made aggressive use of shrink tubing to strain-relieve the numerous joins, and liberal use of solder sleeves rather than a ring terminal setup. The WIlhelmson unit connects to the panel through a 9-pin Molex using the high power .093" pins (rated to 17 amps per pin). After the ease of DB9 pins the Molex is particularly awful to work with, and I had to find the tools to rebuild the harness. I ended up going with individual connections under shrink tubing, as in the Nuckolls method for DB9 pins in small spaces. Finally, I found and used a lot of (Mil Spec) zip ties to organize a challenging set of wires. The gear lift itself is protected by a cover which is easily removable for service, with wires passing through rubber grommets.

One build note - the Wilhelmson nose lift was installed without use of a violet wire in the #1 position: a white 22 AWG was used in its place. I have marked the pin numbers and confirmed the function of each - otherwise the unit is consistent with the technical guide provided by Wilhelmson's EZ-Lift.

1/18/2024

The Wilhelmson design is the best available for the Cozy, but it had a design flaw addressed by a retrofit kit. See the attached write-up.

  Nov 02, 2023     Starter - (10 hours)       Category: C23 Eng+Cowl

I started by installing the starter previously used on this engine: one salvaged from a Honda Civic. It came with a custom mounting bracket milled from heavy aluminum stock, but the fit was not perfect, and at full extension the starter gear engages only half the surface of the ring gear. I was happy when another Cozy builder had a spare starter and took him up on the offer. It's "new old stock", meaning its been on his shelf for years. It's an ElectroSystems MZ-6222, which became part of Kelly Aerospace, then Hartzell. It has a 12/14 gear pitch, the one I need for my 149 tooth ring gear. This model starter is the same one used on the Mooney M20J, the original installation of the Cozy engine.

One bit of good news - using aircraft starter and B&C alternator means I was able to make a good fit for the reinforcing link between the two. Now to find the correct bolt! 5/16x1.25 - 18, drilled head. Add it to my SnF shopping list.

  Jul 11, 2023     C18.17 Instrument Cover - (20 hours) Category: C18 Canopy

I want to have an instrument cover that is light, shields avionics from rain if the canopy should be open, and provides a glare shield for the panel. With a forward hinged canopy, the plans approach won't work. I made and discarded 5 solutions, then decided to come back to the issue after the instrument panel was fully installed and wired, canopy installed with edge seals and the canard on.

  Dec 02, 2020     13. cabin side - part 1 - (60 hours)       Category: C22 Electrical

The Cozy firewall is a non-magnetic stainless steel sheet over Fiberfrax. A lot of planning goes into the layout of both the engine side and the cabin side. The engine side is designed for ease of service, engine layout and clearance. The cabin side is about short runs, logical flow and ease of service.

A lot goes on the cabin side of the firewall. I made up a template and did the layout of components – battery, contactors, relays, battery and engine busses, et al. Cables and connectors come up from the lower corners, then run on the outside of the curve. I aim to have the runs to alternator and starter use the passthrough fittings below the right side of the picture (pilot side). The electronic ignition runs will use the passthrough below the left side of the picture (passenger side).

Ground - Engine side. The alternator grounds through its case to the engine mount. A #2 AWG cable connects an engine case bolt to a brass bolt passing through the firewall and becoming the cabin ground. That brass bolt is the fat point ground on the firewall, connects to a non-magnetic steel firewall by direct contact. Also on the engine side is a supplemental ground for the starter - it grounds through its case to the engine, and also through a #2 AWG cable to an engine case bolt.

Grounds - Cabin side. The busses are grounded by a brass bolt / fat point penetrating from cabin to engine side of the firewall. On the cabin side it is connected to battery negative with a #10 AWG cable. The starter contactor grounds, through two short brass bolts passing through the firewall. A #10 AWG cable connects one of the bolts to a field of tabs. The field-of-tabs has a brass bolt pass through to a second field-of-tabs on the engine side of the firewall.

Power - Engine side. The _normal_ power source is a 60a alternator (B&C LX60). Consistent with the Nuckolls schematic "Z101b", a #8 AWG wire runs from the B lead through a lower right side access port to the cabin side. The field wire is a brown plug on the forward side of the alternator, with a lead running alongside the other power wires, throught the lower right side access port. The _auxilary_ power souce is a permanent magnet generator (B&C SD-8), made to fit the vacuum pump port on the accessory case with a rated output of 8.4a at 3500 rpm (2700 engine RMP). The Nuckolls schematic allows for a generic aux alternator, while his text endorses the B&C generator. I followed the B&C instructions for the SD-8 generator, so that the aux power wiring differs slighting from the Nuckolls schematic (see below). In the future I may choose to replace the 8a with the 30a version from Monkworkz (also mounted on the vacuum pad). See https://monkworkz.com/product/mz-30l/ .

Power - Cabin side. The 60a alternator is fused by a 12AWG fusible link used next to the contactor connection (see schematic detail below). The alternator regulated by the B&C LR3D-14 'controller', consistent with the template schematic Z101. In the detail photo below, the LR3 is the silver box below the red tray. The LR3 is 1) a linear regulator, generating no audio or radio noise; 2) a solid-state, crowbar over-voltage protection circuit; and 3) low-voltage monitoring and warning (yellow light flashing at the panel). The controller also has overvoltage protection, a capacitor to smooth the current delivered and a relay to switch between main and backup alternator. At the panel a clear-yellow warning light that will illuminate if the crowbar over-voltage protection module is tripped or when the alternator switch is left off.

Above the B&C controller is a red tray. It holds a regulator for the backup alternator (B&C model PMR1C), which is a _switching_ series rectifier with a rectifier bridge, adjustable solid-state regulator circuitry, and built-in heat sink. At full load (8.4 amps output) it "rejects" ~20 watts of heat. Per Nuckolls, "The power-path components depend on thermal conductivity of the potting compound for heat rejection. The rectifier/regulator's service life is tied to optimizing thermal management." The installation instructions warn against mounting to a composite surface - the red tray is an aluminum platform for the unit. (On an aluminum tray at 8a output the B&C test results point to a max temp of 125F. Since aux power operations are last only until a safe landing, the heat seems a reasonable risk.) The large blue capacitor below the red tray has two functions – it presents the SD-8 generator with the starting polarity, and it smooths out the noise that comes from a switching regulator. Also on the red tray is a relay, switching the alt aux alt bus feed to the main bus. There are diodes protecting each of the relays and a diode bridge protecting the engine bus (the silver box above the red tray).

[I'm not overly confident that 8a of backup alternator can do the job. There is an alternative for a pad-mounted backup - the Monkworkz 30a. I'll check it out at SnF and can make a final decision after wiring the SDS system.]

4/16/2023. This evening I finished the power lines described above. It continues to seem nutty to have the heaviest cables (#2 AWG) used for the very short term task of starting the engine, while much lighter cables (#8 and #14) carry the power from the alternator and generator.

===============================================

Lots of learning. I had used several methods to secure fasteners (rivet nuts & click bond studs) to the cabin side of the firewall: marine adhesive, epoxy & BID, flox & BID, marine adhesive on aluminum plate holding an installed rivet nut, et al. I decided to test the security of the components. Results were terrifically inconsistent, with some click bond studs easily twisted off with pliers and others pried off with a paint scraper and elbow grease. Bottom line, I need to find a new way to secure components to the cabin side of the firewall.

The new approach is to build the system on the bench, firmly securing everything to a sheet that can then be attached to the firewall, using the existing holes. This will sharply reduce the number of firewall penetrations - I won't be drilling two dozen holes to secure everything. I then prepared such sheets using stainless steel, a coreplast board and my own layup (1/2 foam with 2 layers BID each side). The 22 gauge non-magnetic stainless steel was heavy (3 lb 13 oz). Coreplast was 2 lb 4 oz, my own layup 1 lb 14 oz (before trimming). 22 gauge aluminum was 1 lb 13 oz (after trimming) and it's easier to set fasteners in the aluminum. I'll go with the aluminum.

  Dec 31, 2011     Fuel Level Sensing - (40 hours)       Category: C21 Strakes

Original builder Chad Robinson intended to install an auto engine and run auto gas. He modified the build of the tanks. All of the work had been removed from the plane when I received the project. The remaining differences are
1) Each tank was built with 5 extra openings through the strake, intended for electronic fuel level sensing using magnetic reed switches, all spec'd as ethanol- resistant. 4 are on angle and sealed by BID. One is at the floor of the tank and holds the stub of a switch intended to track fuel level in the sump. All openings are 1/2" male with a brass pipe thread.
2) The tank sump was made extra large - at least 2 quarts larger than standard. Where the standard Cozy holds 29 gallons each side, #1147 holds 30.
3) The tanks are lined with ProSeal.
4) Chad had installed fuel sight gauges using an ethanol-resistant backing plate of polyphenylene sulfide. I removed those gauges and installed the Vance Atkinson standard, which uses an aluminum backing plate. The Vance package comes with an LED, which I have wired to come on with the master.

  Dec 29, 2023     06. Panel Final (?) - (48 hours)       Category: C22 Electrical

My panel plan has had major changes over the past few years, but now the changes are smaller and driven by market pricing of alternatives.

Structurally, the plans composite panel is replaced by aluminum. Thickness of the aluminum is .041. I don't know the alloy, but suspect 2024. Composite "piers" tie the aluminum to the fuselage at the right, left and bottom center. The radio stack ties the panel to the F25 behind it, again adding stiffness. Aluminum angle at the top carries red LED's and adds stiffness.

A busy panel with old school round guages and small screens is now glass panels + ipad. Switches are now (mostly) a row at the top of the panel - see the photot of the test fit. I will work on some other parts of the build, then buy the avionics and cut the final panel. After mounting everything I will set the fuse block locations and do wiring.

(Sooner or later I have to buy a lot of electronics. But not today!).

  Jul 02, 2020     going with a rolled application - (100 hours)       Category: C25 Paint

I took Buly's advice and painted what I could now. Here's why I painted way before it was called for in the plans...
1) the earlier in the process the lighter the fuselage, and a good paint job pretty much requires having the plane upside down for a while.
2) I don't want paint all over the cables and wires
3) Suddenly my family starts to see the airplane

I'll split this into several entries, and this is about process....

Why roll-on finishes. I ran through the list of recommended choices. The local airplane painter (Hawke) declined to quote on a disassembled airplane. I looked for but did not find a moonlighting auto painter who would come to my place. The fuse is too large to fit into the paint booths of the small local shops. A shop with a large booth wanted $8500 for me to be the airplane guinea pig. I went the DIY route and rolled everything. I am satisfied for now, but somewhere down the road I will again disassemble and use the fiberglass cutting compound, then get a professional shop to spray a clear coat.

I painted parts, not the plane. Parts was a _long_ process. The section "Sanding" describes the fill and shaping, ending in multiple coats of pure epoxy. I then worked through all the same parts, bringing each up through a minimum of two coats of primer, then a minimum of 3 coats of top coat. I have now done with the big push in finishing. There are still some parts to do: I need to build fairings, wheel pants, cowling and whatever is required by repairs after installing landing lights, gap seals around windows, et al. I'm not necessarily recommending this, but the slow and steady picking away at the task worked for me. See below and the attached photos.

Process. Big fill, guide coats, Bird epoxy wipe, sand to 180. 2-part Awlgrip primer, 2 coats. Wet sand at 360, clean with isopropyl, then degreaser, then tack cloth, then three insanely thin coats of 2-part Alexseal, wet sanding and cleaning between coats. Finish with a polish intended for fiberglass. Everything was done in the garage, where dust and pollen and bugs all affect the quality of the work. And it's Florida, so we also fight sweat running off the painter and into the paint.

Products.
- I had resisted filling with anything other than micro, but at the end used 2-part Bondo "glazing and spot filler". A phenomenal product that fills, smooths and sands really easily. Two 5 oz tubes were used, most of it sanded off and some still on hand.
- The group discussions arrived at an endorsement of Awlgrip as a reliable primer. Awlgrip is a 2-part primer with slow and fast reducers, and can be rolled or sprayed. Success rolling the Awlgrip made me look into whether I really needed to spray the top coat.
- Internet time + talking to local boat shops persuaded me _not_ to finish coat with JetGlo or an automotive. I went with an updated formulation of a marine paint (Alexseal 501) that has both converter and reducer intended for brushing and rolling. I've already found that it repairs well (I backed the gear leg into a bench and !@#$!). The product is highly resistant to UV and chemicals (up to and including acetone). Interlux Perfection is now available in a similar chemistry and from the same company as AwlGrip. I might look hard at that alternative; both are expensive but the Interlux is widely available at marine retailers.
- The final polish was Meguiar's #67. It smooths the orange peel on the surface but flattens the gloss and leaves swirl marks - I skipped using it on later parts. I came to think of it as a cutting compound intended for fiberglass.
- Miscellaneous products were a degreaser intended for fiberglass (Interlux 202), 6” mini mohair rollers from Redtree, and a fair number of disposable measuring cups and paint pans. The rollers were special order while the pans and filters came from Home Depot.


Learnings.
1) allow for the slow evaporation of cleaning products, which will otherwise give an orange peel.
2) if rolling, use only the Redtree brand mini mohair rollers.
3) A little goes a long way – 300 cc's of product (base + converter + reducer) was more than enough to fully cover the top of the fuse. A single gallon of base is considerably more than you need for the whole project.
4) Each new coat looks even more wonderful and raises the standard. When do you stop? A 3rd party has a useful voice.

Cost. My hands hurt from sanding and I'm using my wife's clear polish to save what's left of my fingernails. I probably gave up some IQ points while huffing paint around the mask seals. The total spend will come out to $800 or so – 3 quarts of Awlgrip + 1 gallon of Alexseal + the related converters and reducers + a lot of paper, mohair rollers and other supplies. Rolling gave me the convenience of working steadily at the project, solving problems and learning the process one part at a time.

Bottom line, if you can find a paint shop that will do the work by all means spend the money and have them do it. If you know how to spray, put on a respirator and go auto paint. But if both of those don't work, consider rolling a high quality marine paint. Now let me get back at it….

  Apr 18, 2023     fuel capacity and unusable fuel - (.2 hour) Category: E01 Next Steps

Unusable Fuel. I've done unusable fuel tests in a high wing tailwheel aircraft, but never in the midwing Cozy, with fuel sumps at the wing root. The Andair fuel valve does not have a setting that feeds from both tanks. The reg reads: "�59 Unusable fuel supply. (a) The unusable fuel supply for each tank must be established as not less than that quantity at which the first evidence of malfunctioning occurs under the most adverse fuel feed condition occurring under each intended operation and flight maneuver involving that tank. Fuel system component failures need not be considered." There was a recent discussion of this on the Cozy builders group, concluding 1) land on the fuller tank; and 2) landing slips should either keep high the full tank wing or limited to 10 seconds.

Fuel Capacity. A fuel specification and tank capacity is to be marked on the aircraft, a guard against the inexperience of the person on the fuel truck. This is normally done with either decals / stickers on the aircraft or engraved fuel caps [see note below]. The process is
a) verify all fuel lines are secure, valves closed and fittings are snug
b) set longerons at level to 2 degrees nose down (inflight is 2-4 degrees up, level is the landing attitude)
c) put masking tape on the sight guages, an ohm meter on any resistive fuel capacitor or capacitor series
d) prepare a tank stick - a paint stirring stick painted flat black or white will work. Mark one side right, one left. Best to call it a refueling aid: if you treat the stick as a fuel guage, it falls under 14 CFR 23.959(a) - see above.
e) support the wing on the side being filled - one side empty and one full could lead to embarassment.
f) fill and mark (in pencil!) in the increments you think will work for you - I like 5 gallon increments, perhaps because that's the size of my fuel containers!

I found that my Atkinson sight gauges are stuck - fuel is barely creeping in and the red ball is not floating free. The Cozy mailing list has the solution...

Notes on resistance (in k ohm) of the fuel level senders:
fill Left
5 g
10 g
15 g 95.0
20 g
25 g 115.0
28 g 115.6


Out in Utah there is a print & layout shop happy to laser engrave fuel caps for all of $8 each. Contact
Aeryk
Jet Blast Designs
2474 South 1500 East
Vernal, UT 84078
435-823-1890

  Mar 20, 2022     Cowling - fasteners - (4 hours)       Category: C23 Eng+Cowl

Piano hinges join to the wing, Dzuz fasteners for the top and bottom arches. See attached write up.

The stronger extruded piano hinges are pricey - $37.50 each, but they are stronger, lighter, and look better than the alternative of Camlocs. Now I'm paying the price for a deviation from plans + the custom cowl. _Lots_ of tweaking has happened. Constantly moving above and below the cowl proved to have its own risks - three weeks of wearing a bandage after I barked my head against a cowl edge.

  May 24, 2022     Cooling - (5 hours)       Category: C23 Eng+Cowl

I've been concerned about cooling since the very first time I saw a Cozy. Pre-flight, I intend to make two tweaks to the air intake at the stock NACA vent and one to the air exit. If post break in I see climb out CHT's over ~420, there more things to try.

At start, I'm using
a) corner tabs on the NACA duct. This is no more than a bracket added to the rear corners of the NACA, aiming to keep in flow the turbulent air that wants to spill out and away from the air intake. See detail on attached.
b) cooling large scale VG's, creating turbulence at the intake. See detail on attached.
c) small scale eductors. Eductors are said to use exhaust pulses to draw air out of the cowl. Maybe. I need to protect the cowl from the exhaust heat, and I made stainless-lined eductors from CF-Cormat-CF using high temp epoxy. I'm not thrilled with the impact on fit of the upper and lower cowl, but I'll give it a go.

As a backup, I made armpit scoops which I can use to force air iinto the oil cooler. See below a Long EZ example of how such a scoop is part of the lower cowl. In my case the scoop-cowl attach point is to be a square the same size as the face of the oil cooler. Nothing off the shelf can plug and play, so I have to either modify an after market part or scratch build. A Bingelis standard says says I need an additional ~30 in2 of inlet area to handle a 200 hp engine. I can do that with a scoop dedicated to the oil cooler. I created a 2-part mold using 4 lb pour foam, and used my new favorite CF-Coremat-CF. No fancy bell mouth. The stem / support was left long for final trimming after I finish the cowl and install the oil cooler. I made one for the right and one for the left - I don't know which side I'll use for the oil cooler, so one of each.

  Nov 08, 2023     08 - back up com - (12 hours)       Category: C22 Electrical

I came across a Becker com radio that was thought to be working. It wasn't, and taking it apart did not show anything obvious I could repair. I am lucky that the comapny is based here in Florida, and they were able to quickly repair and return the unit. The spend including shipping was the radio ($85) + repair ($190) + break-out connectors for the male and female DB-15 connectors. I know where in the panel I want it, and now I have to build the mounting tray. (A shout out to Becker - these units are beautifully made and packaged, and the most durable brand I know. Because the US is still a 760-channel environment, we can use these older models. 5 watt transmit. )

  Nov 03, 2023     Cowl #5 - oil door Category: C23 Eng+Cowl

Procedure for cutting into any composite panel....

Nick,

Perhaps I am misunderstanding in which case forget the following.

It appears that you want to cut a hatch in an already existing front area and you want to be able to have a door that opens and closes and seals on a flange (that fits it) you also want a drip rail for moisture (which of course you want to drain somewhere.

If I am correct, the way that I have done it was:

Before cutting the panel out:

1. draw the outline of the purposed hatch
2. Place one layer of duct tape over this line extending about 1" OUTside the line and approximately 1" inside of the line (this will be adjusted later
3. Make a 3 layer or so BID layup over this duct tape. (creates the flange which will be foxed inside of the fuselage)
4. Peel ply the layup.
5 When cured, separate it from the duct tape and totally cut out the shape. (I don't have to add the word carefully.
6. Neaten up the outside of the layup produced in step4.

7. If you want a drip rail put a round shape of foam on the layup (on the side that was touching the tape) at the inside edge sand and make a layup over this with enough overlap to secure it. if you do the drip rail, you will have to first trim the inside of this layup and let cure
8. Remove the foam and refine

9. sand the inside of the fuselage approximately 1.5" from the cutout hole around the entire hole left after the cutout (for bonding.)
10. make a few layups of BID around this hole the thickness that you want to use for your gasket material
11 Using a relatively thin flox (but still thixatrophic) butter the inside of this fuselage layup
12. fit the new piece that you created in step 9 and or 4 with or without the rail through the hole orientated in the same way in which it was made.
13. Press it into place (upward) (from the inside) and using multiple clamps hold it to the periphery of the hole allow cure and get rid of the goobers
14. When cured create a flox corner in the outside skin, a flox fillet at the junction of the inside skin and the now affixed flange and use some light bid to seal the foam and create a stronger joint.
15. clean the layup and slightly radius the outer skin flox joint
16. Reduce the size of the cutout panel to fit its new flanged hole (remember you will have to address the raw foam edges which will, of course take up some space)

Hinge and lock.

  Oct 15, 2023     wing finishing & attach - (12 hours)       Category: C19 Wing+Ail

Filling, sanding and finishing the wings is a lot of work. I enjoy the scrounging and improvization aspects of plans built, but it takes a lot of time that I could have spent completing the project ahead of life's unexpected events. _Don't_ take my route of slow & cheap - buy the Durabond boards and/or 3M Cubitron paper.

This is also a spot to store the instructions on attaching the wings - plan on buying the bolts in several sizes so that you don;t have to measure, then wait for delivery of the correct ones.

  Aug 04, 2023     Panel glare shield - (8 hours) Category: C24 Seats+Fair

The plans call for cutting out the forward edge of the fuselage when fitting the forward canopy. That edge is to be used to make a glare shield for the instrument panel. My project arrived with the canopy installed and no glare shield piece. I built two different glare shields but don't like either one - I'll redo the instrument panel (now for G3X system) and come back to this issue.

  Aug 24, 2023     Firewall pt 2       Category: C15 Firewall

I have two tasks - the firewall needs to fit the engine, and I have to patch the gaps and then extend onto the strakes. On the first, see the attached Word document with embedded pictures. On the second, see the following.

First builder Chad planned on a Mazda engine, and I'm using the wide deck angle valve Lycoming. The Lycoming is wider, requiring a wider cowl matching to a wider fuse. I defined the fuse shape as I made the cowling, and now I need to expand the area covered by the insulation and firewall. Also, Chad skipped covering the fuel sump and the wing root / aft wall of the fuel tanks out past the wing attach bolts. I want to cover that area and, separately, add the wing root heat shields. See attached photos showing the fuse area I need to cover.

I have a big sheet of stainless that I could cut, but I'm hearing about some modern products that do a better job at lighter weight. Les Laidlaw came up with a high tech product out of Italy: Teknofibra, now a subsidiary of Volkswagen. It is described as "Can manage temperatures up to 800'C only transferring through 35'C on the reverse of the panel it is mounted to." Les used FiberFrax against the composite bulkhead, then the Teknofibra, then stainless steel tool wrap. I'll follow his lead, substituting ceramic felt for the fiberfrax - it is easier to work with and has a higher heat resistance. (Commonly used for kilns and fireboxes, as it has a 2300F temperature rating. I used the same product for the heat shielding of the gear legs near the brakes.) I had a choice of SS alloys for the tool wrap (cut with shears as it is thicker than foil, too thin to rivet or weld). I went with 321, an alloy with extra corrosion resistance. The spend is $150 for Teknofibra + $60 for tool wrap + $20 for the ceramic felt = $230. The three products are shown below. The commerically available products leave me with a lot of left over foil and felt. I'll try to find another builder who can use the products.

  Jun 25, 2023     pitch trim, rolll trim & control sticks - (80 hours)       Category: C17 Trim

Both pitch and roll have trim (actuator + a spring) and autopilot servos. Actuators are controlled from a hat switch on the control sticks.

Pitch. My project came with the Strong pitch trim unit, a modified cordless screwdriver. 12v turns the shaft of the motor, which turns a torque controlling clutch, which spins a threaded rod. The rod moves a cylinder that holds a spring loaded piston. The springs give the piston ~3 inches of play, or 1.5" each way. The piston is connected to the elevator control arm. In operation, a hat switch runs the screwdriver, the threaded rod moves in and out of the cylinder, and the spring loaded piston presses or pulls on the elevator. That's how it is supposed to work. But the unit I got had a threaded rod with a tight bind in the piston. The motor had been abused to the point of sounding, even when free of resistance, like a coffee grinder loaded with gravel. And that motor absolutely would not turn that threaded rod. After some research and emails, I decided to salvage the spring loaded piston and replace the screwdriver / threaded rod with a linear actuator. The actuator has a 3" throw that matches the full range of the spring piston - if I have runaway trim I over power it and solve the problem on the ground. I picked this actuator because of the throw, steel case construction and a super slow rate - this is 0.2 inches/second, or a full 7 seconds from neutral to full up or down. This gives 42 lbs of force for 0.33 amp of current. The actuator is controlled by a hat switch on each of the control sticks.

Roll. For roll trim I made the plans parts, then decided to go with the (common) builder alternative and used electric roll trim. I then made a Wright-Hanka spring, described in the Canard Pusher newsletter of April 2003 (p. 24). I used 8 plies of carbon fiber, and my super strong high temp epoxy - Raka non-blushing, made for use in tropical climates. The spring is connected to a small, low speed actuator, which is connected to the aileron throws. The connection is a 5/64ths cotter pin. The actuator is slow and short: a 4" throw (2" each way vs 1" expected use) and force of 14.5 lbs at 0.6" per second. Note - an earlier spring used 4 plies of BID and the West Systems slow epoxy. That spring was wide but far weaker, cracking under ~30 lbs of force. I recommend the CF and a really good epoxy. From here I can thin the spring to get the targeted give.


Autopilot Servos. The Dynon SV-AP-PANEL NOSE UP and NOSE DN buttons can control up to 2a of pitch trim motor pitch trim is 0.34a at full load. From the Dynon install manual on 18-4 – 18-5:

"Trim Motor Control. In addition to the AP mode buttons, the SV-AP-PANEL includes an adjustable, speed-sensitive trim controller, eliminating the need for relay packs or a separate trim controller. The integrated trim controller is wired to a D15M connector that can be connected to Aircraft power (12V only), pushbuttons (typically, on the stick), and power to and control of up to two trim motors - such as those made by Ray Allen. A set of pushbuttons can optionally be installed for the copilot. The pilot's pushbuttons take priority over the copilot's pushbuttons. Pushbuttons used for Trim Motor Control must be Push Button Normally Open (PBNO) – a momentary switch with a contact that is closed only when the button is pushed. One terminal of the pushbutton is connected to the selected pin; the other terminal of the pushbutton is connected to avionics ground (can be common with Pin 2)…. When SkyView is not powered on (not communicating with the SV-AP-PANEL) the speed scheduling features (QUICKEST TRIM SPEED / SLOWEST TRIM SPEED) are not available. In this failsafe mode, the trim motors will run at their full speed when trim buttons are pushed."

My intended practice is that the airplane is properly trimmed before handing it over to the a/p. The Dynon a/p can be used to control the pitch trim? That sounds like too much system reliance, and I plan to keep separate the trim and a/p.

The no trim force trim speed should be verified to be about 120-130 KIAS, established in phase 1 testing. This is equally applicable to the plans trim (tension springs), any of the leaf spring options or the electric trim options. With "natural" trim speed rigged ~130 KIAS, trim failure leaves the airplane controllable. Before that first flight, pitch trim is set so that at pitch neutral the trailing edge of the elevator is in line with the trailing edge of the canard.

Grips. The grips look to be the teak ones made by Custom Aircraft Grips (http://customaircraftgrips.com/). They were smooth but I could not detect what finish had been applied - a finish helps the grips stay clean. I treated them with "Watco 66041 Rejuvenating Oil" and the grips look and feel good. The left/ pilot grip has a 4-way hat switch (trim), a PTT switch under the index finger holding the stick, and a rocker switch to the left of the hat. The right / co-pilot stick has hat switch and PTT. I'm trying to figure out what I'll use the rocker for... The boots are auto shifter boots picked up on EBay, made as a tall cone sewn from 4 triangles of artificial leather 9” tall. The cone is turned inside out, zip tied to the stick, then pulled back over the stick right side out now. Then stuff it down in the hole leaving the large end of the cone out of the arm rest. I made a bracket to fit the arm rest - see picture below. Trim leather and bracket so the screws go thru the leather to hold in it place, while staying clear of the lids on the map pockets.

The grips have a good size wire bundle - 10 wires on the left, 7 on the right. Those wires are in a super tough shrink-fit protection, and the end beyond the tube were protected with snrink tubing. The wire bundle was passed through the center of the aluminum tube, then past the hinge on the aileron/elevator mixer. On the ouside of the map pocket, aft of the control stick bottom, I installed a terminal junction block where labeled wires from the stick switches can match up with labled wires running to trim motors, PTT and some 3rd function. If I was sure I wouldn't need to remove the stick I would use a DB-15 connector, but this allows me to easily change anything that needs changing. The only hjard part was fabricating some kind of mount for the terminal block - I ended up with nut plates sandwiched in BID, the mount then attached to the map pocket using G1000 Flex, a thickened epoxy.

  Apr 16, 2009     Arm Rests - (32 hours)       Category: C24 Seats+Fair

Chapter 24 step 4 - Armrests

2009-04-12 (1 hour): I decided to do something a little different than the plans for my arm rests. Others have complained about not having access to the hidden area here for maintenance purposes, so I decided to make a removable access panel (as others have done). To do this, I prepared the right-side top arm rest surface to fit around the stick and map pocket. I then 5-minute epoxied two small lips onto it. The first is at a 90-degree angle, and sits between the stick and instrument panel. This will support a vertical filler panel that goes from the arm rest down to the thigh support. The second is at a 45-degree angle from the stick back to the seat. This will eventually be sanded away, but for now forms a lip against which I did a BID layup. My plan is to sand away this section, then install quarter-turn fasteners on it to support an access panel. I'll put a matching lip along the fuselage wall to support the bottom of the panel.

Today I just made the right side because I want to fit it together and see how it works before I do the left to match.

04-16-2009 (1 hour): I now have the second layer of BID on the right side arm rest, and I was pleased enough with how it turned out that I started the left side as well. I did remove the foam under the lips I had made on the right side, so there's now a 3-BID lip for the fasteners. There's plenty of clearance around the torque tube.

2009-05-14 (2.00 hours): Added nut plates. I suppose I could have done this as part of the previous layup, but it was just easier to do it as a separate step. Today I floxed on nut plates along the arm rest edges to hold the side plates. I also put a 1 BID ply on top of them to keep them in place. Job done.

2011-09-11 (1.00 hours): Pilot's side installed. This has been hanging around for a while, never a priority - and it still really isn't, but I had some extra epoxy tonight and couldn't think of anything else to do with it so I made up some BID tapes and installed the pilot's side arm rest. (The copilot's was installed long ago.)

It's a little hard to tell from the picture, but the arm rest isn't all one piece. It has an angled inside edge that stops about 2" down from the arm rest, with nut plates behind it. The nut plates are used to attach a trim piece. It only took an hour or two longer to fabricate it this way, and this makes it MUCH easier to service the controls underneath the surface because basically the entire side of the arm rest comes off. It's still plenty sturdy - the lip the trim piece attaches to is 4-BID and it makes the arm rest extremely stiff - and the trim piece will provide additional support once it's installed.

2023-07-12 (3 hours - PK build). A deviation from plans is a nice custom touch that costs time - but I'm glad to see the nut plates! Working on the lower skirting boards for the pilot and co-pilot, the first step was templates. I used foam board, cycling back and forth to the plane. The control stick throws are significant and I now need to discard and re-make the control stick covers to fit the new trim pieces. I also ordered the fabric I will use as covering - yellow vinyl in back and an upholstery fabric in front. $51 for the two remnants from fabric guru. This was another chance to pull out my Singer 6800C.... it was all about using the right non-stick foot and a super sharp leather needle. Add the right thread and it's more than half way to a solution.

2023-07-15 (6 hrs). For rear arm rest, per plans a 3" and 3.5" wide length of foam joined with micro. To have it removable, both 2 BID outer and 1 BID inner extened to form a lip of BID where the arm rest contacts the side wall, with G10/micarta inserts. 5 click bond studs on each side of fuse, slightly below the level of the top of the CS mini-bulkhead. BID drilled for the click bond stud, and outer face of arm rest drilled for access. Attached photo shows one arm rest curing the inner BID layer, the other pre-glassing. For front arm rest, made a lower skirting board, which will be drilled to match the nutplates in the installed arm rest. One bit of good news - the record heat lets the epoxy cure fast, allowing me to do 2 layups a day. And I again found that my 'tropical weight' epoxy outperforms anything from West Systems. Raka "#900 Medium Resin" + "#350 Non-Blushing Hardener".

2023-08-04. The hardest parts to fit were the upper fairings of the back seats. The angles are hard to fit, then hard to cover - see the photo with all the small bits trimmed from the original too-large parts. For the rear arm rests I added "No Step" patches. Also, in the front I added pockets, embroidered with a nicely detailed honey bee. (Never stop working!) Below the pockets are oval circles - those will be cut open for the seatbelts. I'm glad to have these done and stored away. I'll get the inspections done and then reinstall all these fairings.

  Aug 04, 2023     Cover for EZ nose lift - (6 hours)       Category: C24 Seats+Fair

Used foam board to make a box that fit just right. The box was covered with packing tape, then BID with reinforcing tape on the edges. After inspection I will make it secure against air leaks using aluminum HVAC tape (thanks to Andrew Anunson for the construction tip).

  Apr 19, 2023     Engine mount       Category: C23 Eng+Cowl

The engine mount was bought from Cozy Girrrls, and responds to the early concern over cracking by adding extra bracing. Cleaned with alcohol and acetone,, then painted iwth a zinc chromate primer and appliance epoxy.

The engine is secured to the mount with Dynafocal mounts made by Lord Manufacturing (a division of Parker Hannifin). For the build / dummy engine phase I used the mounts that were used in the build phase for Andrew Anunson, then build phase for Hugh Farrior. As near as I can tell they are the heavy duty mounts sold by Aircraft Spruce for the Acro and Christian Eagle - aerobatic biplanes. I've had them on and off a few times and see no distortion. On 5/26 I corrected an earlier error and moved the large diameter washer to the compression side (for upper Dynafocal cups, the firewall side, for lower cups, the engine side). Moving the washer had the effect of raising the prop. With tires inflated to 60/60/55 and all 4 thick washers on the firewall side, the bottom of the prop extension was 45" above the floor. The new measurement is 46".

Prior to fitting the cowl I wanted to switch over to new mounts. I bought a new-in-box set of the mounts specified for the Cozy: the J9613-40 originally spec'd for the IO-360-A3B6D. When the inner gel capsule burst I used the customer service email at Lord, sending in a description and pictures. That was Memorial Day. Now it is 7/13. I called again (877-275-5673) and left a message with technical support.

I may or may not change over to the new mounts - the old ones have not deformed in any way. In January 2019 I spoke with customer service at the Lord mount division of Parker Hannifin: I described the pusher configuration and mentioned that it was common to use a prop extension. That news made her recommend a stiffer mount, and to avoid the ones with the gel capsule dampers. I infer that the mount is dealing with the vibrations created by the prop as much or more than the vibration of the engine. A prop extension is a longer arm to the prop vibration, and merits a stiffer mount.

  Jul 05, 2023     Sewing machine - (20 hours)       Category: Z-Tools+Tips

There are parts of this build best done with a sewing machine. I worked with fabric adhesives for too long, then bought and learned to use a sewing machine. I'm now a fan of the Singer 6800C. Key seems to be using a specific needle, foot, thread and stitch. I sewed seat belts, then a harness holding the large O2 bottle, then super thin nylon of the sun screen modified to be an instrument cover. Still ahead is sewing vinyl for side panels.

To date, I have consistently used Coats & Clark thread - 100% nylon upholstery CA00011 ART S964 K2 4550. I have used a "locked triple stitch". For the webbing, the stitches are done so as to be stronger than the 3" nylon webbing. The triple locked gives the stitch density that makes this work.

  Jul 04, 2023     Oxygen - (24 hours)       Category: C99 to sort

The experienced Cozy flyers strongly recommend an O2 system - the plane does very well flying at altitudes of 8,000 - 16,000. Flying in the southern Rockies I learned that I get headaches starting at 8,000 and that was 20 years ago. So, some kind of O2 system goes with the airplane's mission.

I installed a cylinder bracket in the right side strake. The bracket is positioned so that an E / M-24 cylinder can be accessed and controlled by either copilot or by the pilot reaching across an empty seat. The bracket is secured by 10-32 clickbond studs, and the bottle is held to the bracket by 3" seatbelt webbing. I also have C / M-9 cylinders that can be carried in a shoulder bag and hung over the co-pilot headrest. Matched to a pneumatic conservator, the two sizes should be sufficient for 25 and 9 hours of supplemental O2, respectively. A fingertip device will track O2 saturation, and the flow rates will be adjusted to maintain saturation at or above 93.

  Jul 19, 2020     Transponder + ELT antennas - (62 hours)       Category: C22 Electrical

Transponder + ADSB + ELT = 3 antennae.

The Dynon transponder unit is "SV-Xpndr-261" and the antenna connection is a TNC male from the antenna locking onto TNC female on the back of the unit. The Dynon ADSB unit is "SV-ADSB472 and needs a BNC male connector locking onto the BNC Female on the back of the unit. The ELT unit is likely to be the Artex ELT 345 with a BNC male connector locking onto the BNC female on the back of the unit.

Along the way to my antenna install I took a few detours and learned a bit. I bought and figured out how to use a vector network analyzer - it was a fun way to master a slick piece of test gear. After carefully making 2 high tech annular antennae, I decided to use the simpler "ball on a stick", and then I made and discarded 3 different kinds of ground planes. I then removed and discarded my homemades in favor of something small, pre-made and inexpensive listed on Amazon (found by Del Schier). See the attached Word file for pictures. As sent from the manufacturer, it worked well only on 1090 MHz and the newer all-in-one transponders seem to require an antenna that works on both 978 MHz and 1090. Del created a simple mod and now it works on both frequencies with < 1.2:1 VSWR. "...I simply soldered .65 mm brass extensions onto both sides of one of the pair of dipoles and it now resonates at both frequencies. See attached photo of the mod and the VSWR plot. Note: the little triangles on the plot are at 978 and 1090." In place of soldering onto the circuit board, I used copper foil with a conductive adhesive. Per Del: "It should have the same performance as a 1/4 wave rod or any BIG $$ dual frequency TSO'd antenna and it is very light and does not require a ground plane." As installed, I have SWR's of 1.2 (ADSB) and 1.3 (transponder). Rather than make cables with SMA connectors, I'll buy those online.

Today's ELT's operate on both the old 121.5 Mhz and the new 406 Mhz (typically with a GPS locator signal). I did a Jim Weir-specified dipole with legs of 22.5" and 6.3" and installed it on back of the pilot seat. Looking at pictures of the Artex ELT 345, I see a BNC connector, so I made up an RG400 coax with one end a male BNC and the other split between center and sleeve, each soldered to the long leg of the copper foil tape. The short leg of the dipole was electrically connected with a conductive adhesive copper foil.

  Nov 29, 2020     Seat Belts - (46 hours)       Category: C24 Seats+Fair

Front Shoulder. The front seat belt mount points are birch plywood embedded in the foam, 1/16th aluminum under that, with a total of 7 (original 4 + plans revision 3) plies over the top. The work on the front shoulder points was done October 17-20 of 2005. The front seat has clearly visible hard points and it was straightforward to locate the nutplate and drill an appropriate hole through the top of the front shoulder support.

Front Lap. The front seat lap belt attach points were done 12/7/05. (See those entries.) The outside attach point are plywood hard points, aluminum angle, then 7 plies BID and uni, then drilled for bolts through the lower longeron. The front center attach is tubing in the seat back

Rear Shoulder. Set into the center section of the wing spar cap are the rear shoulder harness mounts, hardpoints made from spruce and embedded into the foam. Over the top of the hard points go 4 layers of unidirectional fiberglass layup. This was then reinforced with a plans change - add more plies. Here, both wing and center section spar were built by Dennis Oelmann and purchased by the prior builder. I energized a high lumen LED panel inside the spar and was able to see the wood inserts as shadowed areas. I confirmed this with a tap test - tap, tap goes the quarter, and you can "see it with your ears". The rear seat photos below are from another builder - Dennis did the work but failed to provide photos for the log. Dennis put the hard points at 7.25" and 12.75" from the center line. Plans locations are 5.25" and 12.5" from center line, while Aerocad spars use 5.5" and 14". I'll go ahead with the existing hardpoints, and modify if indicated. No nutplate - drill through the hard points and use an AN4 bolt to secure the belt hardware.

Rear lap. The outside attach point are plywood hard points, aluminum angle, then 7 plies of alternate BID and uni, then drilled for bolts through the lower longeron. The rear center attach is a 2" length of 20204 tubing is laid across the heat duct and covered with 7 plies uni and then 2 plies BID. After paint I drilled out the tube. Test fitting AN4 bolts I found AN4-32a was the length that fits the washers, seat belt end and lock nut.

Using auto racing belts. I ordered a set of new-in-box but out-of-date racing belts in a a 6-point configuration. 3" wide shoulder and lap belts, 2" leg loops. Blue that conveniently matches the upholstry fabric.
- The 6 points are lap, shoulder and crotch. After some trial and error, I'm going with 4-point, and not using the cratch straps.
- I'll install the rear belts for the certification, but most flying will be solor or 2-place with significant baggage capacity. I might make the rear lap belts only - lots of space between rear occupant and the back of the front seat. I have the parts ready to install if I change my mind.
- Auto racing belts are good for 24 months from date of maunfacture, while the same material in an aviation belt can be used until it fails to flex. The auto belts are assumed to live on a planet with 24/7 sunshine, while the aviation product lives in a world of perpetual darkness. It makes sense to change seat belts when changing hoses.
- I used a basting stich for the initial fit, then bought and learned how to use a sewing machine. There was a delay as I attempted to restore a Singer 237 (1960's), then a current model made the job doable. The formula for stiching seat belts uses the strength of the webbing, the strength of the thread and number of stiches per inch and inches of stitching. By using a triple stich with "overlock", the webbing is calculated to break (6000 lbs) before the stiches give way.

  Dec 31, 2017     Why Cozy & Progress - (20 hours)       Category: C01 Introduction

PK - My decision process took place in 2016-18. Things stabilized enough that I again had a long term residence with garage space. I still have the PA-12, but I like having a project. My track record was buying and restoring a BD-4, buying and restoring a PA-12, partnering in a RV-6a and EAA BiPlane, and volunteering on the EAA's replica of a Bleriot (steam bending ash!). In each of those I was a shadow and assistant to projects led by experienced builders in Hartford's EAA Chapter #166. I may have confidence greater than my ability, but that risk is balanced by openness to critical review, a willingness to redo bad work and a lot of patience. So long as I 1) listen to experts and 2) recognize when I've made a practice piece, I can do this safely. The decisions that took me to the Cozy were
a) find a fun and different plane that can use the IO-360 I have after selling off the BD-4.
b) flip through the KitPlanes index and narrow it down to a short list - something unlike the PA-12, with a good safety record and fun to fly. I ended up looking at RV's, the T-18, and oh yeah, the Cozy IV.

The final screening was looking for a suitable project. Many or even most kits are never finished, and the completion rate for plans built is said to be less than 50%. Therefore, the optimal move is to find and assume an unfinished project with great workmanship. I ended up networking into the Cozy builders group and its chief guru referred me to a good candidate project. Attached are pictures showing the condition of this project when I did the pre-buy (most parts made, stored uncovered in a dusty hangar) and when delivered to me. There are a lot of pictures, so the files are zipped.

As of 6/09/2023, my progress is as per the attached spreadsheet. Bottom line, some odds and ends, the engine install and then configure the panel. Add 100 hours for final paint and I'll be ready for a move to the airport and inspection.

The original builder had a write up on similar lines...

Why Build a Cozy MKIV? 04-26-2009

My father was a pilot both in the Navy and commercial airlines, and I've had the bug since I was a kid. I had planned on taking lessons and working towards my Private Pilot certificate whenever I had cash to spare.But, as I started planning out this dream, I ended up spending a lot of time thinking about the future and what that certificate would do for me.Rental rates are terrible, (barely) affordable Cessnas are slow and out of date, and there's just something about owning your own plane. I want to be able to go up whenever I like, without answering to another owner, timeshare partner, or invoice.

Buying an aircraft is a lot like buying a car. You have a number of choices, each with different features in terms of safety, comfort, and speed. The problem is that a 1990s Cessna with a decent (but not extravagant) avionics package can run over $100,000, and even a very old, 1960s Cessna can run in excess of $30,000. Mortgages are long enough that these are relatively affordable, but when I did the math between speed, age, features, range, and capacity, I concluded that I wasn't getting enough for my money. It took a while before I came around to the idea of building. I suppose I figured it was a "weird" thing to do - nobody in my neighborhood was building a plane. No, that was something those guys do out on the Salt Flats - you know the ones, the same guys on TV building trebuchets to fling washing machines or "model" rockets bigger than a car. The kind of thing you'd love to do, but you don't know anybody who has actually done it, and all your friends look at you oddly when you talk about it.

As it turned out, once I started asking around, everybody seemed to know somebody who knew somebody who was building, or had built, an airplane. Weird. I live in Connecticut, not Utah. We don't have any salt flats. What's more, I found that there were even more building choices than buying choices - kits, plans, quick builds, aluminum, composite, weird mixes of both, etc. I was overwhelmed by the choices I had to make. I spent literally months researching the options, reading manufacturer sales materials, Web sites maintained by other builders, magazines, and anything else I could get my hands on. In the end, I chose a Cozy MKIV for the following reasons:

1. Composite construction. There's a saying: Composite aircraft are built; aluminum aircraft are assembled. Well, I've never built an aluminum aircraft, but I love working with wood because it's so malleable, and composite construction feels much the same. It's very tolerant of slight variations between builders. Have a bit of a low spot? You can easily fill it. Parts slightly out of alignment? Fill with flox and BID tape over it, and it's just as strong. That's not to say you don't have to make things correctly. But you work the material, you carve the foam, and you shape the finished surface. It feels very much like sculpture. It's also very refreshing to see materials that are so weak by themselves (namely crumbly sheets of foam, cloth you can pull apart with no effort, and liquid glue) come together with such simple techniques into such immensely strong structures that you could quite literally park a car on them.

2. Safety. The Cozy MKIV is a "canard pusher," which uses a positive-feedback front airfoil (the canard) to lift the nose for pitch control, rather than an elevator pushing down the tail. This is not only very efficient, it's also very safe. By setting the canard at a higher incidence angle than the main wing, the aircraft can be made more or less stall-proof. The canard will stall first, dropping the nose back to a safe attitude before the main wing itself stalls. This produces a gentle "nose bob" effect. The clean lines of the aircraft also give it an excellent glide ratio (15:1!), and it climbs to altitude quickly (1500fpm). One can reach 10,000 ft in just a few minutes, and from there glide 30 miles during an engine failure. Name a Cessna or Piper that can do that!

3. Performance. In addition to reaching cruising altitude very quickly, the Cozy is fast. With a turbo rotary engine and a willingness to burn some fuel, one can easily cruise at 200kts+ with a maximum range of up to 1000 miles. That means it is actually possible to use this plane for realistic, cross-country flight operations. I have family and friends all over the U.S. and Canada, and love to travel, so a fast means of transport is a Good Thing (tm).

4. Cost. This has actually risen sharply since I started building. At the time I started building, the plans suggested a budget of $15,000 each for the airframe, engine, and avionics, or $45,000 total. But rising fiberglass, aluminum, and epoxy prices, plus inflation in general, have pushed this number much higher. Today I think it would be safe to double the airframe cost, and add at least a bit to the engine and avionics since few builders install the bare-minimum they can find. Still, for $60k (about the price of a mid-range BMW) you can get yourself a brand-new aircraft capable of 200kts+ at 10,000ft with 1000mi of range. You can even do your own maintenance. Those are good numbers in any book!

5. Support. The Cozy has a big following, and there are a number of good Web sites and mailing lists with other builders answering questions, and sharing information, photos, and experiences. When there's something you don't quite understand in the plans, it can be very helpful knowing that you can ask for clarification on a forum, or visit another builder's site who describes how s/he resolved the issue. Update: I started this project in 2005, and I'm now most of the way through the build. I've had a few "holds" placed on the project due to work, life changes, and moves, but I'm happy to say that the bulk of the airframe is now complete. With luck and hard work, I'll be flying in less than a year. After four years and over a thousand hours of build time, I'm now convinced I made the right choice!

  Apr 09, 2009     Rudders Rigged & Installed - (8 hours)       Category: C20 Rudders

Step 5 - Rigging Rudders

If you look carefully at the first picture you can see that I didn't do such a good job of setting up my rudder cable conduits the first time around - I had a few kinks, and while the cable did slide smoothly in the channel, I really wanted to redo this step.

Today I ripped out and sanded off the patches for the old rudder cable conduit. I also changed how I installed it. Last time I stood there like an idiot at each spot, holding it while the 5-minute cured. This was a huge invitation to holding it at the wrong angle (which I did). This time I put the cable in the conduit first, tied it to the pedals, the gave it tension at the firewall. This stretched the cable out so it was perfectly straight. Then I taped it at intervals along the fuselage wall so it stayed straight, and THEN I put 5-minute epoxy on it all at once, to hold it in place. I'll add some cover patches later - for now the 5-minute is fine. I did cover it where the map pocket goes on top of them, though. That would be very hard to cover over later.

2011-09-03 (4.50 hours):

Rudders Installed

I was initially concerned about my rudder travel because when I bought the wings from Dennis, he said that the hidden belhorns were interfering with the pockets that they sit in. Fortunately once they were mounted this proved to be a non-issue. They do interfere - but only after the rudders have reached their maximum required travel. The rudders are now installed and some of their micro is on.

Rigging Installed

This step went pretty quickly. When I got the wings from Dennis he had already set up the hidden belhorns. All I had to do was run the cables, swage the thimble onto the end of each one, and hook up the belhorns. The only reason it took so long was I spent a bit of time trimming excess micro around the rudders to let them move smoothly.

I temporarily installed the rudders with AN-3 bolts instead of the pan-head screws called for in the plans. Those are a pain to get on and off - I keep wearing down their heads with my junk screwdriver.

PK note 6/7/2020. I received the project with all control surfaces removed and the trim control removed from the fuselage. All control surfaces are now finished and ready for a second round of installation. The build log does not show rudder stops installed - further inspection and work may be necessary.

  Sep 16, 2020     Pitot heat - (4 hours)       Category: C13 Nose+Gea

The aircraft pitot tube runs from the nose, along the left wall, via an aluminum hard tube. I installed a 12 volt, 12 watt resistive heating tape, contacting the face of the tube as far forward as possible within the nose compartment. In addition to heat conducting adhesive, the tape is secured with 3 safety wire wraps and does not touch the fiberglass structure. The location is not subject to motion and this is fully adequate for securing the heat tape. Since the aircraft is _not_ intended for flight in known icing conditions, a short term test was useful information.

I used a digital kitchen temperature probe of known accuracy. First, I tested the temperature of the tape itself. Under 13.5 volts it heated to 180 degrees, with probe in direct contact with the heat tape. Second, I secured the tape as described above - the tape was tied to the aluminum pitot tube and did not touch the epoxy structure. The temperature probe was placed inside the pitot, touching the side wall at a contact point with the heat tape. At 13.5 volts the tape heated the mouth of the pitot, first from 85F garage temp to 127 (3 minutes), then climbed slowly to 141F over 15 minutes. The adjacent composite surface heated to only 115F. I was surprised the glass temp did not go higher and repeated the measurement with the same result. The heat tape came with 24 AWG leads, which were secured with D-sub pins to 22 AWG shielded twisted pair wire. Each pin connection is secured with shrink tubing and the pair covered in a larger piece of shrink tubing. The draw is 1 amp. Pitot heat is switched at the panel and separately fused at 3 amps, giving the wire significant protection. I again used DB pins for the connections. DB pins are rated for 5 amps in low density loads. This load is significantly less than the rating of the DB pin connection.

Because the heat tape is an untested building practice, I will test again when measuring temperature and amperage draw of the completed airplane.

  Feb 21, 2009     Main Gear - building the hoop - (30 hours)       Category: C09 Lnd Gear

It took about five days to do this step, over the course of two months. I decided to go with Wayne Lanza's method of reducing the main gear drag -- we'll see if that pays off.

The gear wraps went as planned, then I installed the brake line tubing. I used tubing large enough that I can slip my brake line down, then covered the trailing edge with foam blocks. Following Wayne's method, I set the gear to the correct angle on my work table, sanded the foam blocks until they were vertical, found the centerline for the new trailing edge, then cut and shaped the blocks into the new airfoil shape. It looks more or less like I expected it to...

I then did the trailing-edge layups down onto the face of the gear strut. Those are visible in the first two pictures below. After trimming and sanding the new trailing edges (I made a cardboard template so I could get them as close to identical as possible on each side), I then did the leadingedge layups today. It took a bit longer than I expected (I initially thought it was two UNI, not four) but wasn't exactly hard. The results are in the second two pictures below.

The plans normally call for the strut to be up on nail heads at this point. I had trouble getting it to stay on there while doing the layup, so I yanked them out and threw foam blocks underneath. I did tuck the layups around the bottom of the strut where the foam blocks are, so the wraps are complete. It was just easier this way.

Honestly, all this work to make these fairings seems a little pointless because the rotary engine can basically bludgeon you through the air even if you DO have extra drag, but perhaps it'll be worth it in a slight fuel consumption savings.

  Jul 02, 2020     dummy engine for the build - (12 hours)       Category: C23 Eng+Cowl

January-March I had an O-235 engine overhauled at Zephyr Engines, in Zephyrhills, FL. I happened to be in the right place at the right time and was loaned a dummy IO-360 engine which they had stored in a back room. (Seems they used it in recruiting for a local technical training institute.) The dummy is the IO-360-C1E6, where I have an A3B6D. Both are wide deck / angle valve engines, and the dummy is an appropriate tool for the build. Based on an online version of the Lycoming engine codes (http://www.meyette.us/LycomingEngineNumbers.htm) both have
- Bendix fuel injection, tuned induction
- a prop governor on the left side of the crankcase,
- the Bendix D4LN-3000 impulse coupling dual magneto,
- counter-weighted crankshafts

Note that I will be moving over to electronic injection and electronic ignition, so I avoid the single point failure mode of the one magneto shaft.
A big difference is that the dummy engine has a rear-mounted air intake. It's a feature I would like to have, since it means ram air when mounted on a canard. Converting my enigne to rear air intake is something to look into.

  May 15, 2023     C18.23 fresh air vents - (3.5 hours)       Category: C18 Canopy

The plans give the shape and location - see attached. The plans location won't fit the Cozy Girrrl extended strakes, and I adjusted to just below the strake and below the trailing edge of the canard.

An alternative to the plans dimensions is using the Van's NACA scoops, and I bought a set. (It comes with NACA scoop, adapter and hose - can I use the extra parts for cabin heat?) That scoop ends in a 2.0 inch outlet which needs to map to a 1.5” mating collar on my eyeball vents. I made some custoom scoops, then some large-ish adapters, fussed with a high performance RTV, then went simple - a 2" plug goes in the CAT hose, with a 1.5" cutout that fits the vent. The plugs are 2 layers of BID each side of 1/2" blue foam, with nested circular cuts made with my hole saw kit. The ends are secured with standard hose clamps.

(This is another instance of where kit builders save a heck of a lot of time - they just install the supplied parts. Yes, I could have followed the plans, but then I wouldn't be building with enlarged strakes, et al.)

  May 09, 2023     CO detector - (1.5 hours)       Category: C22 Electrical

Aircraft need a CO detector. The cool new thing is active detectors like the "CO Guardian" ($640) or the newest, a CO detector embedded in the headset. I went with the most common solution, a $4 2"x2" card with a dot that changes color in a CO danger condition, replaced at annual.

  May 16, 2023     reference - torque values       Category: C23 Eng+Cowl

Below are reference tables for torque values. Where reasonable I write the torque value next to the fastener.

  May 13, 2023     Tires - (2.5 hours)       Category: C13 Nose+Gea

Tires are
mains - 500-5 Desser Monster Retreads, 8-ply, inflated to 60 psi
nose - 10x3.50, Tire Specialty of America, 4-ply, inflated to 35 psi

Note - at full inflation and tirtes as new, the bottom of the prop extension is 45" above the level surface. After moving to new engine mounts, __ inches.

  Jan 13, 2021     14. Firewall - cabin side - (12 hours)       Category: C22 Electrical

This design certainly has evolved - a dozen drafts of how to arrange the components on cardboard, 3 ways to fix components to the aircraft, 4 versions of how to make the removable panel. One of the very last changes came from Ross Farnum at SDS. Seems it takes 6" of clearance for the cables going into the EM-5 electronic ignition modules. I did a last minute juggle to create a clear field for the cables. No, a right angle DB25 won't work. I think all is final, and I'm drilling for the fasteners.

On the fastener map, the green dots are click bond studs, a 10-32 thread and the pink X's are 6-32 rivet nuts. Both use 3/16ths holes in the aluminum, drilled and deburred. I want to prevent turning of stud or rivet nut. The click bond studs were notched and the notch aligned with a 1/16th hole drilled in the aluminum, then set with JB Weld, taped firm against the aluminum during cure. The holes for the rivet nuts are lightly notched before the squeeze, aiming to reduce risk of turning. Back to the fastener map - the purple hex shapes are the through bolts that use existing firewall penetrations. I will put the head of the bolt on the engine side and secure from the cabin side, with AN hardware and thread locker.

Contactor basics and wire size. Contactors are powered switches that control high amperage flows. The battery contactor draws 0.7a continuous, with power flowing from the battery terminal, to a post of the magnet wiring, around the magnet, out through the second post and then through the battery master switch to ground. The starter contactor draws 3.5a intermittent, flowing from the battery to the starter button to the starter contactor, then out through the contactor case to the battery ground. For ease of recognition I used red and black twisted pair, in 16 AWG. Over the 14' run from panel to firewall the 20 AWG is rated for 15a intermittent, 7a continuous.

  Jun 14, 2020     00. 12v, dual alternators, single battery, Nuckolls Z101 - (25 hours)       Category: C22 Electrical

Wiring this airplane will involve a lot of learning. I'm hoping that acknowledging ignorance will protect me in this part of the project. And maybe I'll come away with a better understanding - the FAA says I'm allowed to do build an airplane for research and education. Here's hoping some education happens.

Reliability is darn important. The A3B6D has two mags mounted on one shaft - it has only a little more reliability than a single mag system. I used this engine (on a BD-4) with that mag + a first generation electronic ignition system made by Light Speed Engineering (Klaus Savier). For the Cozy I plan dual EI and the more modern SDS system - weight savings, more powerful spark, said to be more durable, etc.. Dual EI means that if I lose electricity I lose spark, and I'm guessing the Cozy is a poor glider. Therefore, this all-electric airplane needs a fault tolerant electrical design.

Like most builders, I stand on the shoulders of Bob Nuckolls - 50+ years of designing electrical systems for aircraft and probably 30 years of sharing the information through the AeroElectric Connection (first a book and then a forum). Bob recently refreshed his schematic for all electric aircraft - the "Z101" using dual alternators and a single battery. See the drawing below. It is his most recommended design and my starting point.

Continuing the homage to Bob K., a snippet from his "getting Started" (in full at http://www.aeroelectric.com/articles/Getting_Started.pdf):
__________________________

RULE 9A:
A single point ground system shall be established behind the instrument panel with sufficient attach points
for all accessories in the cockpit area. In deference to RULE 2, a forest-of-fast-on-tabs ground block similar
to . . . http://www.aeroelectric.com/Pictures/Grounding/gnd_bus.jpg

The threaded stud on the ground block assembly would penetrate the firewall and be used to terminate
battery (-) leads on either side of firewall and the crankcase ground strap on the engine side of the firewall.
In the case of canard pushers with the battery up front, the ground bus would be mounted forward of the
instrument panel. If the airplane's firewall is metallic, then a brass bolt and appropriate washers and nuts
would be used to provide an engine compartment ground stud and connection of the ground lead to the
firewall. A ground strap like . . .http://aeroelectric.com/Pictures/Wiring_Technique/bbs.jpg or http://www.aeroelectric.com/Pictures/Wiring_Technique/sbl.jpg
. . . will be used to connect the crankcase to the firewall ground stud.

The welding cable illustrated in the photo would also be used to make the short, VERY flexible jumpers from battery (+) and (-)
terminals.

Any ground straps provided around the rubber biscuits of an engine mount will be removed. Engine mounts
are for holding engines on airplanes and not use for any part of the electrical system.
-------------------------------------------------------
RULE 9B:
Avionics and other electrowhizzies on the panel would benefit from an “avionics/panel ground bus” as
described in the latest revision of the ‘Connection. See:
http://aeroelectric.com/articles/Rev11/18Audio_R11.pdf
and . . .
http://aeroelectric.com/articles/Rev11/AppZ_R11J.pdf
and illustrated in . . .
http://aeroelectric.com/Pictures/Grounding/AGB_V.jpg
http://aeroelectric.com/Pictures/Grounding/AVG_RA.jpg
http://aeroelectric.com/Pictures/Grounding/Avionics_Bus_3.jpg
http://aeroelectric.com/Pictures/Wiring_Technique/Avionics_Ground_3.jpg
--------------------------------------------------
RULE 10:
Tefzel wire used throughout with the exception of cranking circuit fat wires where 4AWG or 2AWG
welding cable would be used. An alternative FAT wire could be one of the new copper-clad aluminum wires.
These new materials are as solderable and crimpable as pure copper conductors.
_______________________________________________________

In the next entry I will repeat the electric system build log created by Chad Robinson - he's a skilled engineer and his work deserves consideration. Also, it may help me identify circuits in the tangle of wires I see in the cockpit!

  Apr 06, 2023     95. Manuals       Category: C22 Electrical

see the attached.

  Feb 07, 2021     97. Expansion bus - (8 hours)       Category: C22 Electrical

One fun thing - I ran and will set aside a 12-line expansion block, running firewall to instrument panel. I bought 5 meters of 12 conductor shielded cable, with tinned copper conductors. The individual wires have a conductors with a cross section of 0.75 mm, just slightly bigger than 19 AWG. It's not a perfect choice since individual wires are shielded in PVC rather than tefzel. Bob Nuckolls notes that PVC was used in a generation of Cessnas and he sees it as appropriate for use on the cold side of a firewall. That works for me. One end of the cable was wired on the bench, the other standing at the side of the airplane. I made it once with the full length of the cable, figuring I would have a big service loop. I thought better of it, and cutting to length saved 11 oz. Tested for continuity and cross feed and no problems.

The rear end is at the lower left of the cabin side of the firewall. The forward end is on the passenger side of the panel, secured to the panel with rivet nuts (great when working the aluminum!). I built and installed the one end, then figured out I needed to set the path of the cable to travel alongside my 4 AWG runs. Solution? Chop it off and do it again.
The shorter cable did save 11 oz., and I understand it takes that kind of weight discipline to make this project a success.

  Aug 07, 2021     05 switches - (4 hours)       Category: C22 Electrical

Hardware.

Carling toggle switches, with actuators of either the nylon paddle or steel bat. The switch guards protect
- switches that would affect engine spark
- switches that control the fuel pumps and
- the defeat of the autoextend module for the landing gear.

Battery Contactor / "Master". Battery contactors are the "master" switch, and I used the standard wiring - the panel switches the ground while power is a direct connection from the battery post. So long as the ground is connected the battery will flow 18 watts of power to the contactor. The electromagnet will stay in the raised position and the contactor will stay closed. The panel switch is off-1-both, where the #1 circuit completes the link to the battery ground. I used 16 AWG for the ground connection (a red & black twisted pair) at the switch - on an all electric airplane this is a peace-of-mind matter. The "both" position brings the alternator online by sending the field signal to the alternator. Note that another switch, described below, changes the power source to the backup alternator. Alternator field is a 3a circuit.

Start Contactor / "start button". Using electronic ignition I have a start button and a 1-2-both switch for coil packs. The starter load is intermittent, and that FAA chart allows for 20 AWG. I made a red (starter) & black (battery) twisted pair for the contactors, using 16 AWG - on an all electric airplane this is a peace-of-mind matter.

Buss fuse blocks usign ATC style blade fuses.

  Oct 11, 2020     03 Panel - drafts and discards - (58 hours)       Category: C22 Electrical

I learn and re-learn the common sense stuff:
1) Know and comfortable with CAD? You're way ahead of me. In my case, I need to stay on paper and cardboard for as long as possible. But if y ou can outsource panel cutting with a reliable CAD file, it would be a lot easier than my 8 iterations (paper, cardboard, thin plywood and then aluminum).
2) Aluminum is better for panels because of the design of the avionics mounts. 2024 is everywhere but bending 6061 is lot easier than 2024.
3) The project I acquired had the standard madness of a composite panel cut to spiderweb by all the instrument holes. A spiderweb cannot add strength or stiffness to the structure. Now a radio stack ties the panel to F28 (fore-aft stiffness) and a bit of square tube runs across the panel bottom, tying together the right and left fuselage side.
4) If I knew for certain what my layout would be, I might make it modular. But maybe not – most of the wiring is across the natural panels, and it's just as easy to pull displays as to pull a module.
5) I should have created a wiring channel before I wired anything other than the 4AWG power and ground cables. I clipped open the ugly bundles and am doing it over, threading wire through jumbo Adel clamps hanging on the bottom of F28. The upside is that I have yet another chance to test all my connections.

I worked the last wood template near to death - when I made mistakes I patched it with flox and tried again. I found and solved all my clearance issues and changed the layout multiple times. With the wood perfected I started working on aluminum sheet, first with .063 2024-T3 then .050 6061. See below on aluminum alloys. I worked through multiple colors before setting on flat black with labels of white text on black. Lesson - don't trust "paint and primer in one" over any existing finish and don't spray a clear over anything from another family. To secure the panel to the composite perimeter frame I like AN3 bolts and 10-32 machine screws into 2-lug nut plates, floxed to small pieces of G10 and topped with a layer of fiberglass tape. These suckers hold, and the compressive load is the aluminum sheet on one side and the G10 on the other.
________________________________

Side bar on aluminum alloys....The former panel was aluminum alloy 2024 - readily available and easy to cut and drill. Even so it is easier to work with the recommended alloy, 6061. After the rolling mill 6061 is a lumpy mix of magnesium, silicon, and other elements in aluminum. Solution heat treatment is done by raising the alloy temperature to about 980 degrees F and holding it there for about an hour. This dissolves all the alloying elements into a solid solution in the aluminum. Then quench the alloy in water - cool it rapidly enough to prevent the alloying elements from precipitating on cooling. at room temperature. This is called the T4 temper. If we take this material and heat treat it at a temperature between 325 and 400 degrees F, the alloying elements begin to form ordered arrays of atoms in the aluminum matrix. These arrays are called GP zones, and they strengthen the aluminum considerably. This is T6 temper and how I bought it. Heat it again, the ordered arrays break down and the sheet bends easily. Now the clock is running...an age hardening process takes place at room temperature over a time period of four to five days, with 90% of the hardening occurring within the first day. Because of this effect, aluminum parts often need to be shaped soon after a heat treatment process.

Here's the mechanics of annealing in the home shop. Mark the bend line, then use a candle to put light soot on that line. Think through the bending jig and handling hot sheet, _then_ use another heat source to burn off that soot line. I have a propane torch, and w/o an O2 feed it burns at 2000 F - I play the flame over the bend area to raise the temp gradually. The soot line burns off at 700F, and then I use a spray bottle to quench. I clamp along the bend line and the annealed 6061 bends with hand pressure. To get a crisper bend I use wood blocks and soft hammers. It aircools over 5 days, and is reported at 90% strength within 24 hours. I understand this is a T5 condition, not T6, but has the rigidity I want.
________________________________

For layout I repeatedly used the Cozy template in X-Panel software, muttering all the while. I do not recommend X-Panel - expensive and the ability to export a CNC file is even more expensive. The template doesn't fit this airplane. The outline view is _not_ a cutting guide - it will show the instrument's total dimensions but not the panel hole.

Below is a template I found useful for my steam guages. Print it (use the "original size" option, paste it on a bit of aluminum sheet, market the center and corners, and outer edges of the standard instrument. Save it forever!

  Apr 22, 2023     Gap Seals - (1 hour)       Category: C99 to sort

Took me a while to find where to get gap seal tape. Turns out it's a multi-step process. See the attached pdf's.

  Apr 15, 2023     Safety       Category: Z-Tools+Tips

Canopy exit tool. Cozy pilot Don Berlin had P-mag failures which resulted in a controlled off-airport landing. The airplane flipped and Don was unable to kick through the passenger side canopy. He found a crack in the pilot side canopy and was able to escape through that. He recommends carrying a tool for emergecy exit.

  Feb 06, 2022     Shop-made parts - tips and tricks - (1.5 hours) Category: Z-Tools+Tips

Z-fastners to fit concave and convex surfaces. I make custom brackets and braces as wire conduits and tie downs. Understand that the strength of that attachment is no more than the layer it is attached to. That means you want to use them before paint, or make sure to use the abrasive wheel and get a clean surface for the epoxy to grab on to. Thickened epoxy is my go-to for attaching Z-fasteners and mounting studs, and West Systems G-Flex comes with silica mixed into each component. It holds really well, won't run off a vertical surface, and does a nice void fill under my custom fasteners. After mixing my own and then buying tubes, Santa brought me the 1 qt cans. A little goes a long way.

Gust Locks. The coolest looking ones are being 3d printed with some complex rope work. I went with the lightest that can do the job.... 0.025 aluminum, drilled for 3" lengths of vinyl tubing, painted a painfully cheerful red. Add a ribbon and the job is done. Elevator, ailerons and rudders - it takes 6 to do it right.

Nose seal retainers. I knew lost of anose seal could happen, but it was darn rare. I don't have a super large circle of Cozy builder friends, but 2 of them have reported the seal was either blown or about to go. I grabbed the template file from the Cozy site (thanks, Russ Meyerricks) and cut two from scrap aluminium (1/8 inch 2024). I'll do the match drilling when I next have the prop off.

Link for starter-alternator. A tab of 1/4 steel bar stock, drilled to fit between the starter and the alternator pivot mount. Both items bolt onto the engine, and both are further stabilized by the connecting tab. You can buy the part at B&C ($27). In my case, a local steel dealer laughed at how little I wanted, measured and cut a piece, and sent me on my way. It is mild steel, so it was easy to round the edges and corners, then drill for the two different bolt sizes. Next time I have it off I'll paint it grey to match the engine. FYI, if one has the narrow deck parrallel valve engine the alternator mounting bracket attaches to the engine using both the forward and the rear mounting bosses. That means the bracket is better bale to handle the torque of the alternator belt, and the link is typically not used.

Fuel line sealants. EZ Turn has been the standard fuel-proof lube for a long time. It is incredibly sticky and unreasonably expensive - luckily one tube lasts a lifetime. Another Cozy builder shared the idea of putting the barest film of EZ turn on the flared faces of fuel fittings. I had reason to assemble and disassemble multiple times, and on disassembly found the slightest tackiness to the surface. I think it means my perfect looking flare had a slightly imperfect contact surface with the fitting. EZ Turn makes it just that much better. I prefer Gasoila as the thread sealant for fuel lines. A small tube is more than enough for the project.

  Apr 08, 2023     96. Regulator test rig - (2 hours)       Category: C22 Electrical

Following instructions posted on AeroElectric Connection, I used an automotive regulator to build a test rig. If and as I suspect a regulator problem, I can easily put this in place of the installed one, then test the resulting performance.

  Mar 13, 2021     02 panel - patch & repaint - (6 hours)       Category: C22 Electrical

Getting the labels and switches just right was the path to seeing that the panel color was too dark. I loved the hammered bronze look, but you have to look 2x to read the label, and that won't work. And a poorly fitted notch for the altimeter adjustment could look better, so let's get it done. I removed the panel and instruments (30 minutes?) and used a flox / G-Flex epoxy mix to fill the mis-cut. See the circled area below. Once it cures I'll carefully file the notch for the adjustment knob, and then move ahead with the repaint.

FYI, I'm going with a light grey / dark silver. That goes well with the painted interior and with the labels.

Now in April 2023 I'm recutting the panel. I want a more standard layout, and that means
- engine and aux alternator switches separate from lights and accessories
- lights & accessory switches in the normal location along the top edge, moving engine and aux alternator switches to left edge
- built in red night lighting along top edge (probably a piece of right angle riveted to top edge)
- bottom edge extended full length, with the leg hole cut outs folded back for a longer shelf
- fuse blocks moved to the longer shelf
- round guages in a column to the right of the radio stack
- refinement in the switch alignment - using a blueprint sized print of instruments on a standard Cozy panel
- 6061 alloy! The 2024 was hell to bend and this should be much easier.

I'll change the paint scheme one more time - a flat buff might work well.

  Nov 17, 2020     83. Shore Power - (64 hours)       Category: C22 Electrical

I installed a "Piper" shore power socket, and wired the matching plug to a set of jumper cables. Below you can see the Nuckolls article that describes how to wire the socket, which includes another continuous duty contactor and "crowbar" circuit. In addition to what Nuckolls described, in place of crowbar circuit that blows the fuse, I went with a smart voltage detection circuit, such as would be used in a modern control system. The gist of it is
- flip a toggle switch and the voltage detector runs with a power drain of 3 milliamps - on par with a quartz watch.
- The inserted plug connects to the socket, which has its voltage detected and compared against low (9v) and high (15.7v) thresholds. This is protection against a battery cart trying to send a GA-standard 24v.
- If between the two values, the detector closes a (normally open) relay and power goes to the contactor. Power flows to the battery.
- If and as socket voltage spikes, the relay opens, the contactor opens and the socket is disconnected from the battery. Any fingers inside the socket can get pinched, but not shocked.
- As a final protection, when the toggle is "OFF", there is no power to the voltage sensing relay and it goes back to its normal open, which removes voltage from the battery contactor.

I've found that sometimes I don't want to bother with the socket and I just use a batter charger or jumper cables. In that case, positive goes on the battery side of the contactor and ground clips to the socket cover. Seems to work quite well...

If I were doing it again I would install the unit in the forward hatch - easier access, no need to get the cabin wet on a rainy day, and further from any prop.

  Oct 11, 2021     LIDAR - (25 hours)       Category: C13 Nose+Gea

The Zeitlin auto extend unit triggers nose gear extension on a combination of signals - low speed, throttle closed and near the ground. The "near the ground" bit is handled by a LIDAR unit sold by LightWare of South Africa. The programming of unit was straightforward - below 85 meters it sends a 3.3v signal to the auto extend unit. (I'm hoping this voltage is sufficient - the alternative setting is 2.2v.)

My main power runs on the right side of the aircraft: power and ground travel from firewall to instrument panel. This wiring runs on the left side. The installation instructions are specific about providing a clean power supply and routing the wires away from main power lines. I think any EMR sensitivity is about the signal from the LIDAR serial output, which reports altitude 20x per second for use in drone autopilots. Still, I routed the wires in the pilot side (signal side) bundle and put a power conditioner along side the unit, setting it to 5.00 voltage and measuring the voltage numerous times. 12 volts in on 22AWG twisted pair, and the 3.3v alarm signal on 20 awg unshielded.

The fiddly parts of this task were fitting and re-fitting the altitude senor, and the nuts and bolts of learning enough about circuits and relays to have a reliable device. Note that the plans do not include the power converter you need at the LIDAR unit - you have to deliver power to the unit, then step it down to a precise 5v. I bought an adjustable step up step down converter (single circuit board) and adapted an enclosure - seems to work well. I did the step down at the gear leg mounted LIDAR unit - too much voltage loss sending current front to back. Just put the converter next to the device and all is well. Per mfg., max power consumption of 200 mA (0.2a). See manual.

  Apr 03, 2023     Notes - Cautions - Warnings Category: C23 Eng+Cowl

From the Cozy Builders Group, a recommended secton of the manual:

Notes, Cautions, and Warnings

WARNING – This aircraft uses an electrically-dependent igntion system. With any electrical failure, reduce electrical load and land as soon as possible.

WARNING: Takeoff and land with both boost pumps on. After takeoff, either boost pump may be turned off.

WARNING: Brake lines are to be replaced every 5 years at the Condition Inspection. Brake fluid (5606) is to be replaced no less frequently. Brake fluid absorbs moisture which lowers the boiling point of the fluid and contributes to caliper corrosion. Bottom-filling of brake lines is recommended. Suction old fluid out of the reservoirs with a large syringe and place the extended nose gear on an elevated platform. Pump new fluid in at the brake caliper until fluid appears clean and free of bubbles.

CAUTION – If the nose inadventently rises more than a few inches, a tipback can happen quickly which will damage wheel pants and prop. When parked on three wheels with the cockpit empty: i) insure the 25 lb weight is hung in the nose gear; ii) lower the nosewheel if parked pointed into a strong wind or on an upslope. Apply the parking brake when parked on a downslope.

CAUTION: The nose gear mechanism must be fully extended over-center or fully retracted. If weight is placed on the nose in any other position, it may strip the gear in the mechanism. A stripped gear can be reversed for one more usage.

CAUTION: For safe stall behavior, this aircraft requires 16# of ballast placed ahead of the rudder pedals with a 225 pilot. Cast lead weights fit in the compartments under the nose hatch.

CAUTION: To prevent nose gear damage, avoid taxiing into potholes or ground-wire depressions on the ramp.

CAUTION: Keep the swiveling nose gear adjusted so that that a hard blow with the fist against the tire does not freely deflect the tire, otherwise, the nose gear may shimmy and destroy the nose gear.

CAUTION: There is no warning of an unlocked canopy in this aircraft. Insure that the canopy emergency catch is bent so that it engages when the canopy is lifted.

CAUTION: Leaning during ground operations is recommended, however, lean so that if the throttle is inadvertently advanced for takeoff with the engine leaned, the engine will die.

CAUTION: In the event of alternator failure, lowest current drain will be achieved by selecting E-bus power, then turn off the master switch, then selectively turn off any E-bus-powered equipment that is not needed (i.e., com radio, transponder, D-10A, one fuel pump). Turn off Lightspeed Ignition #2 to save the backup battery for last use.

NOTE: Before removing fuel tank caps for refueling, attach the fueling ground wire to the step. The step is bonded to each fuel cap ring via internal wiring.

NOTE:: For normal operation, maintain Ignition 2 Pwr Select in the “aux batt” position. This checks the condition of the aux battery.

NOTE: Lightspeed ignition #1 is powered directly from the main battery and #2 is normally powered from the aux battery, each through its own switch. If alternator failure occurs, reduce any other electrical load and turn off one igntion (normally turn off #2) until the main battery is exhausted, then turn on ignition #2

NOTE: Spare fuses are stored in a red bag in the right strake hole.

NOTE: If the fuel valve binds, position the aircraft nose-high, disassemble and lube the spool with a tiny smear of Fuel Lube. If too much Fuel Lube is used, bits of it can lodge in the gascolator screen (where it will not desolve!).

NOTE: Tire pressures are 50# main wheels, 65# nose gear. It is particularly important to keep the nose tire inflated to avoid turning on the rim at touchdown.

NOTE: This aircraft uses DOT3/4 brake fluid and EPDM o-rings in brake hardware.

NOTE: The Dynon D-10A EFIS and uMonitor are powered via the E-bus and will indicate approximately one volt less than battery voltage due to loss in the E-bus diode.

NOTE: Water is seldom found in the tank or gascolator drains but if any water is found in the tank drains, then also check the gascolator for water.

NOTE: At high cruise speed and wide-open-throttle, the throttle may need to be retarded slightly for smooth running and peak RPM.

NOTE: This Ellison carburetor set-up exhibits a slight engine surge at medium RPM.

NOTE: For instrument access: remove the canard cover, remove two long pins securing the instrument cover, remove instrument cover and disconnect the connector for cockpit lights.

NOTE: To remove the canard, remove the instrument cover as described above, remove the footwell access hatch, disconnect the elevator control rod quick-disconnect (starboard side), loosen two screws holding the trim spring wires in the trim mechanism (port side), disconnect Dynon remote compass box, unscrew & remove two screws through the canard incidence-setting tabs, unbolt the canard lift tabs, lift the canard up and out.

  Mar 05, 2022     IO-360 sumps & the Cozy - (50 hours)       Category: C23 Eng+Cowl

My Cozy was built using the engine I had on the shelf - an IO-360-A3B6D. The codes mean my engine left the factory as
I Fuel injected
0 opposed-
360 360 cubic inches of cylinder volume –
A type A sump
3 propeller locating bushings rotated 120 degrees clockwise
B with an accessory section using Bendix accessories
6 (1) sixth and (1) eighth order crankshaft counterweights
D (1) Bendix D4LN-3000 impulse coupling dual magneto

Attached is a write up of modifications made to fit this engine to the Cozy.

The angle valve wide deck sump is above the induction chamber, with wings to the right and left. Both wings are tapped 1/2" NPT for oil drains. The standard Lycoming plug has a square head. On one side I installed a stainless plug that uses a recessed hex head / Allen key. On the other I installed a Fumoto engine oil drain valve (p/n T202N) with the lycoming engine thread of 1/2-14 NPT. Spring loaded and brass, vs the standard springloaded and aluminum. This is brass going into aluminum - a strong hand tight is sufficient! Use safety wire for peace of mind. I also have a 1/2" id plastic tube and a solid plug. To drain oil, roll the left wheel onto a ramp, attach the hose, put the drain end in the pan, remove the hose plug, and flip the switch.

  Mar 31, 2023     Exhaust - (2 hours) Category: C23 Eng+Cowl

I bought the standard exhaust for the Cozy.

Here are tips for the flanges (I use the ones made by SDS), studs and nuts.

“While conducting GA and Experimental airplane inspections, I have found many exhaust leak(s) at the interface of the cylinder flange and exhaust stack. Here's some bullet points for your consideration:
1. Ensure the nut that you choose has a High temperature lock device. The Continental part number is a plated-high temp all metal stop nut. I've seen corrosion issues with this particular application and this is not my first choice.
2. Find a system that is corrosion resistant. G#^*'s aero used to send back OH'd cylinders with CAD steel stop nuts and CAD plated washers. In a very short period of time, these would corrode away and allow the stack to loosen and the gasket to be breached by hot jets of exhaust.
3. If a washer is needed, use a stainless or corrosion resistant option.
4. Ensure your studs have a proper interface with the cylinder. If a stud comes out with the nut corroded onto it; Don't be tempted to simply
screw it back in. Replace the stud and ensure you have proper rotational resistance during the process. Oversized studs are available and there's an ID marking on the end of the stud alerting the Mechanic of the existing stud size/oversize.
5. I favor the large profile Continental brass exhaust nuts - like the ones you find at Spruce. The lock is a "stake" or slightly out-of-round scenario. These nuts work very well and are corrosion resistant. Also, they tolerate re-torquing. However, they are NOT reusable. Once they're removed, the threads will be severely compromised.
6. I recommend spending a few extra bucks and getting the No - Blow gaskets. They will last the life of your engine and can even be reused on your next engine.
7. Lastly, address any erosion issues in aluminum flange of the cylinder before assembling your exhaust stack. If it's a used engine, there's a good chance you will find erosion from a previous exhaust leak.
Dave Ronneberg (Berkut designer) prefers - using Stainless 12 sided nuts which are lower profile and easier to put a socket to and seem to resist backing off. No more corrosion that so many exhaust nuts exhibit – sometimes locking onto the stud which results in the stud coming out of the head. He agrees on the blo-proof gaskets being the best – many home-built Lycoming pipes start their "turns" close to the head, restricting space for tools.

  Feb 06, 2021     cowling - Boat Tail or No? - (8 hours)       Category: C23 Eng+Cowl

On 2/5/21 I heard Klaus Savier endorse the boat tail lower cowling. Changes to the plans burn up time, but the cowling is already a custom feature. Here's an explanation and picture of a "boat tail", then a discussion of baffling. I'll be cautious in planning, but it's where I'm headed.
__________________________________________________

The plans version of the cowling involves a large curve that results in separated airflow. Somewhere along the way, someone came up with the boat tail. The idea is that by adding a tail to the back the slope of change can remain at or under the 3 to 1 ratio (for every three inches of length the curve only changes 1 inch or less). This allows a more gradual change in the contour which keeps the airflow connected - see the tuft testing photo below. Then someone went further and moved the exhaust to this location as there was space there and it smooths airflow over the cowling.

Nick Ugolini used an eductor in a standard cowling, and perhaps that can be used with both exhaust and with air exiting the baffles.
_____________________________________________________

Air through the cowling - tractor vs. Canard. A typical tractor-form airplane (think Cub) has pressurized air come through the nose and held tightly over the cylinders before exiting to the lower cowling and out a large and generally neglected exit at the bottom of the cowl. In a canard the air enters at the bottom of the cowl pressurized with intake air from the NACA. That air flows up through the cylinder fins, constrained by baffles, which guide around the cylinders and into a pressure box open at the rear. The air exits the cowling through largely unrestricted openings

Background - plenum & eductor. To increase pressure differentials, a lot of time and trouble goes into constructing tight baffles, with rubber / neoprene strips pressed tight against the cowl to hold air in and increase pressure differentials. When that approach is exhausted, people use plenums - see the BD-4 plenum below. A plenum is a pressurized chamber containing cooling air and forcing it over the cylinder fins. But I have an updraft cooling Cozy and it is darn hard to shape a plenum for the bottom of an engine - the induction pipes and oil drain, are hard to work around. Is there any benefit to a top side plenum with an updraft engine?

I have never seen one on the hot side, but perhaps it could be called a modified eductor. (Wow, straight from the Latin educare, "to lead out"!) Below is a picture of the Nick Ugolini "eductor". He used the eductor to exit exhaust gases, with the cylinder heated air exiting through ducts on the top of his cowling. (There is a slightly more detailed description of eductors at https://groups.google.com/g/cozy_builders/c/-YRfovdF1-E/m/j-1-d7pkOUUJ.) Could an eductor be used in a boat tail cowling - the cooling air exits through the eductors on the right and left side of the prop, while the exhaust pipes exit below.

In this approach the NACA is as well done as possible, pushing into the cowling as much air as possible and losing as little as possible to anything not cooling the engine or oil. The air escapes up past the cylinders into the plenum and out through an eductor exit. In the real world it is hard to pressurize anything, so I'm simply closing off big air exits other than through the baffles. Regarding exhaust, the pipes are kept fairly short, and 4 pipes exit below the prop hub. I'm going to use a 8" prop extension. That should give enough dissipation area for the exhaust not to focus heat on a given strip of the prop.

In summary,
- boat tail cowling
- plenum atop cylinders
- cylinder cooling air exits through eductors either side of the prop hub
- exhaust exits through eductor underneath prop hub

___________________________________________________


And here are the half-done ideas on making cowling and cooling easy to live with.
• The top cowling “plate” should come off with nothing attached to the engine, so I get to see and service most things from the top. Dipstick hole in the top plate.
• The bottom cowling piece includes a ground adjustable lip to the NACA air intake, done via a metal lip on some screws. That lip includes the “fences” that keep NACA air from spilling out the corners.
• Exhaust pipes are pulled in close to and under the propeller – probably without a 4-into-1, just 4 pipes pointing aft. The exhaust exit is framed with an aluminum plate / mask, so airflow can leak around the pipes, but less than otherwise.
• Baffling will be high temp composite - I've ordered a half gallon of "Max HT", being the PCI product for high temps. It looks difficult to work with - at room temp's it can get to gel stage, then put it out in the sun for a "glass" stage (where it can shatter), then cook at 200 for 5 hours to get a hard cure that keeps strength up over 350 degrees.
• A hole in the forward baffle gives a dedicated air exit to the oil cooler, which is mounted on the spar with a 2” or 3” SCAT hose into the baffle.
• Cabin heat will be the EAA standard aluminum sleeve around a door spring around the exhaust pipe, with a bilge blower drawing air forward.
• With Dual EFI I won't need to blow air on a mag, so I don't need a cool air pipe.
• I have a prop-facing induction. I'd like air intake to come with the blessing of intake pressure, but the replacement systems all focus on making a vertical intake into a prop-facing intake (for instance, https://www.aircraftspruce.com/catalog/eppages/superior_sv78950.php). I'll keep looking for something that might work. Right now I expect to end up at an elliptical lip air filter box feeding a fiberglass connection to intake. I can use the Freeman modeling wax sheets to make the complex shape locking onto the induction box.

  Feb 26, 2023     Cowl #4 - (48 hours)       Category: C23 Eng+Cowl

I shaped one half of the cowl mold, then cut profiles (see below) and used them to shape the other half. There were multiple checks against the profile taken from the airplane (pink foam cutout shown below). After plaster and final shaping, it was two coats of epoxy to seal the plaster dust, then 3 coats of mold release wax and two coats of Part All spray mold release (polyvinyl alcohol). Note - you get better results with sealed plaster + packing tape.

I then layered the form with peel ply. I'm making cowls with layers of 6k h5 twill carbon fiber, Coremat 2mm XM and a second layer of carbon fiber. The headache with cutting CF is that it slips like crazy on the peel ply - a shot of spary adhesive fixes that. The headache with the Coremat is that until wetted it is much less flexible than the CF. I ended up with 3 pieces in each CF layer and 8 pieces in the Coremat layer. It all gets topped with more peel ply. Attached is a picture showing the inner CF layer sitting on the inner peel ply, with the next 3 layers on the strake in the background.

The layup was West Systems slow for the inner layer, then Raka + Non-blushing hardener for the Coremat and outer layer. I continue to be wildly impressed with the Raka epoxy - so long as I use their high end hardener I get a superb layup. The tradeoff is a short pot life... The layup day was 6.5 hours of scramble, but I have well wetted layers and it all tied together. The Coremat got a lot more flexible when wetted, so perhaps I could have cut corners in fitting that layer. After a 18 hour cure the mold+cowl went into the Florida sun. 4 hours at 120-140 surface temp finishes the 2nd cure.

Below are pictures of the piece before and after removing from the mold and cleanup. The inside was coated with fuel proof epoxy, and the photo shows the areas I again masked with peel ply. (Note for others - I mixed 300 cc's of epoxy and could have done the job with 210. I hate wasting the good stuff.) Once I had a clean lower cowl I started iterating test fit + trim. The lower cowl was on and off ~10 times, aligning it front and back, marking the edge, back to the table to trim off a bit, then repeat. Picture 10 is a milestone where the front and rear edges work and I still need to trim both top & bottom for good matching surfaces.

  Dec 12, 2022     Cowl #3 - (60 hours)       Category: C23 Eng+Cowl

I can accept the quality of the top cowl (CF + Coremat + CF) but I was sure the commercial ones must be better. In KY, Greg Cross had one he ended up not using, whcih he said was for a wide deck 360. It was a 3-day 2000 mile round trip but I went up, socialized and brought it home. Then I found it is actually a poor fit - !@#$!. The commercial cowl has the oil door too far forward and the exhaust openings are too tight and incorrectly oriented. The upside is that my Coremat cowl is about 60% of the weight of the commercial product and a lot stiffer. Time to double down on making the lower cowl....

This has proven to be a much bigger task than expected, and perhaps it would have been more direct to flip the plane again. But I persevere. Here's what has proven successful.

I need a tight fit at the firewall/wing root, and then at the trailing edge, particularly around my exhaust eductors.
- Lower profile. I used square cross section aluminum to bridge the NACA duct, providing a shelf/lip for the forward edge pf the lower cowl. The lower cowl needs to wrap around the air intake, alternator and starter, reaching up in front of the ring gear to fit behind the (extended) prop hub. I like having at least a 1.5" gap for the lower engine baffle. I used building sheet foam (purple!) to profile the lower cowl - it ended up looking like a hockey stick. I copied that foam shape onto luan plywood and fixed it in place using 5 minute epoxy. The plywood ends up as a keel for my boat tail lower cowling.
- Wing root match. Using the same wing root profile as used for the upper cowling, I built on the plane a foam block lower cowling. The wing root blocks were held in place with bamboo skewers passing thorugh the wood profiles into the foam . At this point I have gone through a _lot_ of foam insulation sheets, and the cost approaches buying something from Velocity. Moving on...
- I did a rough carving of the foam while in place on the plane. I then (carefully!) supported the cowl while I sliced the foam away form the wing root profiles and firewall. I removed the cowl and reinforced the inside with pour foam and building foam. At the sides I added an additional 3/4" wing profile, copying the wooden ones. The block was then moved to the bench, ending the messiest part of the upside down work. Note that with the interior of the foam core now reinforced with pur foam, this core can no longer fit on the plane.
- Firewall profile. I trimmed a piece of foam board to match the exisiting firewall lip. Fitting it to the foam core I saw that it was not a great fit to the new foam template. !@#$!. After 3x checking the template I used both foam and plaster to get a darn good match. I then re-shaped the cowl to smoothly flow from the correct firewall profile.

[Sidebar: Spray foam was an uncontrolled horror, but it took only 2-3 hours to clean that up. Pour foam is much easier to sand, but expensive and it's hard to predict how it flows before it swells, gells and hardens. I was happy to find a pour foam source even less expensive than my bare bones boat supply distributer - Amazon has some, and they sent me 1/2 g containers of A&B when I had paid for only the quarts. Doing it again I would build on the airplane with blocks of foam, remove and add a thick pour foam layer, shape with rasps, then use plaster to get a smooth surface. Model one side, then cut templates and shape the other.]

  Dec 12, 2022     Throttle Cable 2 - (12 hours)       Category: C23 Eng+Cowl

It took a darn long time to 4x measure the correct length and winnow vendors. See attached. The cable is treaded from rear of the instrument panel, a U-turn to run alongside the pilot side of the aircraft, and throught the firewall just left of the fuel. 3 new clickbond studs were added to further secure the shielded, armored throttle cable.

At the engine end the armored throttle cable goes into a steel tube with external threads, then a small dimeter smooth steel tube, from which emerges the moving 10-32 threaded solid rod. To the end of that threaded rod is an engine-rated clevis with a 3/16ths pin, matching the hole in the throttle arm. I modified of piece of aluminum angle so that it bolts (two 1/4-20 by 3/4" high strength bolts, with split washers) to the existing tapped holes in the oil sump. That bracket holds two engine compartment - rated Adel clamps. The forward one is secured around the steel tube with external threads. The one further rear is secured around the small dimeter smooth steel tube.

There was a fair amount of measuring and re-measuring, which meant 1st try success in cutting and drilling the bracket. There is full throw of the throttle are with no conflict with any structure throughout the travel.

  Oct 31, 2020     15. wiring EI - (1 hour)       Category: C22 Electrical

Backtesting the switch layouts.... The instrument panel cutouts plan on using the "Design1" programmer. This is a newish model, and so far SDS has posted the switching and fuses to-from the older style panel-mounted controller head. This looks consistent with what I understood when I did the switch layout. Looks like two electrical cables run from the control head back to the firewall one for the coil packs and one for the main harness. I'm guessing the panel switches are wired into those harnesses, and the harness will be custom built to fit my project. For purposes of the firewall arrangment, the single ECU dimensions are 3.875 x 8.5 x 1 inches, gold anodized, 6061T6, CNC'd billet. I'm doing double, so figure it at 2" tall.

More electrical redundancy - this is a dual alternator system with a single battery. The Nuckolls design standard calls for sufficient redundancy to complete the planned flight. The SDS EFI system draws 1.5 amps and the fuel pumps draw 4.5 amps in normal ops (14 gph), 5.5 at the max expected draw (20 gph) = 5.5-6.5 a. Add 4a for the avionics = a total draw of 9.5 - 10.5 a. The PC680 battery has 16 amp hours, so I get 1.5 hours in which to notice the flashing lights and change to the back up alternator, cut high drain items (not many of those) and plan for a landing. This is an all electric engine in an airplane that "glides" at 150 mph, so the circle in which to find a safe landing is ~200 miles radius. I'm certainly safe, but the next step would be a move to a higher capacity b/u alternator - the Monkworkz ML-30z 30a permanent magnet, also using the vacuum pad. That will shift duration back to fuel remaining.

  Oct 04, 2020     Cabin Lights - (40 hours)       Category: C22 Electrical

For cabin lights I went through 3 iterations before going final with aluminum panels of LED strips fed from a single pulse width modulating dimmer and a selector switch (red-off-white) mounted on the panel. There are red panels in the strakes + instrument panel lip, and white panels in the strakes and overhead rear. Each panel is 0.5 meters long, and 10 strips of 36 LED's, for a panel of 360 LED's. The max draw is 5.3a for the 3 white panels at full output. Below I show both the LED strips and the amperage test set up. The wire runs are short, and I used 20 awg from main bus (15a) to dimmer to red-white switch, and then switch to the point where the current is split right-left-rear. From that point I used 22 awg.

  Oct 26, 2022     Cowl #2 - (100 hours)       Category: C23 Eng+Cowl

My first pass at an upper cowl weighted in at 7 lbs, right on target. But I can see the errors. I had rough edges where CF strips bridged across my 3-part layup. I had not lowered the nose and layers slipped to the rear and down the sidewall, leaving me with edges without the Coremate center. My edge was not smooth where I met the firewall. Clearance of the starter ring gear was only 3/8". The right and left looked symmetrical, but measurement says they are not.
Smoothing the surface and prepping for paint added too much weight. Crap... I decided to do better with the next effort.

I again started with the top cowl. Top cowls are straightforward - start at the firewall and go smoothly to the prop, while staying clear of the cylinders and starter ring. I modified the mold to get a better outcome.

Building a core for the cowl:
- I changed the profile at the firewall to get a smooth line around and across the (wide deck) cylinders
- over the ring gear and rocker covers, add a clearance spacer
- starting from the wing profile and firewall, make a foam core/mold that fits around the engine installation,
- define the exit for cooling air. In my case, a box with the lower horizontal tangential at the lower edge of the prop extension, the rest defined by the shape exiting the wrap around the cylinders.
- rough shaping, then fine shaping, I used foam blocks, foam strips, plaster and micro. For plaster, use the quick set that comes in the 18 lb bag. A 5 lb bucket of plaster will dry too slowly for any practical use.
- fill-sand-fill-sand the mold, then epoxy wipe. Now that one side is perfect, copy that profile to the other side.
- box tape on the mold (you need the epoxy wipe before tape will stick to the plaster)

Layup schedule, top cowl
- Prepare airplane – lower nose so cowl is near horizontal
- Prepare mold: 3 layers wax, buffing between each + 2 layers PVA, using spray bottle
- Using Raka slow cure non-blushing epoxy, paint mold and affix peel ply
- Using oil & gas resistant (GRE) epoxy, carbon fiber tape on edges that will hold Dzuz / Camlock fasteners
- Using GRE epoxy, place and wet out the inner layer of CF
- Using AeroPoxy, place and wet out the 3 pieces of Coremat 2mm XM
- Using AeroPoxy, place and wet out the outer layer of CF
- Afix peel ply and roller, hoping to draw out excess epoxy
- Plastic sheeting for spreader work – draw out excess epoxy and aim for smooth finish

[This task will use a lot of scrap foam, Gorilla Glue, and cheap boat epoxy. On pricing, note that the best epoxy made by Raka (900 resin + 350 hardener) was 75% the cost of West Systems and 50% the cost of Aeropoxy.]

The bottom cowl is a lot harder than the top. I am wildly disinterested in flipping the plane, so I explored making a removable mold. I began with a bridge across the NACA duct, then a "keel" piece to form the center profile (first in foam, then in luan ply). The keel also facilitates a boat tail cowling, which race winners say is more efficient. THe forward edge is glued to the aluminum bridge, the trailing edge glued to a wood disk between prop flange and prop extension. I then filled in with styrofoam 2" board insulation.

  Jul 01, 2020     repairs - (4 hours)       Category: C18 Canopy

The pilot side gas strut failed at the upper bracket connection. This is a threaded connect and prior builder Chad had a mis-matched thread between the strut end and the fitting connecting to the lift arm. He used an epoxy patch amd wishful thinking. Hmmm. I disassembled, cleaned out the old epoxy, used tap and die to clean the threads, cleaned with acetone and patched with JB Weld. Below are pictures showing the strut at inspection, the break, and reinstalled repaired part.

In order to repair the gas strut fitting I had removed the canopy. It's always time to clean, and this time I saw a hairline crack in the new paint job. The frame was made 17 years ago and has been on and off the place dozens of times - I don't know when this happened. I sanded off the finish, patched with 2 layers of BID and put peel ply on it. The epoxy was "G-Flex", a toughened epoxy that I like for repair work. I'm not sure I'll repaint - this is a reminder to stay vigilant.

The rear canopy is hinged at the top of the fuselage with a gas strut running from the fixed portion to a ball connect on the left side. The original gas strut failed (no remaining pressure), so I replaced with the original 10 lb, then upsized to a 20, then 30 lb strut. Installed it correctly (cylinder up and rod down, keeping that oil drop on top of the seal) I think it will last. This strut is 20" long.

  Apr 20, 2007     Piping the Pitot and Static System - (14 hours)       Category: C13 Nose+Gea

For the pitot, I found a small air compressor blower nozzle on McMaster-Carr that had a threaded fitting on one end. A threaded bushing in the nose provides the connection, giving me a removable pitot. An elbow joint completes the picture by connecting to the aluminum tubing that runs up to the instrument panel.

For the static port I found some small machined aluminum static ports that seem like they'll do the job. They (one per side) get installed in recessed holes just under the outside skin and plumbed the same way.

PK note, November 2021.
1) The removable pitot is an M10 thread, with a barbed front. 6mm Barb x M10-1 Male; Total Length: 1.22"; Hex Width: 7/16ths. I have a pitot length of 7/8th inch, with the lip barely forward of the landing light lens. In flight testing I'll determine any pitot-static error.
2) As described elsewhere I have pitot heat from a thermal tape wrapped around the aluminum tube. The temperature test went well. I still need to test for a self-draining ability in the pitot system.
2) Pitot static plumbing went missing from the project as delivered, and I assume they were retained by Craig W. I bought a full pitot-static kit from ACS ("Avery EFIS System Plumbing Kit" at $132) with lots of hose in 3 colors and way too many fittings.

  May 22, 2022     cowl       Category: C23 Eng+Cowl

The engine is a wide deck angle valve. The plans call for buying the cowl - with no cowl available for my engine scratch build is a reasonable deviation from plans. Below I show that this is _not_ efficient or cost effective; if you can use the Aero Composites / Featherlite cowl, by all means do so.

I had to start with the requirements. The cowl has to
- provide a streamlined shape to the fuselage,
- have a stiff and resilient surface durable enough to handle the proximity to prop wash and engine vibration,
- be secured with reliable and easy-to-remove fasteners and
- facilitate engine cooling, with efficient and properly sized intake and exit areas.

Shape and Fasteners. Cowls should have minimal drag and ideally complete an airfoil shape over the fuselage. I've chosen to have a) an upper - lower join that matches the wing trailing edge, b) a boat tail on the lower surface, and c) exhaust exiting through an "eductor", an effort to have exhaust gases pull air through the cowl.

I secure the cowl to the attachment flanges built into the wings and fuselage. I will use Dzuz fasteners every 4 inches on the top and bottom arch, and piano hinges at the wing-cowl mating surface. This is contrary to the builder group discussions, which prefer Camlocks. I have 85 year old Dzuz still working well in a PA12, and I'm comfortable with the simpler install - I'll go with Dzuz and use more of them. For the long straight joins I'll use my RV experience and go with piano hinges. I'll pay up for the extruded hinges, just in case it's a strength issue.

Construction method. The shockwave of prop passage will hit the cowl surface at 2700 rpm x 2 blades = 5400 times a minute, 90 times per second. The closer to the prop the harder the hit. To handle that force I decided to make this cowling from carbon fiber cloth with a Coremat center. This is consistent with a presentation for the Cozy builders group. The presentation is at https://docs.google.com/presentation/d/1vTyFMM8RQ9PijpZ07AT6vDHU164KCHXY/edit?usp=sharing&ouid=111349802257488370268&rtpof=true&sd=true

Intake area. Per _Tony Bingelis on Engines_, the rule of thumb for inlet areas is .35 x engine hp (p.87). At 200 hp, my IO-360 should have an inlet area of 70 in2. The cross section at the end of the ramp of the NACA inlet is 15 x 2.75 = 41.25 inches. Unless the NACA in nutty efficient and has an effective area 1.7x the cross section, I need another 30 in2. My oil cooler has a face of 30.8 in2 - if I feed the cooler with an armpit scoop sized to match the inlet area is 72 in2 and on target with the Bingelis guidance. But most Cozy fly without the armpit scoop feeding the oil cooler. I will build and make initial flights with no armpit scoop, knowing I can add one. I will focus on having an efficient NACA intake, using the corner tabs that keep air from tumbling out of the intake.

Exit area. I have not found a standard for outlet area. Per Bingelis the outlet is at least equal to the inlet. Per the Vans forums it should be around 1.6x the inlet area. Above I plan for 72 in2 of intake, so I target an outlet of 115 in2. The area around the prop flange is a diamond of total area = 119, less the area of the extension prop flange for a net 80. My exhaust pipes exit through the lower cowl; I am shaping my exhaust pipe exits as eductors, hoping they will pull air out of the cowl. The area of the eductor is 23.8 in2, less the area of the exhaust pipes is a net of 17 per side = 34 total. Total exit area is 114 in2. On target.

Construction steps.

1) Install the (purchased) exhaust on the dummy engine, and any eductors. If not done already, install the cowl mounting flanges, keyed to the expected thickness of your cowl. Wrap the engine.

(Eductors are a lower cowl feature, and need to be fitted and secured before the lower cowl layup. I determined the target radius and pop riveted stainless steel shim stock. I then wrapped this in a sandwich of fiberfrax, CF+ Coremat + CF. I let the epoxy set in a frame that gave me a targeted oval shape, matching the exhaust pipes.)

2) Mount the prop extension - the cowls will need to converge and terminate 1-3 inches forward of the mounting flange. The front face of the extension is the back face of the prop.

3) Mount the wings and copy the edge profile required at the cowl-wing join. Securely attach the profile to the spar, but knowing it will be removed after final cowl casting. Remove and store the wings.

4) Surround with foam STRIPS. The cleaner the curves of the foam strips the less time and money you will spend on filler. [A better approach is to build a cardboard shape, then fill around that with the foam.] The cowl is laid up on a casting surface, and the better the surface the better the cowl. Plaster and sanding, epoxy squeegee on the plaster, then packing tape. After all this work I finally have a good casting surface.

5) An armpit scoop was laid on a core carved from a block of pour foam. CF + Coremat + CF for the scoop. This will be held in reserve for use if required.



Materials:
4 sheets of foam sheathing @ $28 each = $112
6 cans gap fill = $24
2 gallon kit of pour foam x 2 = $125 x 2 = $250
1 quart PVA = $26
1 tin miracle glaze mold release wax = $22
10 yards 3k 4hs carbon fiber (needed 7, bought extra) = $182 (Soller Composites)
5 yards CoreMat XM = $0 (the manufacturer sent a free test roll)
1 gallon West resin + extra slow hardener = $120 + $60 = $180. I buy this as I run out, and it was time to get more.
1 gallon acetone to clean arm hairs of gap filling spray foam = $25. As with the epoxy, buy it when in runs out.

Total ~ $825. Chris at Aerocomposites offers a full set for $650 - if you can use a purchased cowl, it's the smart way to go! https://aerocomposites.aero/

  May 13, 2022     stations for CG envelope and W&B - (4 hours)       Category: C99 to sort

An aircraft CG range is the designer's planned relation of mass to lift. The aircraft designer sets a datum plane typically forward of the nose of the spinner - that makes for convenient positive number arithmetic. He then measures the distance aft of the datum of the empty CG of the prototype, setting safe foreward and rearward CG limits by testing that prototype. No matter how closely the plans are followed, EAB airplanes are going to vary from the prototype. If the build varies in the location of lifting surfaces it has to affect CG range. That's why it makes sense to double check measurements.

Builders agree that for the Rutan-derived canards the important reference plane is the wing root leading edge ("WRLE"), since it is a fixed point as regards the wings and thus the center of lift. The plans assign that point FS 113.9 - conveniently it falls at the outmost corner of the strakes and can be measured wings off. Given the WRLE, what are the stations for the leading edge of the canard, our other lifting surface? And what are the stations for the main and nose gear (reference points of the scales), and the several stations at which weight can be added or removed (ballast, pilot & copilot, fuel and rear passengers)? Time to do some measuring.

First step, find someone smarter - in this case Marc Zeitlin and Joel Ventura. I distilled their work into a 3-step process. First, set the WRLE and define the location of the landing gear relative to it. Second, set the aircraft on scales and set the empty CG. Third, measure the actual arms for the stations of interest. Here's the step by step.

1. On a clean floor, level the plane. The first several steps involve measuring distances and can be done wings off.

2. Drop a plumb line from WRLE to the floor and mark the spot (right & left). The plans define this as 113.9 inches, and I used it as the reference point for all other stations.

3. Drop a plumb line from the center of the main gear axle, mark the spot and measure the distance forward of the WRLE (for me, 4.1 inches for both left and right axles, for a main gear fs of 109.8). The plans call for 110, and the template draft POH calls for 109.5 +/- .5. The result is between plans and the POH and within tolerance. The 0.2" bias will very slightly increase the moment arm of the engine relative to the gear, so I might rotate 1-2 KIAS faster than otherwise. The discrepancy has no effect in the air. The landing attitude might be a degree or two nose higher and 0-1 knot slower. I might need to be just a little bit more careful about tip back of the aircraft while moving around the hangar.

4. Drop a plumb line from the center of the axle of the fully extended nose wheel, mark the spot and measure the distance forward of the main gear on the center line. Plans say FS16, which is 97.9 forward of the WRLE reference line. Mine measured at 96.14 - I will fully extend the gear and remeasure.

5. Drop a plumb line from the leading edge of the canard at the side of the fuselage, mark the spot (left and right). The plans say 18.7 but the manufacturer provided draft POH calls for this to be at 18.6, +/- .54. This is 95.3 forward of the WRLE reference. Mine measured at 81.4 - I absolutely need to remeasure!

6. By now there is a lot of tape on the floor! Snap chalk lines and make the measurements. Measure twice. At the end is determining if the build is in conformity with plans, and if not, that the variance is known and not a flight hazard.

7. Now it is time for the scales. My EAA chapter has a set of vehicle scales that are not certified but accurate within a half pound. The airplane should be ready to fly - all standard equipment and normal oil. I hold that you need to have installed all standard equipment - the ELT, POH, et al. Add 5 gallons of fuel and drain all usable. (With aft fuel sumps, the higher risk is level flight - drain with the aircraft level.)

8. The next step sets the arm of the various stations where weight will come and go. The following sequence is keyed to standard definitions of empty weight and to keeping the plane safely on the scales. For each step, add a _known_ weight at the given station. Write it all down - each measurement includes each of the 3 scales + the weight added.
- Empty weight and CG. This may require 50 lbs on the nose wheel to avoid a tip back.
- Add 100+ lbs at the pilot station and remeasure.
- Move that 100+ lbs to the copilot station and remeasure.
- Move that 100+ lbs to the right rear station and remeasure.
- Move that 100+ lbs to the left rear station and remeasure.
- Add max weight (24 lbs) at the most forward ballast station and remeasure.
- Add 50 lbs at the standard ballast station (for me, forward of the nose wheel housing) and remeasure.
- Move that 100+ lbs to the right strake, centered on the fuel cap, and remeasure.
- Move that 100+ lbs to the left strake, centered on the fuel cap, and remeasure.

9. Do some high school arithmetic and determine the station arm for each of the points.

10. Develop a cg worksheet or modify the Zeitlin template. ***Test the worksheet*** before leaving the scales - does the calculated CG match the actual? If so, load the same into Foreflight.

Going forward, look at the placard above the start button. T/O = WT + CG + DA.

  Oct 16, 2020     05 panel - mounting trays and supports - (32 hours)       Category: C22 Electrical

I ended up spending a week fussing with the supports and stiffeners around the panel. It is shaping up as a stiff, strong box with very little flex, while also being fully removable.
- The instrument panel is .063 aluminum bent into an "L", and right, left and bottom center of the panel are tied to the old plans panel with machine screws into my home made G10 nut plates, floxed.
- I have a 1/2" square aluminum tube running across the remnants of the plans i-panel, screwed in on the right, left, and center support. The bottom edge of the panel attaches to the lower edge of the square aluminum tube, using nut plates attached to the top of the tube and machine screws coming up through the tube.
- The top of the panel screws into another piece of square tube, increasing stiffness at the top and across the center.
- Side panels of the radio stack attach to the IP and the cross member used for lift pins into the canard. It's as strong as anything not called a spar. There was some experimentation and dead ends in the shape of the supports, but I finally got where I need to be.
- The audio panel and the 2" GPS-Comm are held in manufacturer-supplied mounting trays, them selves screwed into rivet-nuts on the side panels. The mounting trays add rigidity.

Theoretically I can disassemble and pull out the panel in one piece - I have to take out 2 machine screws from the top edge, three across the face from the bottom and 4 from the leg openings. That is not super easy, and maybe I overbuilt this.

  Mar 05, 2022     air filter - (2 hours)       Category: C23 Eng+Cowl

The A sump locates the throttle body between the sump and prop. It was convenient to directly connect the air filer, putting it at the aft end of the cowl. Per other builders, air "piles up" at this location for a relatively higher intake air pressure. The plans say the air filter is to be "Fram 3647 or equivalent". That filter has a surface area of 66.9 in2 (2.5" height & 8.52" diameter) before counting the pleats. I went with a K&N filter connected to the SDS throttle body with a section of 3" reinforced hose. (I hear rotten lab tests for the K&N air filters, but I think it is the most common make used in amateur built aircraft.) This filter has a pleated surface with an average diameter of 5.44" and height of 5.0". C= pi x d = 17. A = C x H = 85 in2. This is before counting the inner cone, estimated at 40 in2 for total area ~2x the plan filter. The additional area will offset the difficulty of breathing through an oiled filter.

  Nov 06, 2021     Fuel lines & flaring tools - (12 hours)       Category: C21 Strakes

It's been a long time since I last made fuel lines, and those were either Aeroquip 471 or automotive hoses. This time I'm using -8 and -6 versatube (3003 soft aluminum) with AN fittings and the 37 degree single flare. It's a fuel injected system, unpressurized from tank through selector valve to pump, then pressurized at about 30 psi for a short run from pump to injectors. The return lines are after a pressure regulator, and are 6-8 psi. The -8 (1/2 inch OD) Versatube is rated at 380 psi, and the -6 (3/8 inch OD) to 520 psi. The tube isn't the issue, it's the quality of the flare and the correct torque on the fittings.

Flaring tools come in a _wide_ range of styles and price points. After two of the lowest grade I moved up one level, and I'm loving it. Below is a photo showing the equipment. After practice, my process is
- cut with a standard cutting tool
- remove burrs with a step bit
- put B nut and sleeve on the tube
- place tube in the correct sized hole of the multi-hole bar, with the end of the tube proud of the surface
- use a course file on the cut end, then check for a clean and perpendicular cut
- level the end of the tube with the top of the bar
- bar goes into the flaring anvil, centering the cone on the tube
- tighten the anvil on the bar; the anvil anvil is not allowed to twist
- add a drop of oil to the cone and spin it until it self-releases
- disassemble and inspect.

The height and inner surface of the flare has to be perfect, with no burrs or scratches. Tool marks on the outer side wall are acceptable. After practice I like the results.

  May 13, 2022     shuffling everything - (20 hours) Category: C02 Workshop

I want to get the wings on so I can work on the cowling and controls. Those wings need a 31" free span, and my garage can just barely do it - if I move everything. I have pretty much everything in the shop on wheels, so it required a lot of putting away, a new storage bay on the ceiling, disassembling a favorite work table.... it's a project that took 2 weeks of nibbling away. The spend was for additional lumber, drywall mud for the cowling work, et al.

  Aug 22, 2018     more work shop prep post project delivery - (29.5 hours)       Category: C02 Workshop

I kicked off the project by fitting it into the work space. I had to turn a woodworking area into a new space for a family car, install a garage door, build shelves, re-arrange a heck of a lot of accumulated possessions, et c. Everyone reading this will understand the shorthand for weeks and weeks of shop set up time. When the project arrived, I was able to move the wings to the ceiling, canard on a side shelf, miscellaneous in the attic, and fuse front and center.

Friday 8/10/18 Receive Project!
Saturday 8/11/18 8.0 unload, unpack, stow
Sunday 8/12/18 8.0 inventory (missing key, missing bolts
Wednesday 8/1/18 2.0 350.00 lay in basic supplies - laminating & sanding epoxy, BID, peel ply, 4 mil plastic, squeegees, dry wall trowels, sticks, containers, scale, et al
Monday 8/20/18 1.0 65.50 order replacement key
Tuesday 8/21/18 8.0 9.00 design and build device to invert plane (buy bolts), invert
Wednesday 8/22/18 2.5 8.00 build sanding boards
_________________________________

The last photo shows painted wings and fuse in the workshop. The crowding of benches is the result of multiple open tasks, the result of shipping delays, backing up to redo a prior task, and the urge to get something tangible done in some limited amount of time.

  Dec 31, 2025     Project Selection - (40 hours)       Category: C01 Introduction

This entry describes events of Q2 and Q3 2018. I want it as the display entry for the builder's log, so I date it 2025.

I am the 3rd owner of Cozy project #1147, having acquired a "mostly" complete hull, wings and canard. This log is organized by chapters of the 2-volume set of plans, and includes entries by both prior builders and me.

Prior Builder. Chad Robinson is a software designer and developer with an unstoppable personality and a gift for precision. As his first build he picked the Cozy Mk IV, a plans built airplane that will take 2000-4500 hours to complete. He launched this project in September 2003, working out of a shop around New London, CT. At the same time that I was active in the Hartford EAA and networked with the DAR's and homebuilders throughout the state. Even so, I did not know Chad - he was working down on the shore and seemed largely outside the EAA networks.

Eventually a big family and a growing business cut into his build time and Chad sold the project to Craig Westwood, a first time builder in Athens, GA. At the time of the sale this project was inspected by Marc Zeitlin, a highly regarded aeronautical engineer who specializes in canards and does 10+ canard condition inspections each year. He gave #1147 a clean bill of health at the purchase by Westwood. Craig Westwood bought the project, spent some time sanding and thinking about components, then decided he wanted a plane in which he could learn to fly. He bought a flying Cozy and sold this project (removing some high value parts he plans to use). I was referred to this project by Marc Zeitlin. Since then I have been pleased by the work of Chad Robinson and wished that Craig had better documented his holding period actions.

Craig and I hammered on a price for a while, and I probably paid too much, but better to pay a mediocre price for a great project than a great price for a mediocre project. Then we had to find the right shipper, and that occasioned a renegotiation. The airplane was delivered via a specialty trailer, with a hydraulic system for raising and rotating the aircraft. Neat!
_____________________________
Sidebar on Resale Values
High quality homebuilts are the great bargain in aviation - a pristine 20-year-old home built might sell for the cost of the used engine. Unfinished, engine-less projects go for less than the cost of materials - yep, all that labor adds zero value. I've been a happy team player working on airplanes where all or most of the starting project came free to my EAA chapter. We worked as a team, got the airplane flying, and sold it on to a new owner. My favorite project story is a happy little acrobatic biplane donated to the chapter by a widow. We got it back in the air for less than $10,000 and sold it for our cost plus a young eagles donation. The plane is still zipping around and everyone is happy to see it getting loving care and attention.
____________________________

  Dec 08, 2020     98. Discarded - copper tubing connections - (6 hours)       Category: C22 Electrical

I wanted short distances for all the main power connections, and certainly achieved it. On the firewall I have a vertical stack of the battery, my "fat wire point", the battery contactor and the battery bus. The total distance is 5 inches, and the longest segment is 2 inches. It is clumsy to have distances that short connected with cable and lugs.

Consistent with advice in AeroConnection, I am connecting these parts with solid copper, formed from sections of copper tubing. The hydraulic crimper has multiple dies. Reversing a pair to get a flat surface, the 3/8 inch pipe is crushed to a flat of 14 mm x 1.2 mm = 16.8 square mm cross section. This is the same cross section as #5 AWG. which has a resistance of .00103/meter. 5 inches is .127 of a meter, so the length has resistance of .000156. Using an online calculator, at 14 volts it can carry better than 3000 amps while losing less than 1 volt. Yeah, I think it will work for my system.

A headache in using solid copper is the alignment of the attach point - each of the 3 connections has ends perpendicular to each other. Cut it, crimp it, test fit, modify, test fit... there is no time savings here. Once fitted, cover with shrink tubing for some insulation - getting zapped is a PITA. A better approach might be to make the copper bar and put twists in it in order to get the alignment - I'll try that if I have any connection problems.

  Mar 06, 2022     Throttle cable - (1 hour)       Category: C23 Eng+Cowl

Before buying a cable I made sure of the length at full throw. That required knowing the throw. I measured the distance from arm to a known point (here, the throttle body base) at full open and full closed. The distance is 2.75 inches.

The throttle body mounts on studs inserted into the sump, forming a square 2.5" on a side, The throttle body was installed in the temp engine using the o-ring and a film of lube.

  Jan 26, 2021     04 Band Saw Fence - (1.5 hours)       Category: Z-Tools+Tips

I was getting ready to make covers and needed 1/4 inch foam. Rather than break into a big sheet I decided to cut down some left over blocks of "the blue stuff". From past experience I knew I could get close with a hand saw and sand flat from that. But could I do better? I know that serious woodworkers use a band saw fence for slicing veneers from larger blocks, a process called "resawing". I found that a resaw fence runs more than $100. But taking a look at the scrap pile I see
- a Harbor Freight give away, the 14" magnetic bar.
- a piece of cap rail from a deck fence (when and where did I get this?)

The HFT mag bar is a steel U-channel filled with a series of rectangular magnets. I uncovered them, pried them out, and did a test fit in the groove cut in the hand rail fence. Fits! I drilled some shallow holes in the wood of the groove, filled with flox, and pressed in the magnets. Remove the overflow, and let the flox set, making sure of a flat surface at a right angle to the side of the cap lumber. Once it set I had a fence that holds tight to the surface of my band saw, while being fully adjustable.

I set the fence on the left side of the blade and measured the distance from the slot machined in the table. Once square, I trimmed a block to the max height of the band saw, then went to to town on the block. The results are shown in the picture below. This should be more than enough 1/4 foam for the covers, and I have a technique I can use for the cowling.

  Feb 20, 2022     Baffles - composite w high temp epoxy       Category: C23 Eng+Cowl

Talking to experts built my confidence - I'm going to try for composite baffles, done with high temp epoxy. Klaus uses an aerospace product I can't access, so when pressed for a recommendation he offered:
- make sure it's a laminating epoxy and test the sample with a scratch test. There will be a transition temperature where it goes from a hard ceramic-type scratch result to something like a rubber sheet. That transition is the measured temperature for this use.
- The outer edge of a cylinder fin will be a lot cooler than 400 degrees - closer to 200.
- An epoxy with a two-stage curing cycle has a chance of meeting the standard. He likes MGS.

I have West, Aeropoxy, Max HTE (high temp), Max GRE (fuel resisistant) and a number of general purpose epoxies. Only the Max HTE comes anwhere close to the target temperatures, so I'll make a test blank with it, then stress test a toaster oven!

Here's the write up on Max HTE: MAX HTE A/B is a two-part epoxy based system especially formulated to provide structural strength at temperatures of up to 200⁰C or 390⁰F under pressure or load. It is formulated as an impregnating resin for fiberglass, Carbon Fiber, Kevlar, Spectra Fiber and other specialty materials that require heat performance. MAX HTE A/B can also be utilized as an adhesive, encapsulant or potting compound, tooling resin for high temperature applications. MAX HTE exhibits exceptional toughness and other mechanical properties such as adhesion, compressive strength and flexural modulus at high temperature exposure. MAX HTE also provides excellent chemical and water resistance. Fillers and colorants can be added to modify it for a specific use. MAX HTE provides a long pot life at room temperature suitable vacuum assisted resin transfer molding or VARTM, fiber pultrusion, pre-impregnation process with a relatively short heat cure time. It is specially designed to withstand continuous high temperature service, high impact resistance and exposure to harsh conditions. MAX HTE A/B is 100 % solids, low odor and toxicity and does not contain Ozone Depleting Chemicals (ODC).

  Sep 12, 2020     Interior paint - (12 hours)       Category: C25 Paint

This work was done in July, right after finishing the exterior.... I have a fair amount of posting to do.

Looking at the possible mess of a bad zolatone paint, I went with the easily controlled rattle cans. Rustoleum makes a flecked paint in several shades - I went with "Pebble". It seems to be plastic flakes mixed with a high power solvent and a propellant. It sprays rough, then settles into a smooth paint with the look of a high end spatter treatment. Topped with a matte finish clear, it appears to be easily maintained, and certainly repairable if and as required. Note the miserable coverage of the flecked paint - 10 cans were required for the interior, and I've kept 2 cans in reserve. The top coat required only 2 cans. [followup: it does not wear well. I painted before running power and fuel lines, stages where one steps in and out of the back at least 100 times. Lots of bare spots. I'll clean, spray and again top coat with the clear enamel.]

I made the mistake of painting the cabin side of the firewall - I subsequently sanded the surface in order to get a good attachment for mounting electrical components. Lesson - don't paint where you want to glue things!

  Jan 25, 2022     Fuel lines - (100 hours)       Category: C21 Strakes

Fuel tank selector, fuel filters & pumps, fuel lines

First is choosing fuel injection and electronic ignition, going with the SDS EM-5 described elsewhere. That choice meant a switchable feed from each tank to the filters and pumps and engine, then a return line flowing back to the same tank. Each strake has the standard outflow, plus a return line in the upper rear corner of the tank. I bought the recommended Andair “6-port” fuel selector and mounted the valve forward of the instrument panel (a cleaner look, more perceived cabin room). I chose to locate the fuel pumps and filters in the hell hole, increasing the distance between people and potential leaks. I also made covers for the hell hole, both a smaller one just aft of the main landing gear, and the larger one just forward.

With tanks, fuel selector and pumps located, it was time to connect them. I chose to follow the sizing of the Andair selector, with -8 feed lines and -6 return lines. If I had it to do over I would have a -6 adapter on the Andair valve and make all lines with the -6. The flow would have been more than sufficient for the intended IO-360 and -6 would have been larger than the finger strainers at the tank exit. More radically, perhaps the tube bending and flaring and fittings could have been sidestepped with an automotive style flex fuel line. I went with stainless braided flex tube for my brake lines, and my next Cozy will explore flex braided fuel lines.

Tanks. I have the plans fuel tanks, with a finger strainer exit. Fuel lines run underneath the aileron control rod, secured to the side wall with click bond studs and Ugolini-style hold down clips (a customized Adel clamp of sorts). I can remove the sidewall cover and access in-line fuel valves – both valves are secured open and were installed only for future convenience. I have a fuel return located in the upper corner of the tank; any bubbles from returning fuel should be far away from the sump.

Fuel filters & pumps. Below are pictures of the fuel pumps and a “cap” made to hold the filters. This package fit neatly in the hell hole, where it can be put behind a barrier screen. Fuel from the selector comes to the rear, through a bulkhead fitting, across the main gear, through another bulkhead fitting, then into the first filter, the pumps, the second filter and then to a steel firewall fitting low on the firewall. The pressure regulator is mounted to the firewall to the left of the pumps + filters. It's hard to see, but the return fuel lines comes in from a steel bulkhead fitting on the lower left, through the regulator, then curves nicely to the bulkhead fitting. It crosses the main gear, then forward to another bulkhead fitting and onward to the fuel selector. Note that other builders have placed the pump and filters under the front seats - I don't know what's easier, but I've reserved that space for remote mounted avionics, whatever they turn out to be.

(I chose to install cut off valves close to the tanks - the black hexagonal interruptions in the tubing.. They are wired open under the side panels, available only for an annual when I want to service the lines and spill as little as possible (!).)

Materials, techniques & tools. The lines are made of VersaTube, a soft aluminum commonly used for this purpose. A 50 foot length each of -6 and -8 was sufficient. Burst strength varies with size (smaller is stronger), and all sizes are at least 10x the maximum pressure that can be generated by the fuel pumps. Experimenting with the tubing found that smaller sizes (-3) bend easily - I had a 8 foot length I used to template the bending of -6 and -8 tubing. I had _no_ luck with the standard "triple header tubing bender" sold everywhere - I got crushed spots, inconsistent bends and very difficult to get a finished bend out of the tool. I ended up using bending springs - the -6 bends well with an outer spring and the -8 requires both an inner and outer spring. The outer springs come as a set for a wide range of sizes and I found a -8 inner bending spring on eBay. (The inner bending spring for PVC will _not_ work on VersaTube.) My preferred technique came to be putting the tube in a spring and enclosing in a bending jib made from scrap plywood. I recommend you _not_ use the cheapest flaring set, but go up one grade to the $40-50 one. Below is a picture of the tools I used and the scrap I trimmed off when I decided to remake a part. The most common error was to make the tube just a bit too long, so I got to trim and remake the piece. The next most common was to flare before putting the nut and sleeve on the tube!

Most everyone does the flares the same way. My only tip is to add a drop or two of cutting oil on the flare cone - I got super smooth faces. When it comes to fitting pieces, when I couldn't get the nut to to mate with the opposing threads it was where I had the one end secured and was fighting the other. Undo that first end and get alignment at the problem end, after which I was still able to reattach at the first.

Fittings. I underestimated VersaTube, and assumed I would need a lot of 90 degree fittings to get good corners. Instead I was usually able to bend the tubing and get the job done without fittings. (This is preferred by SDS, where they want bends over fittings.) I hit my limit in the 180 degree bends of -8 used between fuel filters and fuel pumps. An inner spring plus an outer spring plus use of the bending jig got my parts aligned and fitted. Deciding how to piece the fuel lines then turns on design for future service. I have 3 pieces in each of the lines between fuel selector and tank, a count that made reasonable the install-remove cycle. 3 x 6 lines = 18 pieces, each with 2 ends. Add in the plumbing of filters and the pumps and you get practiced at making and appropriately flaring the tubing.

Sealants. We pretty much all use Gasoila as the thread sealant for fuel lines. I also used a film of "EZ Turn" fuel lube on the faces of the fittings. A small tube of each is more than enough for the project.

  Jan 26, 2022     fairings, gear leg to fuse - (25 hours)       Category: C09 Lnd Gear

Fairings, Gear Leg to Fuselage

My fairing is detachable, wrapped around the leading edge of the gear leg so that it cannot go through the prop. It was my 3rd set that finally turned out strong, light and with good shape. I recommend the following process:

a) The gear leg is fully faired and painted before making the fairing, with a minimum 3/8ths inch clearance where the leg enters the fuse. The fairing is made with the aircraft on wing stands so that the legs hang as they will in flight.

b) Cover the working areas of the gear leg and plane with packing tape. Do not attempt to form curves with packing tape – it serves as a release wax, not a mold. Have a tape “tail” extending off the gear leg trailing edge.

c) Wrap a 2 layer BID collar around the top of the gear leg. The collar has to extend past the trailing edge, and secure it with plastic clips or clamps. Peel ply the heck out of this collar. Post cure, remove the peel ply.

d) Now use blue foam to form the compound curves. Use adhesive spray to assemble smaller pieces into a block that fits snugly against the packing tape. Shape with a rasp, then cover with micro, then sand to a smooth casting surface.

e) Working on the airplane, mark and cut 2 layer BID to smoothly fit the compound curve. This layer is the upper side of the fairing, extending from the fuse wall, over the transition curve and down to the gear leg collar. The hard part is getting the layer to wrap around the leading edge. Apply peel ply – this will have to be pieced due to the compound curve. Figure out a way to hold this snug against the fuse side while it cures – I used a heck of a lot of plastic sheeting and tape. Post cure, remove the peel ply.

f) Working on the airplane, mark and cut 2 layer BID to fit smoothly and form the lower side of the fairing, extending from the leading edge of the gear leg collar, over the main gear cover and making a smooth join to the trailing edge of the gear leg. Apply peel ply. Figure out a way to hold this against the underside while it cures – I used a heck of a lot of plastic sheeting and tape. Post cure, remove the peel ply.

g) Last layer! On the plane, cut a piece of BID to wrap around the leading edge and back into the upper and lower pieces. Peel ply. Post cure, remove the peel ply.

h) Assess the result on the plane. If 90% on shape, coverage and surface smoothness, OK to remove. (If not, evaluate whether to repair or try again.) Remove by sliding the collar down to the wheel, then removing from this narrower point.

i) Off the plane, inspect again. Repair as necessary. (I ended up adding a quarter sized patch to reinforce a point at the upper edge of the fairing.) Trim the fairings to get good edges and to remove the excess trailing edge “tail”.

j) Fill with dry micro and sand, then epoxy wipe to get a paintable surface. For extra credit, look to the inside of the fairing and use pure epoxy to seal the foam blocks.

k) When satisfied and painted, the gear legs are secured with a trace amount of latex caulk. Dot the caulk on the top 2” of the gear leg, then slide the finished fairing up the leg and into position.

These fairings will have a snug fit on the gear legs and a pleasing shape without excess wetted area.

  Nov 06, 2021     Fuel Pumps       Category: C21 Strakes

I'm using fuel pumps supplied by SDS as appropriate to their system. The twin pumps are Walbro GSL-393 pumps set in a machined alodined frame. Heck, they look so nice I hate to bury them in the back of the plane. The pump can provide 155 liters per hour, or 41 gph. They draw 4-5 amps in normal ops, 5.5 at the max expected draw (20 gph, a pessimistic read of full rich full power) but are to be fused at 20 amps. The pumps run one at time, and I've put the pump selector switch next to the tank selector (switch tanks, switch pumps) hoping to achieve even wear and to always have the backup at hand. From pump selector switch to the pumps is 7 feet each way, 14 feet for the round trip. The wire size calculator says I can use 14 AWG, so another order to ACS... Planning for 20 amps seems crazy, given that the draw is measured at 4.5a. But I'd sure hate to lose a fuel pump on an EI/EFI system.

Walbro's comment on filters: "Note! This pump should never be operated without a fuel filter between the tank and the pump. This pre-filter should be 20- to 40-microns. Additionally, almost all applications require a 10-micron post-filter between the fuel pump and the fuel rail. Putting just a 10-micron filter before the pump is not recommended. We offer a good quality metal cannister filter (disposable) with 3/8-inch hose barb ends as well as the outstanding Holley aluminum billet 390 LPH filter with -6 and -8 AN fittings. The Holley filter can be disassembled and cleaned or the entire filter media renewed." I am using a 30 micron filter pre-pump and a 30 micron after. Both filters are as spec'd by SDS and in SDS enclosures, again beautifully machined aluminum anodized gold.

Here's a link to the Walbro specifications page: https://walbrofuelpumps.com/walbro-gsl393-fuel-pump.html

  Oct 19, 2021     brake lines - (16 hours)       Category: C09 Lnd Gear

- I went with PTFE hose with stainless braid, in a single length all the way from the brake caliper to the cylinder. I bought 25' of the hose, and ended up with an extra 2" - I should have ordered 30'. I ran the hose down the center, inside the cabin heat channel. Routing on the sidewall would call for ordering 35'.
- Braided hose has a limited life - I'm accepting that I will replace the brake every 10 years.
- I used AN3 hardware, with 4 straight ends + 4 90 degree fittings. I should have ordered 2 45 degree fittings for use at the calipers - I think that would have been better clocking. Threads of the AN fittings were touched with Gasoila, the thread treatment approved for fuel, oil and solvents.
- I tore out the NylaSeal that had been installed. I have an artifact - the bit of Nylaflow that was embedded in the landing gear.

Almost all of the Matco hardware was kept back by the Craig W., and I had already bought and installed wheels, axles and brakes. On 10/20 I asked Matco to help by selling the fittings connecting reservoirs and cylinders. That plus the (bad idea) wash bottle are the $45 spend of this entry. See picture for the missing bits.

10/31. Brakes are done, with no visible leads. I'll circle back later and check again for air in the lines.
- I again considered changing over to solid -3 tubing with stainless at the ends, but decided against it. The trade off was burst braided tubing vs extra joints. I like the single piece from brake to cylinder, and running the braid in the heater conduit protects the stainless from most all possible accidents. Prior photo shows NylaSeal exiting the cylinder pressure side; new photo shows stainless braided brake line.
- Fitting the reservoir line required a small adjustment to the clocking of the two fittings on the reservoir. The lines from reservoir to cylinder fitting benefit from a dish shaped cut in the nose gear support, already done when I took the project. See the photo - well done, Chad Robinson. If you want to _not_ do this, space the reservoirs more widely.
- The reservoir is a polycarbonate bowl with a, aluminum cap, and there is a small pressure relief / overflow hole in the cap. A mounting bracket holds the resoir on the wall, above the level of the brake. Note that there is a thin washer between the bottom of the bracket and the top of the cap, creating clearance for that pressure relief cap. One of the washers was lost in the project moves, so I replaced the brass one with a steel one.
- I got some bad advice from a Van's forum, where someone recommended a chemistry lab wash bottle as a better way to pump up brake fluid. Don't do it - the brake fluid nipple is pretty darn stiff and hand pressure won't push the fluid past it. A standard oil can has the necessary pressure. It took surprisingly little fluid to fill the brakes, lines, cylinders and reservoirs - I figure 8-10 ounces. I probably spilled another 4 ounces transferring between containers and having hoses pop off.
- Some Maco brake systems recommend a DOT-5 (silicone) fluid. This one uses the traditional 5606, updated to Mil-H-5606.
- The full kit Matco uses a NPT male-female fitting at the bottom of the reservoir, with a slick washer and integral gasket between the fitting and the polycarbonate bowl. Fitting and bending lines was tight, and I'm a big believer in Gasoila for my fuel fittings. That fuel-proof tread sealant is in place of the NPT + gasket. There is very little pressure at the reservoir side, so I'm thinking I keep an eye on it and figure if I need a different solution.
- Hydraulic fluid seeps, creeping along surfaces and contaminating foam. The bottom of this forward area was already sealed with a fuel-proof epoxy; I went back and sealed the holes for ADSB and transponder antennae. The fuel proof epoxy goes all the way back to the instrument panel.

  Apr 03, 2009     Fuel Valve, Electric Pump, and Gascolator p1 - (5 hours)       Category: C21 Strakes

Note from Chad, the original builder:
"My engine will be a Mazda rotary, which uses a high-pressure fuel injection system with a return. After much debate, I'm following in John's footsteps. Each tank will have a pump underneath each rear seat, which will be tied together at the fuel rail. A return will feed a solenoid that will pass the return fuel to the selected tank. Only one pump will be on at a time. This is a useful setup with some interesting side effects. First, its controls can be simplified to the point that feed and return are switched together, so there is no need for the pilot to manage both devices, and accidental cross-feed is prevented. However, deliberate cross-feed can be engaged when desired. This is a nice feature if you're on the ground and about to refuel. You know the fuel you have remaining is good, so you can pump it all into one tank (which is now a "known good" tank) and refuel ONLY the other tank. You can then take off and perform critical maneuvers on known-good fuel, and only use the new fuel when it is safe to make the switch. If there is contamination, this gives you a response option. You can also use this technique to empty a tank for inspection."

"In the first picture below you can see several of the components in my fuel system, including the pump, tank fittings, and sump level sensor."

PK note 6/8/2020. The project arrived without fuel pump, fuel valve or gascolator. The prior builder even stripped off the AN fittings. I plan on using the fuel flows standard to aircraft fuel injection, with return back to the source tank, and that requires a dual switched fuel valve. I have one of the 6-port Andair valves from a Cirrus, and need to install hard lines bringing the fuel through filters and pumps forward to the dash-mounted valve, then return lines to the source tank and the engine. This is going to be a challenge and learning experience.

(There are wires going into the tank, located at about where the sump would be. This is the fuel level sensor, an array of 5 gas float switches to each side, with a resistor network designed to give output varying with fuel level. It was a clever, if overly complex solution to fuel monitoring, described in C21 "Fuel Level Sensors". I've tied off the wires and will put them under a protective cover for future use by someone else.)

  Oct 17, 2021     Fuel system - (100 hours)       Category: C21 Strakes

Variance from plans - EFI/EI, return lines and location of fuel selector valve. The plans fuel system is designed for a normally aspirated parallel valve O-360; I'm building towards an EFI / EI system from SDS (their EM-5) feeding an IO-360 angle valve engine. This system has a pressurized fuel system with the fuel return flowing back to the originating tank (no sump). The Cozy plans put the fuel valve in the seat back, just like the Long EZ and notoriously difficult to operate in flight. I moved the valve to the instrument panel ("IP").

The SDS system that I'm building towards requires a return fuel line, and getting that return fuel to the correct tank means a 6-port fuel selector. Like most others, I'm using the Andair valve. In my case I got one out of a Cirrus, but it's all the same 6-port valve with permanently lubricated ceramic disks. To fit this large-ish valve in the Cozy, most put the valve on a center console, between the throttle quadrant and the foot of the instrument panel. The SDS FI means that I don't have a mix lever, so all I have is a throttle cable. With only that one control to move, I am trying to have everything on the panel, and avoid having the throttle quadrant taking up people space. That means shoehorning the fuel selector onto the panel. I think I have it done, but it did cost me 100 hours of fiddling, fitting, repairing and replacing. I ended up with a standard Andair face plate, placed low on the IP. Where the Cirrus uses an extension from face plate to valve, I trimmed that to as short as possible, while keeping the extension and the stiffness of that specialty shape. The valve itself is supported at the front by the face plate, and at the rear by a bracket shaped to the nose gear wheel well. There are nutplates on the back of the IP, inside the wheel well, and rivet nuts on the back of the valve. Note how the banjo valves were re-clocked to permit a fit under the rod for the nose gear manual extension. Valve surfaces now have a film of EX-Turn fuel lube. The banjos are safety wired using the same pattern as Andair had on the part when shipped to Cirrus.

Below is the schematic provided by SDS in their implementation guide. Mark Rieger, Cozy builder in California, was kind enough to share his more detailed schematic, with a parts call out. Mark's system is 100% AN-6, but otherwise similar to this build. My tanks have an AN-8 out and a -6 return, and so does the fuel selector valve (as above). I got frustrated with the organization of all the fiddly little AN parts, so I ended up making a paper layout, with post-it notes for each of the AN components (see below). I called out to multiple vendors before finding a combination of 3 that covered the parts I need at a price I can afford. Here is a walk through of the line sizing, as the fuel travels:

Tank to shut-off valve - AN-8 to AN-8
Shutoff to selector valve - AN-8 to AN-8
Selector valve to filter - AN-8 to 3/8th inch NPT
Filter to pump - 3/8th inch NPT to 3/8th inch NPT
Pump to filter - 3/8th inch NPT to 3/8th inch NPT
Filter to bulkhead fitting - 3/8th inch NPT to 3/8th inch NPT
Bulkhead fitting to fuel block - 3/8th inch NPT to 3/8th inch NPT
Fuel block to injectors to fuel block - TBD
Fuel block to Bulkhead fitting - 3/8th inch NPT to 3/8th inch NPT
Bulkhead fitting to selector valve – AN-6
Selector valve to tank – AN-6

Note that fuel lines are segmented for future servicing. (The more segments the more opportunities for leaks, but also the easier to service.

  Oct 17, 2021     Fuel pump, gascolator and filters Category: C15 Firewall

This work is described in Chapter 21

  Oct 17, 2021     Fuel Valve, Pump and Filters Category: C06 Fse Assb

This work is described in chapter 21.

  Feb 12, 2021     07 - Avionics Bus - (18 hours)       Category: C22 Electrical

I made, installed, removed and then repurposed two power conditioners. The more versatile of the two is now used to condition the 5v power that runs to the LIDAR sensor. The brown-out device is still available, and I may use it as a supplemental power input on the Garmin GPS-Comm or the Dynon. I appreciate the several rounds of consultation with Charlie England (AeroConnection guru, #2 to Nuckolls).

At the front of the plane I have gone through a heck of a lot of circuits, and decided to organize them onto 3 buses - a 20-fuse main, a switched 8-fuse for cabin accessories, and an always-on 6-fuse dedicated to the radio stack. At the rear of the plane I have a 6-fuse main and a 14-fuse engine bus. All in, that's 50+ circuits!

  Jan 20, 2021     82. Making cables - (10 hours)       Category: C22 Electrical

I've been making my own power cables, connecting battery to contactor, contactor to starter, main fuse block (rear firewall) to the front fuse block (instrument panel), et al. I started with the Bob Nuckolls recommendation for fat wires - make cable ends from copper pipe. The result was great in carrying current, but way to difficult to get good looking ends with smooth bolt holes. I bought an assortment lug ends and a hydraulic crimper. The result has been much better.

One of the lug ends had too small a hole, so I had to replace it. That was a chance to cut into the lug and check how my squeezing was doing. OK! See the two photos below. The first one shows the cable just outside the lug. You can see the individual wires of the 4AWG cable. The second photo shows a cross cut in the center of the lug. Under compression the wires merge into a solid mass, and there is no space between the copper and the tinned copper wall of the lug. I didn't get that solid metal look all the way across the cable, but there is still one heck of a low resistance connection. I don't know that all the cable ends are this good, but this one looks acceptable.

The spend was on a hydraulic crimper (Klutch brand from Northern Tool) and an assortment of lug ends (Amazon, their larger assortment). Make sure to get copper, tinned, with a tang end opening so you can add solder if you want to.

  Oct 08, 2021     Ballast - (6 hours)       Category: E01 Next Steps

I was able to get lead tire weights from an auto shop. I cast them into the shape of small muffins (the steel mold I found on hand), then got smarter. I shaped wood blocks to fit in a compartment created by structure holding the nose gear. I then built molds to fit the plugs. The weight is far enough forward (and I'm big enough) that I don't expect to use any more ballast than this 17 lbs, even when solo. But I have two more large bricks , able to shift it between the forward compartment and a holder just behind the front seat.

  Oct 10, 2021     Nose Gear refinement - (12 hours)       Category: C13 Nose+Gea

The project as received had some rough spots, and one was the nose gear. Here are the things I found and fixed:
- The EZ Lift attaches to the nose gear strut with a metal clamp. It was held in place by an AN3 bolt with no nut. Replaced with the correct AN4. Where the installation instructions use a lock nut, I see this as rotating through 15 degrees or so, so I went with a drilled nut, castle nut and cotter pin.
- For the past decade or so the user group has called for a bolt securing the metal clamp to the gear leg. The bolt is to be transverse through the no stress center of the gear leg. (The front of the gear leg is in tension, the rear in compression, the center is neutral.) I installed an AN4. Even as this bolt does not move, for consistency I went with drilled bolt, castle nut and cotter pin.
- One gear door was torn out. I prepared flox beds for small screws and installed the missing door.
- I found and fitted a door spring that holds the door open through taxi and take off, then has the retracted gear pull the door closed. This is a fairly standard design for EAB aircraft with retractable gear.
- I cleaned up the gear door edges so that the doors open and close smoothly. This amounted to some light filing work.

  Jul 02, 2020     Prime & top coat - (100 hours)       Category: C25 Paint

Here are pictures of my test panel and then post-paint airplane parts.

I used two test panels and did what I could to screw up the chemistry of the paint. I put the primer and top coat on with and without
- surface de-greasing
- sanding
- conversion time
- brushing vs rolling

Those experiments gave me an understanding of what it might do. There were no paint failures, but I still did my best to follow directions when applying paint to real parts.

Three of the pictures show a board bolted to the wing spar, I used this in rotating the plane; my engine hoist was raised as high as it would go and the nose supported on saw horses. The fuse was rotated on its longitudinal axis and set on wheeled sawhorses. Other builders have much higher shop ceilings or armies of manly friends. I dinked around with various approaches until I found this, which worked for me.

  Jul 12, 2020     Lights: Nav, Strobe, Land and Taxi - (16 hours)       Category: C22 Electrical

I started with the Whelen wing tip lighting, the "Microburst III". There are less expensive choices, but I spent a lot of time in the CT aviation crowd, and learned respect for the Whelen team and products. Worth the $450? I'll find out over the long haul. The sooner you install the lights the sooner the lenses will break. I'm getting all the connections ready, but I won't do final install until I am at the airport. At the wingtip the connectors are DB pins, crimped, with shrink tubing with inner glue used for strain relief. DB pins are rated for 5 amps in low density loads. This load is significantly less than the rating of the DB pin connection. The connections at the wing root are shelled DB connections - there is room for a connector and it is possible the wings will come on and off several times.

The FAR's governing this area were clarified with Advisory Circular 43-217 dated 12/12/2018. At section 11.7.1 the AC says the required light intensity is in 23.2530, which says "Any position and anti-collision lights, if required by part 91 of this chapter, must have the intensities, flash rate, colors, fields of coverage, and other characteristics to provide sufficient time for another aircraft to avoid a collision." The AC also says that Part 23 aircraft certified prior to August 30, 2017 comply with 23.1405, which gives a table for required output measured in "effective candela value". At a measuring distance of 1 meter, the values for candela (lumen per steradian) and lux (lumen per m2) are the same and a candela is to 1 lumen x square meter per steradian. Per the FAR's, the light should put out 400 ECV at the 0 horizon. My non-calibrated light meter shows light output of 421 lux, so I believe I comply with both standards.

Wire sizing From AC43.13-1b: "Wires must be sized so that they: have sufficient mechanical strength to allow for service conditions; do not exceed allowable voltage drop levels; are protected by system circuit protection devices; and meet circuit current carrying requirements." With a nominal 14 volt system the allowable voltage drop is 0.5v continuous, 1v intermittent. As applied to the wingtip lights....... Total amp draw at 14v = nav 0.25 continuous + strobe 0.2 average and 1.2 intermittent. Each wing 14', total single wire run estimated at 16' x round trip = 32'. Fig 11-2 (continuous) says 22 awg handles 1a for 30', so redundant for the nav and average strobe draw. Figure 11-3 (intermittent) says 22 awg will handle intermittant 1.5A in a 40' run, so good for the strobe intermittant. The ground has to handle both loads, so 0.45a average and 1.45 intermittent. I upgraded the shared ground to 18 awg - better safe - and fused the 2 circuits (Nav & Strobe) at 5a each. The old fashioned strobes handled a lot of current and created interference for the com circuits. LED strobes are low current, but still a strobe and run alongside antenna wires. For the wing portion a good twist in the wires, and at the wing root it connects to shielded twisted pair. As a caveat to the tables, I note that Whelan manufactures these lights with 4 leads of AWG 28 or so - those are surprisingly fine wires.

The nose holds landing and taxi lights, a pair of 6-LED lights in aluminum cases which are finned to act as heat shields, with shock resistant design (intended for use on ATV's). I modded the cases for fit in the nose cavity, then attached to the bottom and side ribs with dabs of flox. Connectors are DB pins, crimped, with shrink tubing with inner glue used for strain relief. DB pins are rated for 5 amps in low density loads. This load tests at 0.9 amps at 12.8 significantly less than the rating of the DB pin connection. The wire run is about 3 feet x 2 = 6, so 22 AWG is redundant, fused at 5a.

I made and rejected multiple lenses - 5 in plexi and one in Lexan, before learning how to make a good Lexan lens (heat very slowly, nudging up the oven temperature to a soft slump. The lens was trimmed with bandsaw, then sanded to final profile and attached with silicone. The nose compartment is unventilated and the LED's generate heat. I placed a probe between the lights, sandwiched between the radiator fins. The LED's raised the fin temp from 88F (garage temp) to 120 (15 minutes) and then 141F (30 minutes). The fins were uncomfortably hot to touch. The surrounding composite surfaces were only slightly warmer than the air. With landing and taxi lights typically on for less than 15 minutes and a transition temperature of 165F (50C), I am satisfied that this solution works for this application.

10/18/20 - I came back to the lens and decided it needed a better shape. I rebuilt the base with 2 plies of BID, then a foam micro slurry, then micro. That was sanded flat. I cut the existing lexan lens to have a matching flat, and attached with the flexible and durable "Shoe Goo". See photo. I then faired with a wood flour + epoxy mix.

  Sep 26, 2021     Cabin Accessories       Category: C22 Electrical

Several of the accessories are switched at the device or somewhere other than on the instrument panel. I created a separate bus to handle these loads:
right USB port (3a) left USB port (3a)
right seat heater (5a) left seat heater (5a)
cabin heat blower (3a) defrost blower (7.5a)

I had just finished the radio stack and immediately put the sub bus on the side of the radio stack, and moved the ground out from under the radio stack. The new look is easier to service, with the tradeoff that removing the radio stack now requires disconnecting that cabin accessories bus.

  Sep 15, 2020     cabin heat - (2 hours)       Category: C22 Electrical

There are two blowers that address cabin heat.

A plans blower moves heat across the exhaust muff and pulls it forward. That is done with a marine sump blower, drawing a max of 1.2a and fused at 5 amps. The harness appears to be equivalent to 24 awg. I will follow the FAA rules on load, distance and allowable drop. I will install a heat duct from the engine compartment forward to the passenger footwell, with the motor forward of the passenger rudder controls.

A defrost / avionics fan pushes air across the radio stack and then washes the front of the canopy. This is a 6 amp rotary fan The two work well together, able to spin the motor at all settings and drawing 6v at the half-way setting and full battery voltage at the "max" setting. The manufacturer wired the fan with 24 AWG; my wiring will be to-from the IP connect points, fused at 7.5 amps, wire as per the FAA rules.

I also picked up a nice 10 amp pulse width modulation rotary switch. It is ideal for these motors, but I'm not sure it makes sense. When you want heat or defrost, you want it full on. Is there a use for the switch?

  Jul 01, 2020     seat cushions - (30 hours)       Category: C24 Seats+Fair

The seat panels alone are surprisingly comfortable - that's the right curve for me - but I need something that works for varying size people and, right now, I need some idea of how seat cushions will affect the leg spacing. I ended up with a thin hard cushion permanently attached to the panels, a removable thin cushion that I take in and out, and a set of removable standard cushions.

For sitting in the plane and thinking about panel layout I used a thin pad glued onto the fiberglass. The thin pad is "Anti-Fatigue Foam Mat", at Harbor Freight for $9 for 4 squares 2'x2'. With these seats being the plans 17" wide, an 18" wide cushion will just fit. For a removable layer on top of that I've used a Home Depot distributed Hampton Bay "sling chair cushion" in "chili stripe".

2/27/21 - 3/12/2021. Refining the cushions. For longer term use I made a set of removable standard cushions, as follows.

Front. Foam is 2" of medium density topped with 1" of memory foam, wrapped with two layers of polyester batting. The foam was a 24" x 48" block, which was sufficient for 2 seats and 2 backs. ($65 at Amazon.) The fabric is blue Olefin, made for outdoor upholstery - I bought a 5.25 yard remnant 54" wide and used 2 yards for the front seats. I like the ridges sewn into the cloth and hope that minimizes slipping. There is an auto seat heating panel under the fabric, both seat and backing, which will tie into a 12v switch on the panel, and join the "Cabin Accessories" circuit. The backing is fiberglass (2 layer BID + peel ply) cast to fit the seat and back of the front seats. I found that Velcro just won't stick to Olefin and the cushions slipped around like crazy. I re-assessed. I turned the seat plate into a foam-BID sandwich panel, with front and rear edges that hold the seat panel. I recast the seat back panel for a height matching the setting in the plane. I did the back panels first - on all 4 edges aluminum strip with Tinnerman clips, fabric wrapped around the strip, then short screws through the 2-BID backplate into the Tinnerman clips. Super light, but not much tension on the fabric. For the seat panels I echoed the plans seat pan, with a foam core sandwiched between 2 layers of BID. On both seat and back I used spray adhesive (3M #77) and for the seat I had foam structure I could use to hold staples. Seat back detail - the pilot side does a nice job of holding itself in place, while the co-pilot side has two pins that push into the seat back.

Rear. 2" medium density foam was cut to the panel size, then a diagonal cut taken at the front and back edge or top and bottom edge, for seat and back panels respectively. Wrapped in 2 layers of polyester batting, then the blue olefin fabric - 1 yard for the 4 panels. The fabric is secured with 3M #77, stapled also. The piano hinge were cut free from the back surface, so a small fabric tab covers the visible part of the hinge. The result looks spartan but is quite comfortable.

Heaters: I went with an automotive heating pad - . The heater is the white sheet shown in photo "seats2.png". For each of the 4 heating pads I routed the plug to the outside, where a handy disconnect meets a 2-pin plug. Both back and seat use a panel mounted rocker switch in the lower outside corner, with high and low settings. When off the draw is zero, at low heat the draw is 0.85A per seat, at high 2.70A per seat. Each side has stand alone wiring, with the switch firing a relay that handles the current draw. Each sides comes with a blade fuse at 10A, saying the manufacturer trusts the wire to that current. I don't have that trust and downsized the fuse to 5 amps. I have panel out for repainting. Once back in the plane I'll install the wiring and switches.

  Sep 17, 2021     07 Repairing Seals on MS20271 Universal Joints - (5 hours)       Category: Z-Tools+Tips

Repairing Seals on MS20271 Universal Joints

  Dec 08, 2020     Aux Alt tray - (6 hours)       Category: C22 Electrical

I'm using the permanent magnet alternator spec'd in the Nuckolls architecture, manufactured by B&C Electric and sold as the "SD-8". The kit includes the SD-8, over-voltage protection, a 20 amp relay, regulator, capacitor and instructions. I installed it consistent with the instructions - horizontal on aluminum. (I spent a week of 1-2 hour sessions doing this 4x.) At the end, what you see below - scrap aluminum turned into a tray to hold the switchery of the aux alternator. Two rivet nuts in the aluminum plate catch 6-32 screws holding the red tray to the firewall. Using the tray allows me to horizontally mount the regulator on a cooling surface, and, of course, to do the connections at the bench.

The big blue cylinder is a capacitor, included in the PM-8 kit sent from B&C. That plan was intended for lighter weight engines - think an O-235 in a Long EZ. With the cap the alternative alternator has a starting charge. If and as I had to hand prop the engine, I could switch to the auxiliary alternator and this cap would support the rectifier. There's not much reason to take it out, but it's not in the Nuckolls diagram and not something I'd do again. Nuckolls said not to worry about it - it certainly won't harm anything (so long as I don't lick the terminals!).

  Jan 26, 2021     Test cards - (1 hour)       Category: E01 Next Steps

There is a Cozy-specific flight test program. I supplemented it with the EAA materials. Good job, Kevin Walsh! Good job, Home Builders Council!

  Jan 25, 2021     Certification - (5 hours)       Category: E01 Next Steps

I'm using two reference guides to the certification process. The first is an FAA-assembled selection of federal rules and guidance. I picked this up at an Airventure back in 2002 or 2003. The second is the current edition of the EAA's certification guide. It's got everything from panel labels to sample forms - certainly worth the $15 + shipping & handling.

  Nov 09, 2011     Nobody Will See It... Except the Kids - (5.5 hours)       Category: C25 Paint

I took an extra hour while prepping the wing root hardware to paint the major surfaces (the control tubes and mounting brackets) in contrasting colors. That'll make the boys happy. I know nobody else will see it - but I know it's there, and it looks a lot cooler than I expected it to. You can really see how the parts move together to make the aileron function. Shown here are the FMN10 bearings floxed into the wing root, plus related hardware. I didn't finish the installation because the bearings were still curing - next time I'm out I'll trim those up a bit and bolt everything down for good.

  Oct 25, 2020     06 Jack stand - (8 hours)       Category: Z-Tools+Tips

The Cozy is difficult to lift using normal jack stands, especially if wheel pants are installed. Courtesy of Marc Zeitlin, one solution is to lift the plane at the spar, at or just outside the covers for the wing attach holes. Here's the procedure.
1. Lower the nose gear and put the plane in the kneeling / camel posture.
2. Use saw horses or blocking to create a stable surface 40” high.
3. Nudge the blue block forward under the kneeling / camel posture plane.
4. Operate the nose gear. As the nose comes up there will be a lever action rolling the underside of the wing up the ramp of the blue block.
5. Stop when the nose gear is fully extended. In my case the wheels float 1.5” off the floor.

Having proven that this works and found the required height of the platform, I built custom sawhorses for long term use. If and as I want a jacking kit in my traveling tool bag, I can travel with the foam blocks knowing they work when 40" off the ground. Here are the dimensions for the sawhorses. Using those clamps, the legs were 37", the braces 15" and 18". Top piece was 18", the platform was 16" x 8", made of 1/2" ply. If I had to build a set while on the road, it is four 8 foot sticks of 2x4 (legs), plus one 8 foot piece of 1x4 (braces), plus some scrap 1/2 inch plywood. They seem pretty darn sturdy.

Friggin' Florida. Late October and puttering in the garage leaves me drenched in sweat.

  Jan 01, 2010     05 Hot wiring foam       Category: Z-Tools+Tips

Hot Wiring Tips

My hot wiring experience went very well, thanks in large part to John Slade's assistance. We discovered some tips and tricks that I'd like to pass along to other builders. Our cores looked almost laser-cut when they were done, using the techniques described below.

Saw Construction

I've seen a number of methods of constructing a hot-wire saw. Ours cost less than$8, and was very fast and simple to assemble:

1. Cut 3/4" electrical conduit to 18" (small saw) or 24" (large saw) lengths.

2. Cut a 2x4 to 46" (small saw) or 64" (large saw). Bigger is not better. 64" will handle the biggest cuts you need to make, and you can stand close to the templates for better control.

3. Use an 11/16" spade bit to drill holes in each end of the 2x4 at a 15-degree outward angle.

4. Pound the conduit into the holes with a hammer. It'll be tight, but this friction gives good tension for the wire.

5. Drill small holes in the ends of the conduit for the hot wire.

6. Install the wire and wrap it around itself to hold it in place. I bought 40' of 26ga nichrome wire on eBay for $5.

7. Tighten by turning the conduit with a pair of Channel-Lock pliers.

8. Install conduit couplers at the base of the saw to provide an attach point for the cord. Alternatively, drill the conduit and install a self-tapping screw.

9. Use a 12' extension cord with its end cut off for the power cord.

Constructing the saw took only a few minutes. We used conduit joints to hook up the electrical wires - see picture 2 above. A pair of Channel-Lock pliers made an easy tightener - you just turn one arm a bit, until the wire "sings".

Cutting Guides

The plans include a set of straight edges, but John and I discovered two much more useful options.

The first is the corner of a small sheet of melamine / Formica. You can get this in 12x24 pieces cheaply, and the corners are already perfect 90-degreeangles. Each piece gives you four triangles that can act as squares. If your table is flat, you can very quickly set up a perfectly vertical cut when joining blocks.

For longer, or angled cuts, we used two drywall squares. We drilled out the rivets, and separated the two pieces. We then drilled holes about every 6" for nails. These made great short and long straight edges, and we've since found dozens of uses for both lengths in other chapters of the project.

It helps to make a few cutting tools/guides. Here you can see a pair of triangles cut out of the corners of a piece of melamine, with a factory-cut 90-degree corner. Verify your corners - not all sheets are exactly 90 degrees! With a pair like this you can make perpendicular cuts in just a few moments - set the block flat on your bench, and the triangles will automatically give you a vertical cut! In the background of the second picture you can see two aluminum straight edges, made by cutting the cross piece off a drywall square. The inch markers are handy both for measuring and as hotwire talking points.

Lessons Learned

We learned a number of things that make life easier while cutting:

1. Don't leave the saw on between cuts. This fatigues the wire more quickly.

2. Tighten the wire for every cut, AFTER turning the saw on.

3. Replace the wire occasionally because it will fatigue. Two replacements over the course of cutting all of the cores would be a good guideline for nichrome.

4. Split a mixing stick about an inch down, and once the saw is hot, run it down the wire. This cleans the wire much more effectively and quickly than sandpaper, with no damage to the wire. It is very important to have a clean wire - the quality of the cuts is much higher.

5. If you come off the template and get a bump, don't spline-sand it off. This takes more time than it's worth. Hit it quickly with the hot-wire saw to define the shape, then a bit of sanding will take off the residue left behind. It's more accurate, and much faster.

Half a popsicle stick, carefully split for about an inch, makes a great wire cleaner. With the wire hot, slide the split end over the wire, and run it back and forth a few times. All of the melted foam will get scraped off by the stick, without harming the wire as sandpaper would.

Hot-Wiring Micro

Yes, you can! Wicks had a fire sale on foam blocks that I jumped on. They weren't the highest quality, and they weren't the large size - they were 7x14x41. We had to piece them together to make the wings, and in doing so ended up with a micro joint before we had made all of the cuts (in the aileron torque tube hole area). Think this is a problem for a hot-wire saw? Think again. Our cuts were perfect - circular, and went right through. You can't do this if the micro is fully cured; ours was only a day old. I'm not recommending this, just saying if you happen to have a cut you forgot to make, and now have to make it through micro, don't panic - give it a shot, and you may be pleasantly surprised.

Stopping Midway

Finally, a well-known trick is to put popsicle sticks across the spar cap depressions and cut those in a second pass. The idea here is that the wire will be lagging in the center of the cut, and even a two second pause may not be enough to let it catch up. Cutting a curved trough is not good, so a second pass ensures they will be perfect.

We took this a step further. We started at the leading edge, cut to the center of the popsicle stick, paused two seconds, then lifted the saw out. We then cleaned and tightened the saw, and cut the other side by starting at the trailing edge. The idea here is that the start of the cut is always the best part - you have the tightest possible wire, no lag, and the wire is hot so it makes its best cut. It's not uncommon to see cuts where the leading edge looks great, but there is a curve in the center of the trailing edge that needs to be spline-sanded out due to the wire lag. By making the cut in two passes, from the edges to the center, you don't see this problem. We used the same technique in the spar cap trough, and they looked perfect. This technique also gives you a chance to rest your arms halfway through the cut, shift your feet to a new position, etc.

  Jul 31, 2021     03 Fiberfrax alternative Category: Z-Tools+Tips

The gear leg axle gets hit with heat from the brake, most of which travels by radiation. In response, we cover the gear leg tab with insulation and aluminum tape. I don't have the FiberFrax remnants from the firewall build, and found that Fiberfrax isn't available in the small pieces I need for that work. Instead I found the insulation "paper" used in pottery kilns. It has the same performance numbers as Fiberfrax, can be cut with scissors (far fewer fibers in the air), and Aircraft Spruce recommended it as the new best thing. I was able to get it on Amazon, but the manufacturer dropped the Amazon relationship, so look elsewhere. Here is the description: "CGjiogujio 2pcs 6mm x 12" x 13" Ceramic Fiber Heat Insulation Blanket Paper". I paid $18.50 and got twice what was required for the two gear legs.

  Feb 17, 2021     02 Piano hinge lube - (4 hours) Category: Z-Tools+Tips

Piano hinges get contaminated by glue, epoxy, paint and old-fashioned dirt. When I reinstalled the nose gear covers the hinges were so stiff that the spring would not push the doors open. I found the best lubricant was _not_ LPS-3, graphite or petroleum-based. I got a better, cleaner result using "Lightning Dry Lube", a product made for bicycle chains. The lubricant is a Teflon powder and the medium is a mineral spirit. The product wicks along the pin, so application is a very small amount applied to the piano hinge's interleaving spaces.

  Mar 28, 2021     02 Wing Trolley - (15 hours)       Category: Z-Tools+Tips

Taking wings on and off requires some fine tuning of the bolts and bolt hole alignment. Rather than rely on neighbors or risj dropping the wing, I copied the Bernard Sieu design for a wing trolley. Scott Fish shared the design with the Cozy group, and I in turn shared the cutting plan.