Even though Version 2 of the avionics ground bus was a big step up from Version 1, I found that I still had both a philosophical problem with its construction, and a mechanical one. Here's the first problem: The main ground wires (the ones that run from this bus to the firewall ground block) are still soldered to the D-sub connector pins. Because of the limited visibility and access to the backside of the pins (see previous entry), it's tough to know if these connections were made well enough to survive a long life in an airplane. The soldering was a bit tricky to do, and I just can't risk the slightest question in my mind about whether these connections will be be reliable long-term. If they were to fail, my entire avionics package would fail. Bottom line: Soldering ganged wires across the back of a socket that's not designed for this application just seems a bit... chintzy. Especially for an IFR airplane. So today, I taught myself to use some design software and used it to design a printed circuit board (PCB) that will solve this problem. The D-sub socket will be soldered to the PCB, and the connections will be ganged together by the foil traces on the PCB. The ground wires will then be soldered to pads on the board and strain-relieved inside the case. This should be a foolproof way to ensure the quality of these critical connections. I sent my finished design to a prototyping company for rapid manufacturing, and I should be receiving a prototype custom board in a few days. If it works well for me, I may offer the boards for sale, either as a kit or a complete unit. Second issue: Version 2's pins will not click into place like Version 1's, since Version 2 deleted the male socket entirely. In other words, the pins need to click into a female socket mechanically, or they'll just pull out. Duh. I'll go back to using two sockets mated together.