« Archives in August, 2011

Back to it: ELT and radio stack.

4 hours.

Finally got back into it today. First thing I did was dismount the EFIS in preparation for making a hole for the ELT remote switch. Then I reinstalled the radio trays back in their spot so I’d have a stiffer panel to work with during the process of making the hole. I still need to figure out where the ELT wire’s going to make its way to the panel. Oh, and I realized I need another roll of 3-conductor shielded wire, because I also have to get GPS data back to the ELT as well. I guess it transmits the last known position while it’s screaming for help.

I need to start finishing things in the back of the baggage compartment, because not only am I tired of crawling around back there, I need to not have that to worry about while I’m finishing up the wiring.

I also have some 4-40 screws on order from ACS, because the ELT kit doesn’t actually provide any. I also have a BNC-terminated antenna on the way.

Wiring harness done. I think.

10 hours.

This is what I wound up with at the end of yesterday:

It might look like a mess, but it’s not. Every pin in every connector has been checked twice, and the coils are grouped by device and function. Except the one or two I missed. This has been an exercise in patience and squinting at little wires. It’s like everything else in this project: If you look at the whole thing it looks insane and impossible, but if you look at it one piece at a time, it’s not that big a deal.

Each wire only has two ends, so that makes a good place to start. First I did all the power and ground connections. There are several. The 430 has 3 devices in one, but only two have their own power feeds, unless I’m driving a superflag device off the NAV unit, which I’m not. The audio panel is the hairiest part, because there are so many connections and a couple of them share pins, which sucks, because you have to make pigtails that combine two wires, usually audio returns, into a single pin that plugs into the connector. The shields absolutely suck. I probably should have used 24ga wire, but I had green 20ga wire, so I used that. That’s where I learned the Aeroelectric (and probably standard) way to bridge shield to a standard wire. That wasn’t that difficult, but there were a lot of them, and now they’re all ganged together and fastened to the audio panel’s frame. The deal is, you ground the shields at the audio panel to reduce noise, or at least attempt to do so. Supposedly this is how you keep EMI noise out of your audio connections, which tend to run all over the aircraft and often past some noisy things like strobes, relays, etc. Now, hopefully, in crushing together all those cables, I haven’t compromised the insulation integrity, and during the soldering process, didn’t melt through any signal wires. I still need to verify connectivity and isolation, but I’m off to Vancouver for SIGGRAPH, so that’ll have to wait.

I couldn’t just leave it be, though. I powered up the audio panel, which lights up fine, but I didn’t want to risk powering up the radio. If something is screwed up in my wiring and the mic key is stuck on, the RF power with no antenna will fry me out of $7500 worth of radio, GPS, and NAV equipment. Not to say I didn’t try. I had some leftover RG400, which I put a BNC end on, and was going to connect the other end to my COMM antenna, but that didn’t happen. RG400 cable has two layers of shield on it, so the usual trick of pushing the shield down and pulling the wire through doesn’t work. My current RAMI antenna needs pigtails for wiring, and dammit, so does the marker beacon antenna. So I need a COMM antenna that uses a BNC connector like everything else.

One thing I missed while I was wiring is the COMM remote recall function. This uses a discrete input (fancy name for an external button) to scroll through a list of preset frequencies. It looks almost useful enough to take apart the main connector on the GNS430W and add a tail for it, but I don’t actually have a free button for it. If I didn’t do remote ident on the transponder I could. The other suck part about the GNS430W is that it doesn’t have the same remote function protocol as the SL30, which can be fully controlled via RS232 communication. Sure would be nice to be able to set frequencies from the EFIS, but this setup will be fine as is.

I also got a phone extension cable to make up for the shortcomings of the one shipped with the ACK 406 ELT, which doesn’t quite reach all the way to the panel after threading its way through the wire bundle. I also figured out how to install phone cable ends, a skill useless in every other facet of life, since I don’t even have a landline anymore and subsequently will probably never wire a phone jack in this house. The new cable, with the end cut off, goes fine through the portholes I drilled for the strobe cable. Yes, I know it’s probably bad, but as long as the strobe noise doesn’t trigger the ELT’s emergency transmit mode, I can live with it.

When I get back, I’ll also have to dismount the EFIS to install the ELT remote switch and verify GPS communication. Then my time in the tailcone is DONE. I’ll reinstall the flap arm and the flap motor and call it a day.

The other thing I have to do is run another wire for push-to-talk on the pilot side. The audio panel has facilities for multiple push-to-talk sources, diagrammed by shorting the PTT circuit to the MIC return, which is the same thing as shorting it to ground. Right now, I have both PTT buttons coming out of one wire, and I need two, one for pilot, one for copilot.

I also need to lay out my wire codes on here so I have a record of them.

Phone cables suck.

2 hours.

I don’t know whose brilliant idea it was to use terminated 4-conductor phone cable for the ELT’s GPS data, but they should be kicked in the groin until they pass out. First, the ends don’t fit through the pass-through holes in the bulkheads so the cable can get from behind the baggage compartment all the way up to the panel. Second, the cable provided is too short, by about 5 feet. Third, the end connectors are single-use, and taking them apart (to run the cable through bulkheads) pretty much destroys them so you can’t use them again.

After utter failure on that front, I went back to wiring the avionics stack. It’s progressing nicely. Still, Shelley commented that it looks like C-3PO had diarrhea all over the shop, indicated by this shot:

The GMA340 audio panel is a wonderful thing. Not only do I have inputs for music, 4 radios, various navigation equipment, and an antenna. I also have an altitude warning audio input. I need to do some research, but that seems like the logical place to run EFIS audio. So far I’ve got almost everything half-connected. The next step is to connect the devices together and make nice wire bundles per AC-43.13 or whatever it is. I’ll be using zip-ties, because as cool as lacing is, I want to get this done fairly quickly.

I’m about to bugger off to Vancouver for a few days for the Gathering of Geeks known as SIGGRAPH, so I’ll probably be out of the game until next Friday, but if I can get some stuff done this weekend, I will.

Avionics harness continued.

7 hours.

That’s 3 yesterday, 4 today.

The two biggest challenges for me in this wiring harness debacle seem to be keeping good records (which colors go where), and puzzling out the connections between pieces of equimpent. The GNS430 and GMA340 have way more options than I’m ever going to need. Most of the 430’s outputs are for various navigation instruments that I don’t have, and the audio panel has options for 6 passengers and various pieces of navigation equipnent like DME and ADF, neither of which I will be using. The 430W to EFIS connection is dirt simple, six wires and the EFIS makes use of the ARINC429 data feeds for fully legal IFR navigation. The connectors themselves, however are quite intimidating. The 430W has 3, two of which are high-density d-sub connectors, but very few pins are actually in use.

But the key is to take good notes. I’ll post my wiring system here next chance I get, but I’ve basically set it up this way:

For mono audio, White is HI (+) and white/blue is LO (-). For stereo audio, White is return, white/orange is right, white/blue is left.

I wrote something down for the ARINC and Mic connections, but I don’t remember it offhand. It’s important to take notes, because when the backshell is on the connector, you can’t see the wire colors anymore. I’m sure this is old hat to you avionics experts out there, but my experience runs more toward DJ and club sound systems, MIDI keyboards, and computers, so this is indeed different.

I also make my first shield ground pigtails today. This is the Nuckolls method. You strip off the insulation on the shielded cable, then cut the shield back so only 1/4″ is left. Then you make a 22ga pigtail. On this, you strip off about 1″ of insulation, then unwind 4 strands, then cut the rest of the wire off to 1/4″ of length. Lay the 1/4″ of wire along the 1/4″ of shield, then wrap the strands around both. Then flow a little solder into the connection, cover with heat shrink tubing, and done. This is how all the audio panel connection shields are grounded.

Shields are all grounded at the audio panel, because that’s the best way to eliminate noise.

OH, and for some reason, I seem to have the worst luck with all the litte attachments for my d-sub pin tool. I’ve already mangled the lo-density attachment, and today, I managed to get one caught in the mechanism of the AFM8 crimper. Freak accident. But fortunately it wasn’t an attachment I’ll need for this project.