What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Converting a GTX-327 from Parallel to Serial Altitude Encoding

snopercod

Well Known Member
I decided to join the 21st Century and change the altitude encoding inputs to my Garmin GTX-327 transponder from parallel to serial, driven by my Dynon D-6. Sadly, six years ago when I had the harness made up for the GTX-327, I didn't instruct the shop to add the shielded pair(?) for the RS-232 serial encoder input to Pin 19. My first plan was to extract the unused wire going to pin (socket, actually) from Pin 20, and move it over to 19. No matter how hard I tried, I was unable to get the pin extractor in the hole, so I gave up and moved on to Plan B. I begged a socket from the FBO and added a new wire.

The wiring diagrams for both the GTX-327 and Dynon D-6 were unclear, especially regarding the shield grounds. The Garmin just showed a ground symbol for the shield, but didn't specify case ground, power ground, or signal ground. I chose signal ground but, in retrospect, I don't think it made any difference. The Garmin schematic didn't show a "return" or "common" wire for the RS-232, so I guess the ground was it. I still don't know if Gamin intended a shielded pair or a single shielded wire for the RS-232 IN. The Dynon schematic didn't show shielded wire at all, but there actually was a shielded pair coming out of the harness. It contained a "return" line for the RS-232. The wire color in the manual was incorrect, but I figured out which wire was which with my Fluke.

Without going into the boring details, I connected the Encoder Out wire from the Dynon to Pin 19 of the GTX-327, and all the shields and the second RS-232 "return"? wire from the Dynon to Power Ground with ~8" #22 pigtails. I programmed the ALTENC setting on the Dynon to "Format 4" for the GTX-327, and on the transponder, set "RS-232 Input CHNL 1" to "Icarus" per the manual. I was as surprised as anybody when it actually worked. Now I guess I'll need to get another transponder check in order to remain legal.

To work on the GTX-327 connector, it must be removed from the backplate with a long, skinny, screwdriver (two Phillips screws). Then the backshield clamp can be removed and the shroud retracted, exposing the back side of the DB-25 connector. Hint: The backshield clamp can NOT be reinstalled with the connector in place. Ask me how I know. :rolleyes:

The two black wires are the "power ground" and "signal ground". You can see they are spliced together so apparently it doesn't make much difference. (Ignore the tygon fuel vent passing through the aluminum without a grommet. I'll put the grommet back tomorrow.)
IZbw7I.jpg

The die cast "keeper" is open on the bottom, and fits over the connector and holds the connector to the back plate. It's screwed in from the inside with a long, small, screwdriver.
86o4CA.jpg

First, gently cut the outer jacket with an Exacto knife and pull it off, exposing the shield braid.
YQWDHV.jpg

Push the braid toward the cable to make a "bird cage" at the bottom
9v0fIr.jpg

With a sharp pick, make a hole in the braid and enlarge it so you can insert the pick and pull the conductors out the hole
bQt00w.jpg

All done. Snip off the messenger strip.
ESplo7.jpg


Next, I soldered a #22 wire to the shield and protected the junction with heat shrink (no photo, sorry.) A solder sleeve would have been better, but I didn't have any.

Tools
9JTFkr.jpg

On the GTX-327, I changed RS-232 Chan 1 from OFF to (in my case) Icarus per the Garmin instruction manual
bH3Tbc.jpg


Success. The darn thing actually works in serial mode. The parallel wires are still hooked up so, in theory, I could change back by resetting CHNL 1 to "OFF"

Tomorrow I have to re-dress and Ty-Rap all the loose wires in back of the panel.
 
Congratulations. Those connectors are a pain, especially if you have to work on your back.
RS232 protocols are ?high level? ( several volts). Many people have used unshielded wires without any problem, although shielded is always a better bet. As you figured out, return signals are via a common ground. No twisted pair like ARINC. Standard technique is to ground just one end of the shield, to avoid ground loops.
I probably shouldn?t say this, but I probably wouldn?t pay for a recertification (unless it is due soon anyway.) You didn?t change the static system; the altitude is still coming from the same source. Just a different wire. Hard to see how the calibration could change. And since the 327 displays the pressure altitude, it?s pretty easy to confirm that there?s no gross error. One good thing about digital systems: for the most part they either work fine, or not at all.
 
Standard technique is to ground just one end of the shield, to avoid ground loops.

I agree with Bob except for the above, data line shields should be grounded at both ends, audio lines are pretty much the only exception to ground at both ends.
 
Standard technique is to ground just one end of the shield, to avoid ground loops.
That was my thinking, too, and I'm lucky. At the Dynon, the shield wasn't connected to ground so no ground loops. If I had known what I was doing, the shielded wire from the Dynon was long enough to put a socket on the signal wire and plug it directly into the transponder. Oh well, my splices ain't pretty, but they work. So now I can see my pressure altitude from the Dynon displayed on the GTX-327, and the pressure altitude from the parallel encoder on my Davtron 5-in-1 gauge for a cross check. Another bonus is that the Dynon doesn't have a 10 minute warm-up time like the encoder does.
 
...data line shields should be grounded at both ends...
I really don't want to get into that debate, but if you know a good method for extracting DB-25 pins, I would be very grateful. I was just trying one of those cheap plastic extractors and it just wouldn't go.
 
I really don't want to get into that debate, but if you know a good method for extracting DB-25 pins, I would be very grateful. I was just trying one of those cheap plastic extractors and it just wouldn't go.

No debate, just fact.
The metal extractor tools work great, ditch the plastic ones.
 
The metal extractor tools work great.
What tool do you recommend, please? Something like this? https://www.aircraftspruce.com/catalog/elpages/edmo91285.php?clickkey=37399

I had difficulty installing the transponder back into the tray. The Garmin installation manual says:
2. Looking at the bottom of the transponder, make sure the front lobe of the locking mechanism is in a vertical position. This can be accomplished by using a 3/32 inch Allen wrench through the face plate.
3. Slide the unit into the rack until the front lobe of the unit touches the rack.
4. Turn the Allen wrench clockwise until unit is secured in the rack.
I took "front" to mean the front of the aircraft, and that just didn't work. What Garmin meant is the front of the transponder as you are looking at the front panel. Doh!
 
for me the hardest part of using these tools is the relearning curve. they work when you get the hang of them.then you are done and by the time you need it again, you have forgotten ''the hang of it''
 
Insertion & Extraction / Red & White 24-20 11-04433 $4.95
That's what I tried to use. I was using the silver colored end, not the copper. Oh ****, was I using the wrong end??? I'm on my way to the hangar now and I'll confess if I was.
 
Last edited:
Look at the images in the SteinAir link. The white end is the extractor; the colored end is the insertion tool (pushes directly on the pin, so you don't have to push on the wire).

If it doesn't seem to work on the 1st attempt, rotate the tool 90 degrees and try again.

edit: The ACS descriptions are kinda weird. I've never seen the tools described by wire size anywhere else; they should be spec'd for the size of the *pin* they're used on.

Charlie
 
Last edited:
If it doesn't seem to work on the 1st attempt, rotate the tool 90 degrees and try again.
I tried that and even broke off the two little "leaves" and tried inserting them one at a time. There just wasn't enough room. Need stronger, thinner, metal. Oh, and yes, I was using the "extraction" end.
 
Walt is right. That's the first thing taught in the experimental avionics installation class at AEA. Shielded audio wire shields grounded at one end (preferable the ICS/Audio panel end). And all shielded data wires have their shields grounded on each end to the component unless otherwise stipulated. That's standard industry practice, although there are many instances where only one unshielded wire is run in an RS232 feed without interference. Garmin's diagrams in their manuals use symbols to designate grounding of shielding runs, using triangled "S"s for shield drains. Go by the official manuals at a minimum.

I agree with Bob except for the above, data line shields should be grounded at both ends, audio lines are pretty much the only exception to ground at both ends.
 
Unfortunately, there are no universal truths when talking about shielding. As mentioned, many data lines really don't need shielding at all (even when spec'd by the mfgr), and there are many situations where audio shields *must* be grounded at both ends. The 'tell' with audio is to look for a return. If there's just one (unpaired) signal wire in the shield, then the shield is the return and must be grounded at both ends, unless you want the a/c's chassis to supply the ground (with the small signal audio return riding along with the noise from your strobes, etc).

Charlie
(no Holiday Inns in my recent past, but many years as an audio and industrial tech)
 
Garmin's diagrams in their manuals use symbols to designate grounding of shielding runs, using triangled "S"s for shield drains. Go by the official manuals at a minimum.
Sadly, that isn't the case for the GTX-327 Installation Manual. They just show a standard earth ground symbol for the shield, and no return conductor inside the shield. The manual was unclear. I guess I'll find out if the way I wired the shield grounds is adequate when I start up the engine after these mods.

P.S. Since we're comparing credentials now, I was an instrumentation engineer on the Space Shuttle in a former life and a nuclear instrumentation engineer in another former life. I've never forgotten, though, that EMC is mostly magic. :)
 
Back
Top