Intermittent glideslope needle

[jmaynard]

My Zodiac has a GNS430W and a Garmin GI106 indicator, connected via the Approach Pro-G wiring system. Recently, the glideslope needle hs gotten intermittent: it stays dead center, even right as the 430 is powered up and on the self-test. I can enter maintenance mode and show the problem.

Three times now, I've gone in and disconnected the plug from the back of the GI106 to try to troubleshoot the problem. Each time, it's started working just as soon as I've plugged it back in. The DB-25 in the cable was originally not fully mated with the one on the indicator when the slide lock was locked; the second time, I put in a couple of little washers to make it mate more fully, and thought that had fixed the problem. It didn't.

Yesterday, I went to the airport, meter in hand, to measure the voltage at the GI106's connector, just to see if I could see the problem. You guessed it: there was voltage at the appropriate pins of the connector when I measured, and plugging it back in showed it working again.

This is about to drive me nuts. How am I supposed to show folks what a GPS LPV approach looks like at Gaston's if the glideslope needle doesn't work?

Any suggestions? I'd like to isolate the problem before I take it to my local Garmin dealer for a warranty repair...

 

[ScottM]

My Zodiac has a GNS430W and a Garmin GI106 indicator, connected via the Approach Pro-G wiring system. Recently, the glideslope needle hs gotten intermittent: it stays dead center, even right as the 430 is powered up and on the self-test. I can enter maintenance mode and show the problem.

Three times now, I've gone in and disconnected the plug from the back of the GI106 to try to troubleshoot the problem. Each time, it's started working just as soon as I've plugged it back in. The DB-25 in the cable was originally not fully mated with the one on the indicator when the slide lock was locked; the second time, I put in a couple of little washers to make it mate more fully, and thought that had fixed the problem. It didn't.

Hmm up to this point I was going to suggest a bad crimp on the db25. But...

Yesterday, I went to the airport, meter in hand, to measure the voltage at the GI106's connector, just to see if I could see the problem. You guessed it: there was voltage at the appropriate pins of the connector when I measured, and plugging it back in showed it working again.

This was on the connector that would be attached to the GI106?? Or were you actually on the GI106 side of the connector? I suspect you mean the connector that attaches to the GI106. If that is the case, did you try moves the cable to see if the voltage drops out on the pin? This would confirm a bad connection in that connector. If that is good you could also try the same thing on the other connector that is hooked up to the 430. This would be to check again for a bad connection in a connector but to also rule out the interface cable.

The next step is to suspect the Gi106 itself. If possible could you install another one and test? If that one works fine all the time then you have a bad GI106. If it makes no difference I would then suspect the 430 itself.

 

[gismo]

I think a bad crimp is the most likely cause but there could be a problem inside either the indicator or the 430W in the form of an intermittent connection. I'd also make absolutely certain that the appropriate pins (there are two pins involved IIRC) are properly latched into the connector(s) and not sliding back as the connectors mate. The most common cause of this kind of issue is a crimp on the insulation with the bare wire just touching the pin without being crimped tight. One fairly intrusive way to diagnose this problem would be to splice in a pigtail to the appropriate wires in the harness and test for voltage during a failure without disturbing the connectors. A less intrusive alternative would be to install a "patch cable" between the existing cable and the indicator with the appropriate wires tapped.

With an intermittent like this you need to use different troubleshooting techniques than what you'd do with a solid failure. There are four approaches to intermittent fault finding: Swapping, Substitution, enhanced precipitation, and trapping. Swapping only works if you have two or more identical units in service. To isolate the problem you swap things one pair at a time, waiting for a detected failure before making another swap. When the fault begins to occur in the previously working subsystem you can generally assume that it's located in the unit you most recently swapped into that subsystem. Substitution is often the easiest course if you happen to have handy substitutes (e.g. another 430W and indicator from somewhere else) but kinda tough without them. Precipitation involves attempting to find a way to make it fail at will or at least reliably and very often, at which point you can use standard isolating techniques. Trapping means setting up some way to catch the failure when it occurs in a way that provides information about the root cause. The tapping of wires fits this category.

Good luck.

 

[SCCutler]

Breakout-box.

 

[jmaynard]

Breakout-box.

Yeah,. Lance's message has me thinking in that direction. Since it's a standard DB-25 connection at the back of the GI-106, I can hook that up and wait for it to fail. I've already done the wiggle bit, with no success, and I don't think it's a bad crimp - but I could easily be wrong there, too.

 

[jmaynard]

I talked to the folks at Approach Fast Stack Systems today. They had some suggestions for me, involving measuring and wiggling. They also offered to send me a replacement cable for the GI-106, free, just to see if that was the problem.

I've now spent 4 hours inside the panel over the last few days. I can't make the ^&&$%^& thing fail. This is beginning to seriously irritate me.

I didn't hang a breakout box on it yet, but that's next - however, I'm going to wait until it fails again.

 

[ScottM]

Hmmm. Could this be a problem with the glideslope antenna and/or coax? Does the problem manifest itself on LPV or VNAV approaches? If so we can throw that idea out completely. If it is only on ILS....well, maybe we should do a little testing in the RF feed area.

 

[jmaynard]

Hmmm. Could this be a problem with the glideslope antenna and/or coax? Does the problem manifest itself on LPV or VNAV approaches? If so we can throw that idea out completely.

Yup, and it even manifests itself during the 430's power on self test if the problem is happening.