New gremlin and I'm stumped! No trim??

[lancair360]

Spoke with Dynon today. We're pretty much settled on it being some issue in a screen / network loader. Like I said my second screen is an old serial number and old hardware version (1). I may as well upgrade to the HDX now. Dynon is talking options, waiting to hear back, but we have the end in sight.

Also, I've had zero failures in the last 24 hours. And we haven't even done anything yet. Very frustrating but at least the airplane can fly.

 

[lancair360]

100% failure rate today regardless of how I boot the system up. Ball is in Dynon’s court, waiting to hear back.

 

[lancair360]

Intermittent times to fail now but still 100% failure rate. Cannot correlate it to anything. Even removing the suspect screen from the whole equation and it still failed.

Got my options from Dynon today. I chose option 3. Out of pocket full retail for two new HDX screens. Then I’ll send mine back for repair and be able to sell them on eBay and get some money back.

Hopefully we’ll find out what’s wrong. And hopefully the new screens will fix the problem! At this point I have no idea except to try.

 

[lancair360]

Two new HDX screens installed today. Coping config files off the old screens meant setup was easy. Updated software indiviually. Did network detect. Added tach and hobbs times to new screen. Configured bottom band for engine gauges. Done.

And NO failures. No AP panel offline and trim failure. Old screens on their way back to Dynon HQ, cannot wait to hear what the hell is wrong with them. Then they're up for sale. So if anyone needs two freshly gone over screens by Dynon, 7" Skyview, hit me up.

So happy that after two months and over a hundred emails with Dynon we can put this to bed. I'd rather not be out $7000 in new screens but it is what it is. As usual working with Dynon is fantastic and why I prefer to play in their sandbox than brand G.

Whenever I learn what was wrong I will update this thread.

 

[Rhino]

Thanks for letting us know the resolution.

I guess this means you can't do the impossible anymore though.

 

[lancair360]

Spoke too soon. Failed after takeoff today. Same failure mode. We've ruled out the 12v trim section of the equation. Have two new HDX screens, new AP panel. and something in the system is still seeing a runaway pitch trim. Also something causing power to the trims to be shut off. and given the failure mode those two things are NOT related according to Dynon. And no, trim isn't running, wiring all rung out, phone call to Ray Allen folks to verify info.

Possible fix - removed the trim wiring from the Dynon completely and just hook up normal trim relays. Keep the elevator position indicator on the EMS though.

 

[vlittle]

I recall that if you connect the two SV screen Contact 1 inputs together, some wonky things happen with the autopilot. It's a WAG guess here, but easy to try, by disconnecting one of them.

Vern

 

[lancair360]

What’s contact 1 input?

150 emails back and forth with Dynon including sketches and photos and video.

 

[Marc_J._Zeitlin]

Spoke too soon. Failed after takeoff today. Same failure mode. We've ruled out the 12v trim section of the equation. Have two new HDX screens, new AP panel. and something in the system is still seeing a runaway pitch trim...

I haven't been paying extremely close attention to all the failures you've had here, and you've stated that you've now replaced the screens and the A/P controller, and have rung out all the wiring.

What I don't recall hearing about is the A/P servos. They're smart and have electronics in them that communicate on the bus. MAYBE one (or both) of them are brain damaged and are telling wrong info to the rest of the system, causing the faults.

I'd suggest replacing one servo at a time with a loaner and seeing what happens (if you haven't already done this).

Also, what Dynon SHOULD (but almost certainly won't, as it's expensive) do is send a tech with a network sniffer to watch what signals are coming from where on your bus right before the system Sh!ts the bed.

 

[lancair360]

No servo errors have come up. And taking the servos offline doesn’t change any behavior. Autopilot is not failing, just the panel. Dynon cannot figure out how anything else on the network would cause these failures either. None of it tracks. And it’s not repeatable now. It was for a while, dependably failing. Now it’s intermittent again with the new screens installed.

For some reason the computer is seeing a runaway pitch trim. But the trim isn’t moving. But that would cause a “trim switch stuck” message. Not an AP panel fail message. The problems happen together but aren’t related according to Dynon. The AP panel fail is the AP panel losing power at pin 9 and leaving the network. But no one knows why. And the runaway pitch trim the software guys is seeing cannot be correlated to anything. It’s all weird.

Common denominator is the AP panel. But it’s brand new form Dynon. I think we should rule it back in. Also they haven’t opened up the AP panel I sent back to them yet, I think there will be answers in there.

Software guys are digging through code, see if we learn anything next week.

In the mean time I’m going to remove the trim wiring from the AP panel and wire in regular trim relays. Airplane will be flyable and trims will always work then. Also buys some breathing room for me and Dynon. I’m going to build up a wiring harness and connect the two trim relays to a D15 connector so I can I unplug trims from the Dynon and plug into my harness. And quickly swap back for troubleshooting with Dynon.

 

[CanardMulti]

Just thinking way, WAY outside the box here: Are there any high-ish current wire bundles running in proximity to other sensitive bundles or components where a current could be induced that could spoof electronics into thinking a servo is running away? I just have a gut feeling that when the problem is indeed isolated, it is going to be something from way out of left field.

When you finally isolate the problem(s), perhaps some billable hours by you as a consultant might be in order? (Yea, I know. But hey, it IS April 1st after all ).

Ken

 

[lancair360]

Worked perfectly from 2015 to 2021. Got intermittent AP panel offline with failure of all trims. New AP panel fixed the problem, old one had every solder joint on the main board broken.

2 years later, same problem. Only this time a new AP panel didn’t fix the problem.

There have been no changes, additions or deletions of any wiring or components.

I’ve asked Dynon where the computer gets it’s from trim moving information from. If it’s from the relays on the AP panel then that’s still suspect even though it’s brand new. Trying to get them to focus some efforts on the AP panel again.

Inducing noise to cause a runaway trim indication in the computer wouldn’t cause an AP panel offline. It would trigger a trim switch stuck error, which isn’t happening. Also what could be intermittently causing EMI that only effects that one very specific thing?

Agreed it’s either gonna be way outta left field or something stupid simple. But we’ve thrown everything at it so far and the fact Dynon is baffled is saying something. Network sniffer would be cool if we could get it to fail consistently again.

 

[lancair360]

Installed Ray Allen trim relays. Flew for an hour tonight, no squawks.

 

[jakej]

It’s not been mentioned before ? But what electric trim motors do you have fitted ?

Rhino’s post #23 asked if the Dynon Trim Amp unit is fitted - answer ?

 

[djones]

Jake,
No trim amp installed, it's a very odd case. I am curious to see if the a/p panel remains online after removal of the trim system wiring from the panel. That will be a smoking gun there is something in the trim wiring that's causing the panel to go offline.

 

[CanardMulti]

Just rereading the history. Another admittedly even farther outside the box thought:

1. Worked fine for a half dozen years.
2. When it began failing, EVERY solder joint in AP panel was failed. Note that it did not fail when just SOME of the joints had failed.
3. While certainly not impossible that Dynon had poor QC on just one panel board's solder joints, a perhaps more likely explanation might be some sort of vibration node in vicinity of the AP panel.
4. Vibration causes wear. Just the right-wrong vibration might cause wear in the insulation of wires, inside or outside of a cable or harness, leading to a short of some kind.
5. First new panel fixed the problem, perhaps due to some wires or cables getting moved a small amount and moving poorly insulated spots away from each other just enough in the process.
6. First new panel continues to vibrate, wear spots on wires get larger, problem reoccurs.
7. Second new panel now does not fix the problem, but IF said theoretical wear spots were now two years larger, they might not have been moved enough to move shorting wire(s) away from each other.

I'm in no way saying this actually is the problem. It is instead simply a series of linked conjectures that fit the physical facts of the situation.

Good luck.

Ken

 

[lancair360]

Trim motors are Ray Allen. Nothing has changed there. Flew a couple times today, no squawks. Diagnostics sent to Dynon to see if the computer is still showing a runaway pitch trim now that trim is removed.

Wire harnesses have been rung out and the two Dynon network cables connected to the back of the AP panel are brand new. No shorts in any wiring. Also failures were intermittent, then constant, all while sitting in the hangar. Then intermittent again. It behaved for a while like a component on a circuit board going bad. On one evening of troubleshooting it failed after 7 minutes of power up. Then 2. Then 1. Then immediately. All sitting in a heated hangar totally undisturbed. Testing one fix it worked for a day, then leaving the fuel pumps it failed. Power cycle fixed it the. It didn’t fail again for over an hour of flying.

Spoke with the Ray Allen people and the trim servos are just a dumb 12v system. If the computer thinks the trim is moving and it’s not, it isn’t the trim system. Any shorts in the trim wiring or current being induced or anything like that would show up as trim actually moving, which is isn’t.

While it was up and running on the old screens and not failed I tapped and twisted all the wires and components and could not induce a failure.

I don’t know if they’ve opened up the AP panel I sent back a while ago, if solder joints are again broken that’s a clue. Really curious about this replacement AP panel though. Also, like mentioned, with no trim connected to the Dynon now let’s see if we get any kind of failure.

So far, no failures. It’s a dill of a pickle. Really thought we had it down to some computer/software/network loading problem. Every time we nail down something correlated to the failure it changes behavior. It’s fascinating.

Keep the theories and ideas coming. We haven’t thought of everything yet but honestly we’ve covered about everything in this universe.

 

[jakej]

Ok, I’ll add some other points into the mix -
If you are using the ‘white’ covered Dynon network cable then that may? be putting a strain on the a/pilot panel board connectors resulting in fractures of the solder.
What what type of fuse or circuit breaker are you using?
Did you bypass it temporarily with another one?
Are there any high current load wires close to your a/p panel wiring ?
This is how I do all my installs and, other that having issues with the first batch of a/p panels, all subsequent installs have been solid.
(I never use the 24/26 awg wiring as it’s to ‘weak’ IMO)
2 core shielded (braided) 22awg wires from Ray Allen Trim motors to a/p panel.
1 x 4 core shielded 22awg from a/p panel to control sticks for roll & pitch control.

Your experience reminds me of an ‘issue’ around 2005 I had with an Alternator field circuit going offline right from the get go, sometimes it would when using the PTT or at other random times. ( I had wired 2 x RV7’s, side by side, with the same firewall forward wiring method & routing)
After much research, including contacting a couple of ‘gurus’ there did not seem to be an answer until months later I just decided
to replace the 18awg field wire with a shielded one (it was only 3.5 ft long) & the problem has never reoccurred since, go figure.

ps - the other RV never had the same problem & I’ve never had to or use shielded field wire since then.

I do understand your pain however there will be a solution, talking, analysing & more troubleshooting will get you there - sometimes it can be a pita .

 

[lancair360]

Yes, using white Dynon network cables. When first installing the new AP panel during this round of troubleshooting the failure occurred on the first boot, this was with things plugged it but not mounted and secured so there wasn’t any stress on any physical connections.

The behavior of the failure will sometimes change with no changes in any other parameters. I went through dozens and dozens of power cycles while troubleshooting and diagnosing, 99% of which were in the hangar and not even sitting in the plane, so no movement whatsoever to the plane. Then all of a sudden it would work and nothing you did could make it fail. This plus other clues is what led us away from a physical issue to a code issue. But we can’t be certain that’s even right.

No high voltage in the whole aircraft.

2A breaker for the trim power. Never bypassed it because if there was a 12v issue on the trim side we would be seeing it in the trim itself.

It’s a lot to dig through but when we got repeatable failures it only failed after the slave screen booted. Pulling the CB for the slave screen and the failure occurred in the exact moment the slave screen would have booted up, exactly.

Removing that screen from the network and then completely (D9 first then also D37) still resulted in the failure.

For a minute Dynon thought an overloaded network may be the cause so I booted with two devices on the network unpowered and it didn’t fail. Then booted with one device connected and the failure would occur in a couple minutes. With everything online it failed immediately. This test was repeated with every device on the network unpowered at some point and all results repeatable 100%. Came in the next day and with two devices off the network booted up and it failed immediately. This is the type of changes in behavior we’ve been dealing with. We could then get no failures on several boots and a half hour on ground power or battery alone just sitting there, then it would fail on the next boot and every one after.

I keep circling back to this new AP panel. The trim relays inside there tell the computers what’s going on and if we remove the trim switches from the equation things seems fine. But with intermittent failures you cannot be sure. We left the AP panel D9 connectors plugged in and jumped power and ground from the D15 trim wiring to the AP panel and didn’t get anything failures.

With the AP panel not on the network I would still get a trim failure. With the AP panel off the network, the trim CB pulled, I could boot up the system and count to ten and pushing in the trim CB and the trims were dead. No AP panel fail cuz it’s off the network already. But if this is a runaway trim thing that takes 4-5 seconds to happen and there was no AP panel and no trim power and instant failure of trims, they were already turned off by the system when I pushed in the breaker.

And the trims would work until the slave screen booted. Never failed before that. That’s a computer problem, not a wire problem.

Flown the plane 3 hours since removing the trim wiring from the AP panel and installing Ray Allen relays and still no failures. Diagnostics sent to see if the computer is still trying to shut off trims. Trims working 100% now because it’s just a dumb 12v system. If there was a power or ground and wire issue or high voltage EMI or RMI we would still see it in the trims. All points to a computer issue still.

And just for fun remember the AP panel offline master caution and a stuck trim switch are two different failures and not related. But here are 100% correlated. It’s baffling. A stick trim and runaway results in a master caution trim switch stuck message and we’ve never received that caution. AP panel offline only means it’s not connected to the network which is a power or ground being turned off. It’s not a network data problem or you would get a data one or data two check pins x and y message.

 

[jakej]

To me it appears to be only happening to 1 aircraft & therefore at least, probably, not a manufacturing fault or software Issue or we’d be hearing of other cases - unlike the issue with some earlier a/pilot panels.
Without have to troll through all the ’posts’ - did you try another a/p panel ?
Do you have several troublefree hours now with the trim/s relay bypass ?
I made up a trim/s motor bypass wiring harness using the existing 15pin connector with the switch mounted to a plate can be attached to the a/p panel hole- wondered if u did the same ?