Taming Runaway Trim

[rocketman1988]

"...The G3X does everything you need for the trim (and flap) so use that (it is also much better developed than the VPX). I would power the whole G3X system from c/bs that are independent of the VPX (but I understand that will not be a view held by all)..."

Posting a statement like that, you would have to expect the next question...

Where is the data that supports your statement?

[pvalovich]

-8A with TruTrak ADI II.

During build I added three switches to the right side switch panel: Autopilot Disconnect, Elevator Trim Disconnect and Aileron Disconnect.

Have never had to use trim disconnects but do sometimes turn off autopilot with disconnect switch vice display switches.

During Phase 1 I landed with full nose up and full nose down trim. Easily manageable with the caveat of pre-planned conditions. Would probably take a few seconds to figure things out if runaway occurred during approach.

Bottom line - fly the airplane.

[gereed75]

I do not know anything about the trim features built int any of the more sophisticated panel integrated auto pilots out there. I did do some fairly extensive testing of the ?stock? trim system in my -6 equipped with the typical Mac servo.

It is my opinnion that this system provides way more trim power than necessary. It does not surprise me that Those who have tested in relatively slow flight (landing configuration) have found surmountable control forces. But Try it sometime in cruise flight and I think that you will find that control forces generated by the stock trim very quickly build to dangerous levels

For those reasons I have taken some measures to mitigate the risk - first I have installed a speed and time limiting trim controller. Then I have biased the servo travel to provide only just enough nose down trim to balance the elevator at max cruise speed. My logic is that a run away up trim is easier to deal with than a hard trim down condition (runaway up - reduce power, speed slows, trim effectiveness and resulting force go down until it is manageable)

I know this is not something most on this board want to think about, but I urge you to think through the various scenarios carefully and and come up with a risk mitigation approach that makes sense to you.

Off the soapbox. Fly safe

[jliltd]

So there he was. Excited to have his new G3X Touch glass panel upgrade from steam. Full GSA 28/GMC 307 autopilot with auto trim. Lined up the RV-8 on the centerline for first flight with me in the rear position. The shop installer recommended he turn on the autopilot as soon as possible once off the ground. I told the pilot that was a super bad idea and we would wait until at least 3000 AGL before touching the autopilot.

Takeoff was fine although the pilot was a little slow on the glass panel display having been so used to legacy steam gauges. He initially kept asking me how fast and high we were and I would crane my neck to look at his panel and issue call outs. He caught on fairly quickly.

Once at altitude I had him push the straight and level blue button to initially try the autopilot. Things were smooth but the pitch had a mind of it's own. I had him disengage and we had close to full aft trim and would have looped out if we hadn't both got on the stick pushing. He got on the trim button and got no help. The trim button didn't seem to work. Much fiddling and trouble shooting didn't improve things. We decided to land ASAP. It took two of us pushing forward on the stick to bring it back in for the landing. He did a fantastic job with me holding lots of forward pressure from the rear.

So what went wrong, beside a loose and sloppy pre-flight briefing? The installation shop had the pitch servo wired in reverse so as the plane climbed or decended the autotrim would run the wrong way and increase elevator force, up or down, rather than trimming it out. This meant when we disengaged we were in a terrible out of trim position. Our futile attempts to trim were also wrong way. **** poor preflight control check on our part. The trim direction is an easy fix through the G3X installation setup screen -- on the ground.

So here's the subject tie in to the trim time limiter. The pilot was so panicked he would hold the trim button continuously to fight the pressures and the system would time out thinking his death grip on the trim control was a runaway trim. If he had instead pulsed the trim button the trim timer would have reset each time and it would have worked fine (albeit reversed). But neither of us knew this at the time. Then again in this reverse hookup scenario we were holding the trim button the wrong way anyhow. So it could be debated whether the trim timer hindered or helped us. There is always a tuition involved when learning things. And this lesson cost us a trip to WalMart for some new whitey tighties.

[vlittle]

Several years ago, I developed an autotrim system using many of the techniques described above for fault mitigation in manual mode. My system predates most of the commercially available autotrim controllers.

I, too was also concerned about failures in autotrim mode due to EFIS system failures or unknown unknowns.... i.e. MCASish failures. The disable modes for my runaway autotrim are:

Any one of the following:
-Press any trim button to turn off (for 30 seconds)
-Disconnect the autopilot
-Pull the trim breaker
-Pull the autopilot breaker
-Turn off the EFIS
-Push the big Autotrim button

The idea here is that in an emergency, when the pilot's IQ drops, any logical course of action will disable or turn off the auto trim.

On a side note, I dislike trim switch timeout in normal mode. That is not an intuitive way to correct faults (as discussed above). I prefer to lockout a stuck trim switch by detecting when the opposite trim switch is depressed. The recovery method for this event, if it can't be rectified manually, is to engage the autopilot, which engages auto trim. The autotrim will then neutralize the trim settings.

Since my autotrim system also controls my flaps (for go-around and landing presets) I have a bunch of other failure modes to consider as well.

I stopped selling my system years ago because of liability concerns.... which is an entirely different discussion.

V

[rv8ch]

Seems like the trim runaway problem could be solved by not having a trim system that activates based on ground, but on 12v. Of course, taking care when wiring will help, but it seems like many people have experienced a trim runaway problem.

I've got a trim power switch on the panel, next to the autopilot servo power and the flaps power switches - just in case, since I don't have CBs but fuses.

[penguin]

Originally Posted by rocketman1988

"...The G3X does everything you need for the trim (and flap) so use that (it is also much better developed than the VPX). I would power the whole G3X system from c/bs that are independent of the VPX (but I understand that will not be a view held by all)..."

Posting a statement like that, you would have to expect the next question...

Where is the data that supports your statement?

How many development $$ have been spent on the VPX in the last 5 years and how much on the G3X series? What testing has been documented of the features advertised for the VPX, particularly the EFIS/VPX integration, as the EFIS and VPX software is developed? When was the last VPX software update released? When was the last G3X (Touch) software update released? You don't have to look very far to answer your question ...

[rocketman1988]

Sorry, that's not data...it's your OPINION.

Let me state that I have the entire G3X suite.

That said, comparing the Garmin system and the VPX isn't really a valid comparison. The primary purpose of the VPX, since it's inception, has been to provide an electronic circuit breaker system, something that Garmin does not do.

There is some overlap between the two systems as far as flap control, wig wag lights, specifically related to the GAD27. This unit can mix the the trim inputs eliminating the relay decks, control two wig wag lights, control interior PWM lights, and provides for advanced flap control options. It also provides several more discrete inputs.

The VPX has basically the same functionality as the GAD27...as an aside. It's primary function is providing electronic circuit breaker functions and user addressable input/output...which the Garmin does not.

If you read the history, the VPX was introduced around 2010 and was subsequently acquired by Astronics, which has been around since 1968.

Your point about $$$ is not really valid, as Garmin has a multitude of different products even within the G3X line up...of course they are going to be spending more money.

Updates mean NOTHING. Garmin is continuously updating, optimizing features, and adding features and functionality...BUT, if we are comparing apples to apples, when has Garmin made any significant changes to the GAD27?

Listen, everyone here has their own opinions. I chose to use BOTH systems as they each provide a different set of features...and because of the overlap, I opted to not use the GAD27, which was also recommended by my avionics shop, and saved $600 in the process without sacrificing any features...

Build (or fly) On and have a Happy New Year!

[sailvi767]

Since this is a discussion about trim safety I thought I would copy the Dynon HDX features. It provides what I feel is a reasonable level of protection.

? When the trim controller is connected to two sets of switches or stick controls, one of them will always take priority over the other when the requested trim commands disagree. This is normally the "pilot" position, but whether or not this is the left or right seat or set of controls will vary by aircraft.

? Although it is contained within the SV-AP-PANEL, the electronics in the trim controller are powered by aircraft power and are independent from SkyView. Trim control will remain available even if SkyView itself is unpowered or offline. In this failsafe condition, the trim controller's speed scheduling is disabled, and the trim motors will run at their full speed when trim is commanded.

? To protect from a "runaway" trim condition, the longest that a trim motor will run continuously even when a trim button or switch is continually depressed is five seconds. After three seconds, the trim motor will cease responding until the trim button is released and pushed again.