Crazy full-up elevator issue...

[RV7A Flyer]

Originally Posted by gasman

Then a safe way to wire the trim is to have it active only when the auto pilot is disengaged..... unless you HAVE auto trim.

You're just moving the failure point around. You'd have to have a relay in the trim circuit, activated by power to the autopilot servo, and there are ways for *that* to fail, either open or closed. Not to mention that failure of the relay might cause a failure of the autopilot itself (which could be particularly bad if, say, in IMC).

FMEAs are not as simple as just adding more complexity to the system.

[markserbu]

Originally Posted by RV8JD

Yes, Dynon does offer auto trim. As I said in my post, you need the separate autopilot control panel to get auto trim. You can add it to your existing SkyView system.

Link --> https://dynonavionics.com/autopilot-control-panel.php

Attachment 31309

"The Autopilot control panel is also a full featured dual-axis trim controller. It can control your electric pitch and/or roll trim motors, replacing the relay decks normally used.

With Autopilot Auto-Trim, the SV-AP-PANEL can keep the the aircraft in trim while under autopilot control. This eliminates the need for the pilot to manually actuate their trim to follow the Autopilot’s built-in trim advice. "

Ah, gotcha. Thanks. Now if I could just *get* a Skyview HDX...been trying to buy one since November and they've been out of stock!

[rzbill]

Mark,
I read the thread but did not see the following asked.
Are you engaging the pitch autopilot during the climb phase after takeoff and asking the autopilot to level the plane at a set altitude, with zero trim adjustment after AP engagement in the climb?

If so, this is the problem. It would correlate with time and distance of repetitive flights. My RV needs a major nose down trim change from climb to cruise with the largest fraction due to speed increase rather than attitude. I have had the AP kick out once or twice due to bad pilotage of not enough trim adjustment for proper AP operation.

I assume you have hand flown the suspect flight path with no autopilot but electrically connected trim with no issues?

[markserbu]

Originally Posted by rzbill

Mark,
I read the thread but did not see the following asked.
Are you engaging the pitch autopilot during the climb phase after takeoff and asking the autopilot to level the plane at a set altitude, with zero trim adjustment after AP engagement in the climb?

If so, this is the problem. It would correlate with time and distance of repetitive flights. My RV needs a major nose down trim change from climb to cruise with the largest fraction due to speed increase rather than attitude. I have had the AP kick out once or twice due to bad pilotage of not enough trim adjustment for proper AP operation.

I assume you have hand flown the suspect flight path with no autopilot but electrically connected trim with no issues?

Yeah, I know what you're saying. Yes, I engage the autopilot on climb-out as I always do, setting an altitude and rate of climb. The plane is trimmed before engaging the autopilot, and if I'm not properly trimmed or it changes during the climb, the autopilot is very "vocal" about setting it correctly. As I've said it shows either "UP" or "DOWN" on the display along with an associated arrow, and (what I've forgotten to mention before) it "bucks" several times. It's impossible to miss this bucking...it really gets your attention! And it's never bucked before this runaway trim issue, so now I'm not so sure about the "autopilot kicking out" theory now.

[gasman]

Originally Posted by RV7A Flyer

You're just moving the failure point around. You'd have to have a relay in the trim circuit, activated by power to the autopilot servo, and there are ways for *that* to fail, either open or closed. Not to mention that failure of the relay might cause a failure of the autopilot itself (which could be particularly bad if, say, in IMC).

FMEAs are not as simple as just adding more complexity to the system.

Are you over thinking this, or is it really that complicated. If you have one wire that disconnects the AP, and one wire that removes power to the trim, then one simple toggle switch on the panel would give you one or the other and you would never have a runaway trim that was hidden by the AP.

Just the fact that you chose electric flaps and electric trim also added failure points to your aircraft.

[RV7A Flyer]

Originally Posted by gasman

Are you over thinking this, or is it really that complicated. If you have one wire that disconnects the AP, and one wire that removes power to the trim, then one simple toggle switch on the panel would give you one or the other and you would never have a runaway trim that was hidden by the AP.

Just the fact that you chose electric flaps and electric trim also added failure points to your aircraft.

ETA: I don't think I'm overthinking it by positing a relay. You don't want to simply switch power between AP and trim, because pulling power from the AP to use trim means (at least for Dynon) that the AP servos are no longer visible to the network and the EFIS, and I don't believe that reapplying power to them will make them operable again without rebooting the EFIS. And you D/C the A/P from the stick via shorting a line to ground, but it's hard to see how that would help if you just want a toggle switch.

And that toggle switch can fail. Maybe it fails so that you can't use the trim at all. Or it fails such that it fails to lock out the trim when the AP is engaged. Or it fails so that neither one can be used. Etc.

This is how "normal failures" occur...people start adding components that they think increases safety, but in fact they *decrease* it. Safety Hazard Analysis can be tricky.

FWIW, I've not had (nor heard of from any of the half dozen fellow RV pilots I fly with locally) a runaway trim on an electric RAC trim servo. I *have* had a mechanical system fail (Piper Cherokee) in the past.

[RV7A Flyer]

Originally Posted by markserbu

Yeah, I know what you're saying. Yes, I engage the autopilot on climb-out as I always do, setting an altitude and rate of climb. The plane is trimmed before engaging the autopilot, and if I'm not properly trimmed or it changes during the climb, the autopilot is very "vocal" about setting it correctly. As I've said it shows either "UP" or "DOWN" on the display along with an associated arrow, and (what I've forgotten to mention before) it "bucks" several times. It's impossible to miss this bucking...it really gets your attention! And it's never bucked before this runaway trim issue, so now I'm not so sure about the "autopilot kicking out" theory now.

This is like pulling teeth...

Again, CAN YOU POST THE SCHEMATICS. And do you have any sort of datalog for these flights?

And most importantly, what is the *exact* configuration of the vehicle before and during these events? A/P on *in what mode*, when was the trim adjusted, what was the rate of climb or airspeed selected, etc., etc.

Here we are several pages in and you're just now telling us about some "bucking" of the system.

Have you tuned the autopilot parameters according to the autopilot manual?

[RV7A Flyer]

Any update?

[RV7A Flyer]

Guess not. Drive-by posting with no further info.