Garmin GSA28 and Trim Motor Control

Some background:

You can attach your trim motors to the GSA 28 autopilot servos such that they automatically adjust trim as they do their autopilot work. The servo has input and output pins for trim, so that manual trimming gets "passed through" the unit.

During the beginning of my flight testing, I had this automatic trim motor control function disabled in the G3X software (enabling/disabling it is described in the installation manual, 29.4.13.1&2). This was so that this automatic trim functionality would not interfere during my initial autopilot tuning. When this function is disabled, the autopilot will occasionally ask you to manually adjust the trim. This works fine.

The issue:

I recently decided to enable this automatic trim adjustment function, since I believe I have the AP servos' strengths dialed in and controlling both pitch and roll properly. Ever since then, I have noticed a problem with its command of roll trim. When in HDG mode, it will favor commanding RIGHT trim. It will occasionally trim left, but the vast majority of its trim commands are right. To the point where, if I let it go for a while, it will eventually end up with my roll trim actuator fully-right. Fortunately my RV's roll trim is that spring-bias system, which can be easily overpowered even at full trim deflection. Still, I question this behavior.

This does not happen when the AP is in "NAV" mode, following a track. Only when it is in "HDG" mode holding a particular heading. Pitch trim control is always fine. It trims up when it needs to and down when it needs to, and its behavior never results in full pitch trim deflection--only roll trim.

I've found that the problem is most often reproducible when I'm flying slowly (<90KIAS). Other than that, I have not done too much diagnosis to isolate what flight conditions cause it to happen.

Has anyone seen this kind of behavior and managed to correct it? Ray Allen roll trim servo, if that matters.

Roll Trim GSA 28

I have a very similar problem with the A/P roll trim on the RV 14. Have opened a query with G3x experts. Forwarded the flight test raw data and await a response with hopefully a suggested resolution.

You can tell pretty quickly from the servo torque column in the data log whether the servo is trying to hold a sustained torque for long periods of time. I've seen this before when the aircraft is flown uncoordinated.

Draker said:

lot will occasionally ask you to manually adjust the trim. This works fine.

The issue:

I recently decided to enable this automatic trim adjustment function, since I believe I have the AP servos' strengths dialed in and controlling both pitch and roll properly. Ever since then, I have noticed a problem with its command of roll trim. When in HDG mode, it will favor commanding RIGHT trim. It will occasionally trim left, but the vast majority of its trim commands are right. To the point where, if I let it go for a while, it will eventually end up with my roll trim actuator fully-right. Fortunately my RV's roll trim is that spring-bias system, which can be easily overpowered even at full trim deflection. Still, I question this behavior.

This does not happen when the AP is in "NAV" mode, following a track. Only when it is in "HDG" mode holding a particular heading. Pitch trim control is always fine. It trims up when it needs to and down when it needs to, and its behavior never results in full pitch trim deflection--only roll trim.

I've found that the problem is most often reproducible when I'm flying slowly (<90KIAS). Other than that, I have not done too much diagnosis to isolate what flight conditions cause it to happen.

Has anyone seen this kind of behavior and managed to correct it? Ray Allen roll trim servo, if that matters.

Click to expand...

Are you sure the trim polarity is correct? It sounds like it is reversed for roll. Without auto trim, the 28 just passes voltage along at whatever polarity it got from the switch. BUT, with autotrim, the servo produces the voltage and does so with the polarity set up in the configuration. If reversed, it keeps trying in vain to add more trim to reduce servo torque, but keeps getting worse. It is not human, so cannot resolve that discrepancy and just keeps trimming more, assuming that that actual human set up the polarity correctly. Because the springs are much weaker than the servo, the plane still flies ok with the full trim.

Right rudder!

Famous words from you first flight instructor. RIGHT RUDDER. I often see the autopilot on many aircraft trying to “compensate “ with the only thing they can-ROLL TRIM if the ball gets just a tad out of center. This happens most in climbs and slow flight conditions.
Not saying this is your issue, but it sounds like what I see often. The GMU11 is sensing a slip and correcting with the only thing it can.

mburch said:

You can tell pretty quickly from the servo torque column in the data log whether the servo is trying to hold a sustained torque for long periods of time. I've seen this before when the aircraft is flown uncoordinated.

Click to expand...

toolmanmike said:

Famous words from you first flight instructor. RIGHT RUDDER. I often see the autopilot on many aircraft trying to “compensate “ with the only thing they can-ROLL TRIM if the ball gets just a tad out of center. This happens most in climbs and slow flight conditions.
Not saying this is your issue, but it sounds like what I see often. The GMU11 is sensing a slip and correcting with the only thing it can.

Click to expand...

Yea, this seems like the most likely cause now that I think about my flying!

What happens if you disconnect AP when it moved the trim all the way to the right? Is the plane fairly trimmed or out of trim?

Another possibility, to other suggestions, if you have a bit of wing heavy which might be more pronounced during slower flight.

Autotrim

There are 3 columns in the Flight Data Log you can cross reference to explore this a little more:

• Lateral Acceleration (G)
• AP Roll Torque (%)
• AP Roll Trim Motor

If the aircraft is uncoordinated, you will see relatively greater values in the lateral acceleration column. During periods of uncoordinated flight, the roll servo will compensate by applying greater autopilot servo torque. It takes 5 seconds of sustained high autopilot servo torque to trigger trim tab movement. Between those 3 columns you should be able to roughly determine if that is your problem or not.

Thanks,

Justin

I re-checked my configuration, and the Roll Trim Motor Direction setting was "Reversed". I believe the way I have mine wired, this should be Normal. I'm going to give it a try using "Normal" and see if this corrects the problem.