AP servo chattering

[sparxfly]

I have a problem with a roll servo fitted to a Glasair 3.
The issue shows up on the ground using the AP servo checks- it will not settle at a selected position, but chatters around that point unless 'helped' by a little bit of stick pressure.
In flight on a headling change/roll command it frequently gives up and AP disconnects- particularly if the roll is to the right (???).
Stick forces on the ground and in flight are light to moderate- not noticeably heavy at all.

The Glasair 3 has the roll servo out on the left wing, with long linkages across via the stick to the right aileron.
It appears to be a mechanical resonance due to the stepping motor impulses being reflected back through these linkages.
The ailerons don't show any significant(?) play against either the stick or the servo.
Reducing servo torque dosn't make any noticeable difference until it runs out of sufficient torque.
I get the feeling the system needs a small dashpot to smooth the individual steps of the servo to a more gentle linear slew.

I will next try reducing the step rate for a slower overall slew (but still with the individual step impulses) and see if that helps.
But I would appreciate any suggestions from others who may have encountered this problem.

[John H]

Had a similar issue with a Questair Venture on both roll and pitch.The Venture uses control bias spring cartridges to tighten up the flight controls as well for trim. I changed out the springs, not a noticeable change in control pressure but the extra loading of the servo made a noticeable difference. The Glasair 3 is twitchy much like the Venture, you might try adding some bias springs to help dampen the servo stepping. Again IMHO..

[sparxfly]

I changed out the springs, not a noticeable change in control pressure but the extra loading of the servo made a noticeable difference.

Just confirming- did you increase or decrease the spring loading?

[matthunter1448]

So I just installed my autopilot servos. My pitch servo has chattering when moving from the Forward(down) position to the center position but if you help it it's fine I don't know if it will do this in the air and I haven't tried yet but not sure if that's what it's supposed to do or not. It is acting as if there's somebody fighting it when moving but there's not and it's super easy to move when the servos electrically disconnected. Just want to know if this will be a issue in flight. It will chatter all the way untill it gets to the position it needs. Then moving to another position.

[John H]

I have seen, and it does the same on my RV9 and a recent install on a RV4. They will chatter, if you will, but if you put any loading on the control surface it will stop. That being said the aero loading while in flight you wont get it.

[JimJab]

Hi All,

I am yet to get my AP running as I would like, the roll servo is chattering way to much for me to use it in flight. Manly because I don't know much about servos and I am thinking I could be damaging it with this much chattering.

I have Trig Gold servos, one day whilst sitting in the plane on the ground, I turned on the pre-flight routine in the EFIS for the AP and the roll servo did its thing without a chatter, I thought wow.

I then turned on the AP lining up the heading bug my planes heading, only to find the roll servo was chattering away pretty badly without the sound of the engine going. I turned off the AP as quickly as I could, it sounded like a tooth had stripped from one of plastic cogs inside the servo.

I unscrew the servo cover and looked around inside the servo and everything seemed good, but for this chattering. Then I notice the servo chattering nearly stopped when I was sitting very still in the plane, when I moved a little the servo would start chattering away excessively again.

When I touch my AHRS mounting box on the inside of the cabin roof, I found I could move the box ever so slightly and this would make the servo chatter and carry on. I then held the AHRS firm against the roof and the servo stopped most of its chattering.

I have tighten the AHRS roof mount and there has been an improvement, but I am thinking I need to look into the AP settings for the roll in the EFIS, it may be just to sensitive.

But that' is job on my list for another day.

Cheers
JimJab

[sparxfly]

Following up on my original post on this thread re the AP in a Glasair 3....we are very close to a satisfactory solution, just some more tweaking to fine tune it over a series of flights.
Ther were two major problems- servos slipping/AP disengaging, and erratic pitch and roll control when engaged- plus one minor one of servo/control surface chattering when hit with an impulse.

The most significant cause was the power supply to the servos- particularly the roll servco.
It is out on the left wing with a relatively long cable run, in an almost inaccessible position.
Finally in desperation I made up an adapter harness so I could tap into and monitor wiring at the servo.
I found that with 12V at the battery, I was getting 9.5V servos disengaged, and about 7.5V engaged and under load.
Holding torque was quite weak and easily over-ridden.
Directly powering the servo with 12V and it had plenty of grunt!
The customer had used some existing 3 core screened 22AWG wiring in the wing for power and CAN bus to the servo.
Rewired with 18AWG power and ground, the servo voltage and holding torque was fine.
That fixed the slipping/disengaging problem- the AP held S&L through some moderate/high turbulence last test flight.
It will still slip/disengage if a 180 is commanded with the turn rate set above 2.5 degrees/second, but that is satisfactory for the moment.
The servo torques are currently set at 80% to enable torque sensing trim prompts- next flights we will set torques up to 100% and see if we can add a bit more crank and bank to the roll axis.

Fine tuning the pitch and roll magnitudes- with the servos now behaving themselves, I started with magnitudes set at zero and progressively increased them.
The Glasair is a fast, slippery and sensitive ship- roll/heading magnitude was best at 5, and pitch at 3- any higher and we started to get undamped overshoot/correction...
It would be nice to have finer/less sensitive control in both axes, but the servo linkage has to be set to handle full control throws without over-centre locking, so the throws to hold/correct in S&L and moderate turns is quite delicate.

Re the chattering under an impulse on the control surface- on the ground it is easily damped by hand, and is not evident in flight, presumably damped by air loads on the control surface.
So not an issue worth chasing.

Happy customer, relieved dealer!