Pitch servo always climbs!

[fabiocala]

HI!
I bought an used Tecnam P92 Super with Rotax 100cv.
It has:
Dynon D10-A, AP-74, servo SV42 on roll and servo SV42 on pitch.
It has also a Garmin 695 and a remote compass and an OAT sensor.
DSAB and Network Status is ok (no alerts/error message).
The electric trim is only manual (it is not connected to the AP system).
Everything is ok with Roll (on TRK, on HDG, on NAV).
The problem is on Pitch-servo, and the previous owner told me that this problem arises from the beginning.
I came back to the professionist that installed the whole AP system:
he tried to change the pitch servo but the problem continues.
The problem is that the servo climbs all the time!
We tried all combinations of settings (TORQUE, SENSITIVITY, PITCH GAIN,ALT GAIN,PULL RATE,VSI GAIN,G ERR GAIN,G ERR LIMIT) but the problem continues.
The MAX AIRSPEED e MIN AIRSPEED are set correctly.
For example:
I fly in no-turbulence and without small bumps and I trim to maintain the altitude with AP off;
when I activate the AP on ALT (to maintain the altitude) the servo, after 4-5 seconds, acts to climb until I turn it off: I activate ALT to maintain 800 ft and just after less than 2 minutes I'm at 2000ft (but the slope is not constant)!
If I trim to descend during it climbs, it continues to climb; if I turn off the AP the plane immediatly descends!
We did calibrations lots of time and the servo-tests are always ok.
The informations provided by EFIS are always ok (IAS, bank, VSI, Horizon line, pitch and roll indicators, etc etc).
We tried to activate only ALT (no HDG/TRK/NAV) but problem continues.
The trim indicator sometimes flash on and off but doesn't remains on for several seconds.
We already tried to change the Long-arm variants ("holes") on pitch servo but problem continues.
I noticed that if I activate the ALT AP when the plane is in hangar (engine off) and I want to 'maintain the altitude'  the servo tries to climb, as when it is in flight.
Any idea?
Thanks!
Fabio.

 

[Dynon]

Nothing obvious is immediately jumping to mind, but a few thoughts: First, if the servo's shear screw were broken, and the servos weren't actually controlling the aircraft, you'd expect the aircraft to basically not obey the autopilot commands most of the time. The residual friction between the arm and the servo action might cause you to experience some limited autopilot control, and perhaps more in one direction than the other, so this is something to look at. But, I think we should troubleshoot this with you directly. Can you send an email to support at dynonavionics dot com with the information you've provided here, but also ALL of your autopilot settings? Additionally, if you have the ability to take a video of the EFIS screen while you're experiencing this behavior, that might give us some hints about what is going on here.

 

[fabiocala]

Hi,
the shear screw is not broken, anyway we already changed it and it is ok: infact the servo tests are always ok.
We tried starting with default parameters (TORQUE:100%, SENSITIVITY:10, PITCH GAIN:20, ALT_GAIN:6,
PULL RATE:10, VSI_GAIN:15, G_ERR_GAIN:10, G_ERR_LIMIT:25) but problem continues.

Are there any other AP settings?

We tried almost any combination of parameters but problem continues; high values and low values, the servo always climbs!
P.S.:
- I have a Ivoprop propeller with manual governor (NOT automatic!) and I don't change the propeller pitch during tests.
- The EFIS-D10A Firmware is 5.5.0.

I will take a video as soon as possible; for the wile I send an email as you suggest me.
thanks!

 

[dynonsupport]

Please check your G-meter in the D10A. If this is not accurate, reading 1G in level flight, the autopilot will not work properly.

 

[fabiocala]

Already done (I forgot to tell you that): it is 1!

 

[airguy]

Is there a max airspeed value that the system is trying to protect by pulling up, like it's thinking you are exceeding Vne?

 

[fabiocala]

No.
As I already wrote, the MAX AIRSPEED e MIN AIRSPEED are set correctly, and during tests I have constant VSI and constant IAS, far from MAX-IAS and MIN-IAS.
It's a very mistery!

 

[fabiocala]

...maybe the Tube/Pushrod (of Mounting Brackets & Kit) is too long or too short or has an wrong angle? (see the photo above)

 

[dynonsupport]

At this point, please contact Dynon Avionics Technical Support - email support at dynonavionics dot com to receive instructions to send a datalog to Dynon Avionics for our AP engineer to evaluate. In your email, please provide a link to this discussion thread and mention that you have an EFIS-D10A.

 

[fabiocala]

Hi everybody!
It took time, but  now  my problem is fixed!

To recap, the solutions are the followings:
1) I had to add 2 more arms to the mounting bracket to fix the mounting bracket  strongly (someone, that often writes on this forum, helped me by phone/mail to verify if the old mounting bracket was moving while the AP was engaged)
2) I had to substitute SV42 with SV52 to eliminate the slip indication on pitch servo.

After installed new mounting bracket and SV52, I did a flight with default parameters: there were lot's of clouds and small bumps, but the autopilot held the altitude all the time: it is ok, now!!!!!

I HAVE, AND I WANT, TO THANK THE DYNON TECHNICAL SUPPORT AND THE INSTALLATION ENGINEER (who writes on this forum) WHICH, WITH GREAT PATIENCE, HELPED ME TO FIND AND SOLVE MY PROBLEM.

Fabio.