Intermittent but Persistent Problem

[bhilgy]

Hey all!  Looking for ideas to help diagnose this one:

I have a intermittent but persistent problem with the Autopilot/Nav in my M20J.  I have an Aspen PFD with EA100 coupled to a KAP 150 AP and 530W.  Not sure it's important but also have a GTX-345 with ADSB-IN/OUT and Garmin Flightstream

Problem showed up on a flight to Chicago from Branson last week and has displayed on each of the last 3 flights since that one.

When:  At cruise, between 30-90 minutes into a flight (happening sooner during each subsequent flight).

What:   The heading begins to drift to the left with AP engaged and GPSS engaged on the Aspen.  When it starts to happen the 530W gives me a “set heading to XXXdeg” alert.  Take the GPSS off and the AP will not following the heading bug.  Cycling the ASPEN, 530W, EA100 has no effect.  After hand-flying for a while, the heading function may come back but GPSS will not engage and plane will eventually roll to the left and fails to maintain heading.  Problem resets for initial phase of next flight, then returns.

Grateful for any ideas or next steps to help solve this one.  Hate to start replacing equipment to resolve.

[Marauder]

Hey all!  Looking for ideas to help diagnose this one:
I have a intermittent but persistent problem with the Autopilot/Nav in my M20J.  I have an Aspen PFD with EA100 coupled to a KAP 150 AP and 530W.  Not sure it's important but also have a GTX-345 with ADSB-IN/OUT and Garmin Flightstream

Problem showed up on a flight to Chicago from Branson last week and has displayed on each of the last 3 flights since that one.

When:  At cruise, between 30-90 minutes into a flight (happening sooner during each subsequent flight).

What:   The heading begins to drift to the left with AP engaged and GPSS engaged on the Aspen.  When it starts to happen the 530W gives me a “set heading to XXXdeg” alert.  Take the GPSS off and the AP will not following the heading bug.  Cycling the ASPEN, 530W, EA100 has no effect.  After hand-flying for a while, the heading function may come back but GPSS will not engage and plane will eventually roll to the left and fails to maintain heading.  Problem resets for initial phase of next flight, then returns.

Grateful for any ideas or next steps to help solve this one.  Hate to start replacing equipment to resolve.

I have an Aspen but with a
GTN 650 and STEC 60-2 autopilot. I know there have been issues with the King APs and the EA-100. That would be my first suspect.

The Aspen’s failure mode is wings level. Since it isn’t doing that, I would investigate whether the EA-100 is functioning correctly.



Sent from my iPhone using Tapatalk Pro

[bhilgy]

Thanks, Marauder.  I asked my Avionics shop to check into having Aspen send me replacements if I commit to a MAX upgrade as soon as they are out.  Crossing my fingers.  ;-)

 

[larryb]

I had a similar problem and it was a problem in the roll servo. The motor had become loose in its mounts. A trip for the servo to mid continent for repairs and reinstall by my avionics shop and all is well.

 

 

Testing the AP on the ground showed an obvious problem in the servo. It sounded sick and moved sluggishly.

 

[Warren]

After my GTN650 upgrade I had similar issues — I learned a lot about how things work.

The Aspen only provides heading references to the autopilot as a direct heading from the Aspen heading or a passthrough when in GPSS mode.  The non-Aspen attitude indicator is responsible for the analog attitude indications which the autopilot uses to make corrections to course.  With these analog signals and the autopilot there is no error indication except the autopilot appears to lose its mind.  During my install/upgrade one of the wires was broken (poor previous install with a DB9 connector and no strain relief) between the attitude indicator and the autopilot — of course it was intermittent, adding to the challenge.  After chasing it for a long time I repaired the wire, added a shell for strain relief and it has worked flawlessly since.   There are only 4 wires, so relatively easy to trace and confirm good connections.

Edit (added after I realized this was not accurate in all cases):  This does not apply if you have the EA100 adapter.  That can provide attitude information from the Aspen to the autopilot.

Edited August 8, 2018 by Warren

[khedrei]

16 hours ago, Warren said:

After my GTN650 upgrade I had similar issues — I learned a lot about how things work.

The Aspen only provides heading references to the autopilot as a direct heading from the Aspen heading or a passthrough when in GPSS mode.  The non-Aspen attitude indicator is responsible for the analog attitude indications which the autopilot uses to make corrections to course.  With these analog signals and the autopilot there is no error indication except the autopilot appears to lose its mind.  During my install/upgrade one of the wires was broken (poor previous install with a DB9 connector and no strain relief) between the attitude indicator and the autopilot — of course it was intermittent, adding to the challenge.  After chasing it for a long time I repaired the wire, added a shell for strain relief and it has worked flawlessly since.   There are only 4 wires, so relatively easy to trace and confirm good connections.

This seems to be what my centry 23 is doing.  It used to follow the nave signal fairly well. Then it only worked to follow the heading bug.  Now the nav doesn't work at all and the heading mode is intermittent. Sometimes it simply wont follow it and does nothing but still engages the servos. Most of the time it violantly banks the airplane in one direction or the other. 

 

Where did you find the bad connection? And what kind of connectors did you used to fix it?

[Warren]

17 minutes ago, khedrei said:

Where did you find the bad connection? And what kind of connectors did you used to fix it?

My bad connection was on the wiring from the backup AI.  About 1 foot away (middle of the panel on the pilots side).  In my case it was a DB9 connector that had been wired in.  However, the shells were not installed on the connector and thus the wires had no strain relief.  Thin wires, no strain relief -- it was only a matter of time before failure.  It was a quick fix to replace the pin on the broken wire.  I ordered DB9 shells and added them to the connector.  These are readily available and pretty simple to install - search DB9 connectors and you can find lots of options.