I utilize a S-Tec 50 with their heading gyro and driven by a Garmin 430 plus GPSS. My 320 has about 600 hours, most of the time (400 hours) is from cross-country flights and most of that time is with the A/P engaged.

 

I have frequently complained about how the use of the tilted "rate" gyro in the turn coordinator translates turbulence into turns - big turns. I want to retract "big."

 

Last year in the Kittyhawk to OSH race, I wanted to utilize GPSS and the A/P to manage the most efficient flight line (for race purposes).  Of course, flying low at midday usually results in turbulence and those flight legs were no disappointment in jostling that surely took an inch off my height after each flight.  I had to discontinue the use of the A/P because moderate turbulence (light chop to ATPs) led to some deviations that were as great as 12 degrees with very poor recovery to the course line. The A/P had never been this bad although its' usual use is at higher altitudes and in smoother air.

 

Later, I consulted with the local avionics shop and they suggested that carbon buildup on the armature/brushes makes the servo motor not very responsive.  They made two suggestions: 1) Remove the servo motor and clean it or, second best was; 2) Run the motor at 40 VDC for 5 or 10 minutes on a yearly basis to clean up some of the carbon.  I had them perform #2 because, for me, removal of the servo motor from the wing is no simple task.  Test flights indicated a significant improvement in that "big" deviations had become less big - more like 4 or 5 degrees, still not the best.

 

Recently, during a cold and miserable day, I pulled the servo and removed the black plastic enclosure (one screw on the back), secured the wee springs holding the brushes in place, took out the two screws that hold the bearing/brush block in place and, using q-tips dipped in alcohol, cleaned the brushes and the armature of quite a bit of carbon. Reassembly included securing the two screw heads and their lock washers with a touch of Locktite. The following test flights were much more satisfying in that turbulence would only produce a 1 degree (or so) deviation.

 

So.....

 

If you are experiencing sloppy course tracking during turbulence (under "perfect" GPSS guidance), consider cleaning the servo motor.  Report back if you find improvement.  This suggestion is meant for those whose A/P has been in service for some time and performance has degraded even though the best possible GPS signals are being used.