Autopilot Question

[INA201]

2 hours ago, philiplane said:

Roll oscillations are caused by:

1-Using a narrow track width setting when in GPS mode, as in 0.3 mile setting instead of 2 mile setting in cruise flight

2-dirty slip rings on the roll servo. This can be tested by using a precision power supply hooked to the servo. The servo is supposed to start moving with less than 3 volts applied. If it takes more, the brushes and slip rings are dirty, so the computer will keep sending power until the servo responds. This causes overshooting. Pilots virtually never do the autopilot pre-flight check, which by itself will keep the servo clean through exercise. With a recent rebuild, this shouldn't be a problem, but it should be checked to rule out excess resistance.

3- loose bridle cable, allowing the servo to move excessively before the ailerons catch up.

We got the autopilot servo to engage and determined it as a connection issue.  I’ll try the GPS setting. Another setting I was told to look into was the sensitivity coming from the G5 changing it from a 1 to a .75. I’m holding off on the G5 setting as it’s possibly a threshold adjustment on the amplifier causing the mild oscillating behavior. 

[Bob Weber]

 The rigging should be checked as well, The Mooney's I've worked on have a bellcrank and a push/pull tube for the servo, if the Heim joints are worn they could be contributing to the issue.

 I always used a variable power supply and monitor voltage and current for the test. Slowly increase the voltage until the motor "breaks over", both directions. Record the readings, then run it for 30 minutes at full voltage (printed on the side of the motor) both directions. Recheck the break over. I've found at least 1/2 of the motors failed this test. The motors I used to build would break over at around 850mV drawing less than 250mA. Anything less than 1.5v and under 300mA after the burn in (full voltage runs), should perform correctly. That computer's threshold adjust is set around there, and, depending on the computer, is adjustable in flight.

  Once we get it back on rails, I would fly it in HDG, at cruise, with feet on the floor, and hands on your lap, for several minutes. We want to look critically for wings level flight, if it holds a wing low, indicating a skid, put a toe on the low wing peddle until it holds the wings level. At this point, check where the ball is, this is where it needs to be. I can't recall positively, but I don't think that aircraft has rudder trim, if it has a bendable trim tab, I would set it to hold it level in cruise.

 

 When an autopilot flies in even a slight skid, the computer is commanding the servo, the servo runs until stalled to hold the wing low. This can cook both the motor and the drive transistors, quite often shortening their lives.