Peter wrote:
You need the servo MMs with the circuit diagrams. […] KS271CThat’s the one I was referring to. But I haven’t look at the actual circuit diagram — just the block diagram. More interesting bedtime reading, then…
You need the servo MMs with the circuit diagrams. KS270C KS271C TB20 TB21 STC IM. The last one is for the installation overview. I have all the schematics of course, including the KC225 
Peter wrote:
The “valid” output from these servos is a fake signal. It is derived from the amplifier output, only, and is thus valid even if the motor is burnt out. All it checks is that the amplifier output is the right polarity
According to the block diagram on page 4-5 of the KS 271C servo MM, the valid signal does make use of the tach output — or rather it compares the amplifier output to the speed error (the difference between the servo command input and the tach output). I can’t quite make out how that is supposed to work and I can’t find any explanation in the text of what exactly is required to get a valid output.
See here
The “valid” output from these servos is a fake signal. It is derived from the amplifier output, only, and is thus valid even if the motor is burnt out. All it checks is that the amplifier output is the right polarity 
I don’t think you can buy the motor assembly alone. I tried but failed. It is listed on the usual aircraft parts dealer websites but none of them had it in stock when I looked; they just list every possible P/N in the hope of getting an enquiry. The motor can be rewound, with some difficulty, by any motor repair shop; see my notes. Obviously it can be done only off the books; you won’t get a certified repair unless you send the servo to Honeywell.
I’ve also checked our broken autopilot using a PC. It turns out that while the roll servo clutch works, the ailerons don’t move when you give a command to the roll servo motor.
So it looks like the servo motor or its driving electronics are broken. What confuses me a bit though, is that the “valid” output from the roll servo remains true. I would have thought that if the servo tachometer doesn’t register any movement, this output would be false. So it could be a mechanical problem?
@Peter, would you have any idea? The KAP 140 uses the same roll servo as your KFC 225.
Today we had the exchange roll servo installed.
Now the autopilot is working again.
It was already clear when the old came out that the shaft was free wheeling. We did not open it but definitely some mechanical failure.
Now we hope it is accepted in return so we only have to pay the exchange fee.
Today we had the PC connected to the KAP140 again.
We first verified the 28V and found that OK.
Then we measured between CMD and CMD GROUND
Forcing varying outputs with the PC we measured between -11 and + 11 V
As I understand that should have been -9 to +9 V, anyway the motor did not move at all.
We are going to order an exchange servo and get that fitted.
pmh, you might want to get a hold of Martin Scheifl, at Avionik Straubing, EDMS.
He’s about the most knowledgeable guy I know about autopilots. We had a problem with our KFC150 servo once; while two other avionics guys were guessing around and wasting our time, Martin found the reason real quick. You might want to try to call him and see what he suggests.
Just my two cents.
I want to give it another go and try to measure the voltage reaching the connector before the servo when forcing an output from the PC.
Is it correct that +9v to – 9v shall be measured between
D – COMMAND REF
and
L – COMMAND
and the measured voltage shall be linear in respect of the output setting between -1 and 1 on the PC?
