The pic shows the other two horizons which I have, which were both running straight.
After about half an hour, the KI256 straightened itself out.
Could it be dirt in the bearings, which later got dislodged?
The vacuum gauge was showing a good vac the whole time.
The ball was in the middle, and the fuel tanks were balanced to compensate for me flying solo, but the two other AIs not agreeing is a bit of a give-away.
my KFC200 autopilot has stopped working, the flight director on the KI 256 is not functioning.we are not sure if is a computer fault ( although it runs the the test function ok) or the KI 256 (which was o/h less than 100hrs)
Probably a long shot but can anyone help who has a spare KI 256 that I could borrow to hook up to check it out .
The KI256 has a simple FD interface which is easy to check with a DVM.
The only UK shop which IME had any competence on King APs was Lees Avionics / Gama. Try to make sure Harry Lees is around.
Thanks Peter, where is the FD interface to be found?
KFC200 FD driver is inside the KC295 AP computer
even IF the FD part of the KI256 is unserviceble, the AP would be working.
I did not mention, the flight director works in roll but does not come up in pitch.
Rob – if you are doing this yourself you need a very good electronics engineer, with access to the maintenance manuals (those are not hard to find if you have contacts in low places).
The electronics involved is simply 1960s/70s stuff but to troubleshoot it one needs to actually understand electronics. This is why the vast majority of avionics shops can’t do this work – they are just wiremen, with a phone line to Garmin if the thing doesn’t work.
The KI256 acts as the pitch/roll source for most King autopilots. It contains what is basically an LVDT, bent into a circle, for pitch and another for roll. This is excited with supposedly 400Hz (actually more like 500Hz in some cases) from the AP computer.
The FD is an indicator only, with nothing going back to the AP computer.
Unless you find a friendly avionics shop you may find it difficult to borrow a KI256 because they are worth about $10k and if somebody has a known-working one, they will not be keen to lend it out because it will be a valuable spare part.
I have seen this twice so far.
One was years ago.
Another was today, FL100, in cruise i.e. a ~ 2.5 deg UP pitch. I suddenly saw the autopilot take a dive, losing a few hundred feet in maybe 10-20 secs. So I did the usual debugging, flying by hand and engaging the FD-only mode and see if the FD is asking for the dive i.e. could it be the pitch servo. It turned out to be quite simply the KI256 indicating about 10 degrees UP when it should be indicating a “few” degrees UP (the KI256 is not good at showing these small angles but my Sandel system is very much more precise).
The KFC225 will compensate for small deviations e.g. 1-2 degrees but it can’t cope with the pitch returned by the KI256 going up by best part of 10 degrees in such a short time. That causes a de facto altitude hold disconnect.
So I hand flew for a bit, and watched the KI256 pitch indication settle back down to the more normal 4-5 degrees (the little plane indicator on it may not be in the right place ). Then re-engaging the AP normally was OK.
The whole episode lasted about 20 mins.
How is this possible? There are only the bearings in there…
The vacuum was always perfect.
I have a spare KI256 on the shelf, overhauled by a now-defunct UK firm but should be working fine. But the old one is working fine too…
Looks like a dirty gyro. I had the same kind of problem in the roll channel of my AI several months ago (on my IR check ride, no less). An avionics repair company here in Prague cleaned it for some 125 euros.