You previously wrote about small VSI indications. Are these related?
Does it happen at all altitudes? Can you simulate the issue on the ground?
I couldn’t notice FD indications, could you hand fly in FD mode, and see if the FD would be oscillating as well?
Didn’t we have a discussion on this before? Can’t find it now? That was on KAS297 altitude fluctations. The KAS297 uses and altitude encoder input (100 Ft incrediments) while the KFC150 uses it’s own altitude sensor which much high resolution and accuracy. Issues with this sensor would cause the same symptones both when flying on the autopilot as well as flying manually with FD on.
There is (was) a modification on the KS270C/KS271Cto replace the tach generator with a more solid one.
But even those have sometimes unstable output.
When you monitor the Tach-output on the test bench it looks stable until you start tikking on the tach-generator.
this could maybe be your problem ? or maybe the Att-gyro output ?
You previously wrote about small VSI indications. Are these related?
I don’t think so.
Does it happen at all altitudes? Can you simulate the issue on the ground?
On the flight today, Brac-Shoreham, it started doing it and after about an hour it was fine! God knows what it was…
I couldn’t notice FD indications, could you hand fly in FD mode, and see if the FD would be oscillating as well?
Good point. I didn’t try that.
The FD oscillation was there in ALT & PIT & VS modes (all modes where the AP is driving the pitch servo) but very small in amplitude and only just visible.
Didn’t we have a discussion on this before? Can’t find it now? That was on KAS297 altitude fluctations. The KAS297 uses and altitude encoder input (100 Ft incrediments) while the KFC150 uses it’s own altitude sensor which much high resolution and accuracy. Issues with this sensor would cause the same symptones both when flying on the autopilot as well as flying manually with FD on.
I don’t think that was related because the altimeter (KEA130A) never moved even a tiny fraction.
There is (was) a modification on the KS270C/KS271Cto replace the tach generator with a more solid one.
But even those have sometimes unstable output.
Unbelievable! Do you have any details? I looked into this extensively and found a higher quality tacho, with proper long life carbon brushes, but it was out of production. The crappy tacho in these servos use metal brushes.
When you monitor the Tach-output on the test bench it looks stable until you start tikking on the tach-generator.
this could maybe be your problem ? or maybe the Att-gyro output ?
It will be easy to test the tacho, for sure. But even replacing it is easy – they are £40 from Faulhaber.
or maybe the Att-gyro output ?
Yeah, that would be really great (not). A duff KI256.
But as I say the issue disappeared today, which makes it all the more puzzling.
My suspicion is that the pitch servo has an excessive gain which makes the control loop unstable, for certain aircraft loadings.
Peter wrote:
I don’t think that was related because the altimeter (KEA130A) never moved even a tiny fraction.
I think you make an incorrect conclusion. The KAS297 uses the KEA130 output, which is in 100 Ft steps. So for example if you fly at 8000 Ft, the output will be the same from 7950 upto 8050 Ft. Within this area the internal pressure sensor inside the autopilot computer is used. So variations from within for example -20 to +20 Ft could be caused by internal pressure sensor issues. For example due to poor capacitor, which makes input, reference or output of this sensor unstable.
Peter wrote:
My suspicion is that the pitch servo has an excessive gain which makes the control loop unstable, for certain aircraft loadings.
This is where manual flying with flight director would be helpfull, if it is in the computer of gyro your would see this on the flight director, if the flight director stays rock solid something is wrong with the servo.
Peter wrote:
My suspicion is that the pitch servo has an excessive gain which makes the control loop unstable, for certain aircraft loadings.
I have seen this in a Pa46 with an Avidyne DFC-90. At certain loadings you got some roll oscillation.
pitch servo overcompensating. Can be adjusted.
Euro flyer wrote:
‘pitch servo overcompensating. Can be adjusted.’
How can it be adjusted?
could be caused by internal pressure sensor issues. For example due to poor capacitor, which makes input, reference or output of this sensor unstable.
I thought that the internal pressure sensor is used only once altitude has been captured. However it probably is used during climb and descent too, to extract more resolution out of the large steps of the KEA130A.
However I don’t think the pressure sensor is used in the PIT mode, when the pitch is controlled according to a combination of the internal accelerometer and the pitch output of the KI256.
I have seen this in a Pa46 with an Avidyne DFC-90. At certain loadings you got some roll oscillation
It is really bad that the DFC90 does that. They obviously haven’t tested it properly for that airframe, all around the loading envelope. Roll oscillation is extra weird though. That said, I got it once in the TB20, but it never came back. That was about 10 years ago. Another one-off issue…
The pitch oscillation issue is otherwise well known with STEC autopilots most of which were never fully tested and their STCs were generated on the back on some STC – it was the only way to get the large number of STCs.
I saw similar behavior on KFC150 in my TB20 at certain loadings at FL120 and higher.
Peter wrote:
However I don’t think the pressure sensor is used in the PIT mode, when the pitch is controlled according to a combination of the internal accelerometer and the pitch output of the KI256.
Yes I agree, Checking F/D during manual flight should be the next troubleshooting, then you know if it is the input or ouput side.
