Peter: I have no clue on how it works internally. I think it gets its data from the AHRS in the PDF. All I can say is that from my own experience, the DFC90 autopilot is dead-on on e.g. an ILS even with strong cross-winds as compared to the STEC-55 autopilot. The DFC90 performs like the GFC700 which is part of the Garmin Perspective setup I fly now.
The DFC90/SR22 install seems to have a conventional roll servo - OK.
But an AP which has no pitch servo and which (apparently) uses the elevator trim to control pitch is going to have a relatively poor performance in turbulence.
I was just trying to get my head around that bizzare description which frankly is written for idiots and doesn't explain "whatever it is" very well.
If I'm not wrong, every AP mounted on a Cirrus uses, only, the trim servos. And, for sure, it's the case with the STEC-55.
The STEC-55 is not stable in turbulent weather. The DFC90 is solid. Why, I don't know. They say that the DFC90 takes its signal from the AHRS module in the PFD while the STEC-55 not. I am not expert in this stuff, except that I have used both.
I am not expert in this stuff, except that I have used both.
Hmm, this reminds me of the SR22 accident in Zurich! I would think that a pilot should very well know where the autopilot takes its input from and which failure modes exist...
I read the manual and POH where it is specified how to use it. I am proficient in using the AP and can disable it or pull the fuse. I know of quite a few pilots who fly the Cirrus and have not a clue to how the autopilot works. I have no clue to how the wiring is. There is nothing in the POH or AP manual that gives me that kind of information and to be honest, I have no real interest in them. Maybe the SR22 accident in Zurich will make me think twice?
I think a reasonable level of systems knowledge is important.
But I suspect the SR22 crash could have been avoided rather more easily. For example, what was his bus voltage at the time he thought he was running on battery alone? If it was say 23.5 then yes you have lost all battery charging and you need to get down ASAP unless the conditions are CAVOK everywhere of relevance. If it was say 28.0 then at least one alternator is working fine and you are good to continue. Obviously an SR22 will have a voltmeter - somewhere... If you don't know stuff at this level then you should not be flying a C150, never mind an SR22.
I don't think a pilot needs to know where the autopilot gets its pitch data from, but he ought to know which systems depend on what. For example my heading gyro is on BUS1 so if I pull the CB on that, I will lose everything that depends on the heading: the EHSI(s), the HDG mode of the autopilot (ROL will still work, and you still have altitude hold, so hugely better than nothing), any overlays of TCAS or stormscope data on your GPS map will be disabled, etc.
During climbout, the voltage fluctuated between 23.6V and 28.7V, when the decision for diversion was taken, the voltage was 24.4V with 0.1V/min downward trend, during approach it fell to 12.1V where measurements stopped.
SUST found a few causes, but one of them was that the pilot apparently didn't really understand the electrical system of his plane.
There is nothing in the POH or AP manual that gives me that kind of information and to be honest, I have no real interest in them.
Actually there is, both in the S-TEC 55X manual and DFC90 manual you find detailed descriptions how they work. Imagine you're in IMC and your turn coordinator is causing problems. Proper systems knowledge will tell you that the S-TEC can no longer be trusted because it is driven by the TC and the DFC90 will probably turn off with an ADHRS miscompare alert because it uses the TC to monitor the ADHRS information it uses as primary input. Autopilots can have a few very nasty failure modes, also of the slow and steady nature.
I would think that detailed systems knowledge is more important in the Cirrus than with any other 4 seater because it has by far the most complex system and it is equipped to take you into situations where equipment failure can bite you.
A friend of mine did an emergency landing in his SR20 last week after a complete failure of the electric system. He was very surprised and worried by the extreme forces he had to apply to control the aircraft, he thought the electric issue (which started the usual way, first one device, then the next, then all) was "propagating" to the control surfaces. Well, the reason is simple: the Cirrus only has electric trim, there is no mechanical trim wheel. Had he spent more time studying the systems, then he would have been more relaxed without the fear of his elevator being stuck.
