I've researched the forums extensively on this topic and most of the posts I've read have been helpful but are now fairly dated so I think it's worth asking for some present-day advice...
My PA-30 has an Altimatic III Autopilot. Here's a pic of the control head:
Also related - it has a selector switch to govern what heading / nav source it's tracking and just under that is a separate on/off switch and trim adjustment for the backup "Autoflite" system. Here's another pic:
I've read the book several times and experimented with all this in flight. Basically the Autopilot itself is Attitude based and is based on the vacuum driven attitude indicator. As a backup Piper incorporated the "Autoflite" system which is based on an electric turn-coordinator type system and is really just a wing leveler. The idea being that if you are IMC and your vacuum-attitude based autopilot fails you can switch it off and turn on the Autoflite wing leveler as a secondary / backup system.
I have to say that from what I've read / experienced and understand, this autopilot was way ahead of it's time (perhaps not unlike Comanche aircraft in general). If working properly there is really nothing this autopilot won't do. It has altitude pre-select (albeit on a fairly "gross" scale), auto-trim, a glide slope coupler (where installed). and other features. It even incorporated an air-speed limiter to prevent the aircraft descending at such a rate it exceeded airspeed limitations (Vno / Vne) and on the climb it prevents climbing into a stall by swallowing the climb to avoid getting too slow. As an attitude based autopilot the turns are crisp and decisive. I am used to flying with a rate-based S-Tec 55X and have now come to realize how much more "crisp" the turn in and response are in an attitude based autopilot like this old Altimatic (also our S-Tec 55x in the Cirrus is not too good in turbulence being rate based).
On the down side... the altitude pre-select is on a pretty "gross" in it's scale. It's not like setting a digital altitude in tens of feet but rather turning a knob to rotate a drum (see the first pic). Also it has to be calibrated on each flight due to baro changes. The disadvantage of being attitude based is that it's based n a vaccum attitude indicator spinning 3x as fast as a turn coordinator. I've heard it argued that a rate based S-Tec and the turn coordinator gyro are more reliable.
My autopilot performed well on the test flight pre-purchase. I've read a post from Kristen Winter (pretty sure it was Kristen's post) where she advised the autopilot as an item to test thoroughly and carefully and I have to agree with this. I probably did not test mine enough pre-purchase and have discovered just how crazy expensive these things are to repair or replace.
Here's where I am today... my Altimatic III will track a heading or nav source (VOR/GPS) very well. On a LOC it will track but did seem to S-turn quite a bit before locking on. Probably an adjustment issue. I don't think I have a GS coupler or if I do it's not working. It will hold altitude if I play around with it. However I have never seen the autopilot adjust the electric trim so I think the autotrim is probably inop. Also, the altitude preselect works some of the time but not often and not with any precision. My impression is there are dirty contacts in there somewhere and I can get it to climb but not descend. Also it's out of calibration. The little round window with the horizontal line used for calibration is slanted slightly down when it thinks it's level. Finally... my Piper Autoflite system is completely Inop and I don't think it worked when I bought the plane. The prior owner said it did but candidly I did not test this and I've read alot about mis-understandings of how these systems work. For instance, if you don't turn the Altimatic Off before turning Autoflite on then the wing leveling you are getting is coming from the Altimatic not the Autoflite.
I spoke to Bob Furguson out at Autopilots Central. He said the systems can be made to work. He said the only part they can't get / replace is a "bellows" that is used in the altitude hold function. He said if mine holds altitude at all (which it does) then that is probably not a problem. He said to fix it they would really need me to fly the plane out to them in Oklahoma and let them pull the whole thing apart, overhaul it, then reinstall it all and test / calibrate. Cost wise he said he could only give me an "idea" but worst case it could approach $10K. So I might spend $4K if not much is wrong but maybe I'd spend close to $10K. If I went this route and everything was made to work like it should it seems like I would have a very capable autopilot with features that would cost more like $30K installed if I were going with a new S-Tec of similar capability. That said I worry that I'd be "throwing good money after bad" investing so much in a 45 year old autopilot with limited parts support.
Another idea is to spend say $15K installing a basic S-Tec 30 or 50. That would give me a new autopilot and while it would cost more (no knowing how much more since I really don't know if the Altimatic repair would be 4K or 10K) it would at least be a known outcome (or should be). it would have altitude hold but no vertical speed or altitude preselect or even a glideslope coupler but no big deal I'd just hand fly approaches, I often do anyway.
While I've had no problems with the S-Tec 55x in the Cirrus I have noticed quite a few posts on the web on different forums about S-Tec issues - particularly since they were acquired by Cobahn. I've read comment about lower quality components being used in servos and the like but I don't know if that's really true.
Another question... what about Century? A Century 2000 is attitude based and the costs seem lower but avionics suppliers and shops I've spoken to seem to recommend away from Century and towards S-Tec. Many seem to say that S-Tec is the "only game in town". I've heard shops say that Century won't work with shops they only work with owners so shops prefer S-Tec.
I've also called Avidyne and ended up emailing with the VP of Product Development in charge of the new DFC-90. Predictably he would not give me any info on if or when the DFC-90 would be available for PA-24/30/39 aircraft but he agreed it would be a good platform. He said a Mooney 201 or similar would probably be the next platform and that they need to get through their existing pipeline of commitments before considering other airframes. So for now the DFC-90 is not an option, maybe it will be one day.
Anyway that's a lot of commentary... I'd appreciate any thoughts or feedback from those who have already been down this path, especially recently...
Thanks,
- Charles