[ Autopilot posts moved here from the Twins thread ]
boscomantico wrote:
Of course, depending on where you come from, it has great features
Well, what is the alternative? All i hear is the system being garbage but when asked for an alternative system with STC e.t.c. there usually is silence.
Clearly, the DCF90 or G700 offerings are more advanced but that helps preciously less if there is either no STC or they are not available for retrofitting in the first place.
boscomantico wrote:
Butto be honest, GPSS and preselect via the PFD is nothing new. This has been available for over 10 years. Again, the verdict depends on where you are coming from.
Not in the integrated world no and it was possible via the GPSS converter by… S-Tec. The variant often used in retrofits is via the Aspen, which can do GPSS via Heading Mode for legacy AP’s. Difference is, the S-Tec 55X can do GPSS even without a PFD. Apart from the DCF90 and S-Tec55x I don’t know of any other retrofit AP which has GPSS integrated in that way, most need expensive converters to do it.
Clearly the AP Suites of the G700, DCF90 as well as the usual airliner offerings such as in the Airbus series e.t.c are more advanced. But they are either in a totally different price league or not available for the airplane due to lack of STC, not being available for retrofit at all or not relevant.
So what should be retrofitted onto a legacy aircraft? Mind, most Cirrus were delivered with the 55X before the Avidyne offering came along. And before that, I wonder where all those who point fingers now “came from” before that? As I said, I have flown airplanes with the King KFC Systems, they are much more expensive to retrofit and from what I hear from Peter for instance I am not so sure of their superiority.
boscomantico wrote:
What’s bad about the S-TEC is the price, and the technology that you get in return for that.
And the alternative being what? The S-Tec 55x System costs around $20k plus 5k for the Autotrim if you want it. Any other system for this kind of price with the features it has? AFAIK if you want to buy the Avidyne system even for those lucky guys who do have an STC, you need an existing 55X installation first.
Most legacy planes which have a working AP such as the older KFC systems or in the case of the 414 Cessna systems would not change for an S-Tec anyhow, but for those who want to replace a legacy Brittain or similar system there is preciously little choice. Which of course also explains why those systems can be so expensive and still are bought. But that goes for all avionics, not only AP’s and certainly not only S-Tec.
Emmanuel wrote:
I noticed my PA34 (Stec55X equipped) had problems to capture a localizer course correctly. It always overshot the axis and sometimes I even had to capture by hand (switching from AP to FD).
In the test flying we did and the subsequent familiarisation flights we had some slight Loc overshoots with crosswinds exceeding 30 kts during capture which however were corrected by the AP quite adequately, no manual intervention was needed.
What Turn and Bank do you use with it? Did it come with the AP? When I purchased the system, a T&B was included and came brand new.
The one bit I don’t like about the 55X is that you need an expensive remote display to use the FD function. That was definitly better thought through by King and Avidyne.
Most of the 55X criticism seems to be related to its rate-based nature. I wonder if this is less of any issue in a larger/twin aircraft like the 414 with more roll inertia.
If the vacuum pump or AI in your attitude-based AP fails in IMC you might be wishing you had the 55X, so it scores higher in this regard.
And presumably the second axis (alt hold / VS) performance of the 55X is fine as not affected by the above. I don’t remember reading otherwise.
Rate based autopilots are entry level simple products for C172 and the like. They act based on after-the-fact-information (aircraft has turned because it changed attitude). There is simply no comparision to attitude based APs, especially for faster airplanes. A Cessna/ARC 400, KFC225/335 or any other would be suitable but an S-TEC 55X is a joke in an aircraft like that. The TCDS has others but they’re old of course.
I would never do fancy seats and stuff and then fly with a totally inadequate S-TEC rate based AP. Not in this class. Ask any Cirrus pilot who upgraded to an attitude based digital AP. Worlds apart. And a C414 is quite another class and was originally equipped with an attitude based AP!
No I don’t think so.
A turn coordinator is sensitive to roll rate as well as rate of turn. So a rate based AP can actually detect a pending change of attitude before and earlier than a attitude based AP, it doesn’t need to wait until a turn has started. That would only be true if you had an AP working off a “turn and bank” rather than “turn coordinator” instrument. If you hand fly in turbulence (assume VMC) you are probably making corrections by feeling a wing start to drop, which is feeling roll rate rather than waiting for an attitude change, which makes the flight smoother, it’s the same thing.
Secondly, an AP than is driven by an AHRS with MEMS gyros is using similar rate information anyway, and integrating it to get attitude, so there is no reason why the 55X couldn’t do something similar to derive reasonable attitude information.
Now if the 55X software is poor and doesn’t perform well in practice, or if the turn coordinator performance is inadequate compared to a MEMS gyro, that is a different matter, but I don’t see why rate-based APs inherently have to wait for the aircraft to start turning before they can do anything useful. If anything it is the opposite.
Thread drift. Needs an “Attitude vs Rate Based APs” thread.
ortac wrote:
feeling a wing start to drop, which is feeling roll rate rather than waiting for an attitude change
Huh? Roll rate is by definition the derivative of attitude, there cannot be a “roll rate” before the attitude changes. Therefore, an attitude based autopilot has all the information a roll rate based one has, plus it lacks the integration drift a roll rate autopilot suffers from. That means an attitude based autopilot knows when the aircraft is upright, but a roll rate autopilot can only infer that from heading changes.
Fuji_Abound wrote:
the Aztec has the old Century Analogue Autopilot
This one?
I once found this thing in a Seneca I. When I did the introduction to that airplane the owners (a flight school who had acquired the plane in a take over and clearly did not like it much) told me “leave that miserable piece of garbage, we could never make it work”. The same went for the two Rajay Turbos on that airplane.
That was before the internet really took hold so no Euroga community to ask. During the intro I insisted on trying it in HDG and ALT hold and it worked just about fine. Later I took the manual and self taught myself about the altitude preselect system and the rest of the functions and used them extensively. They worked just fine, including the alt preselect, which, if you kept an eye on it, did the job pretty much as advertized. Couple of weeks later I did an IR check on that airplane with the same lady and she was stunned that I flew most of the flight including a coupled approach with it with no discernable problems. Ah yes, I also demonstrated that even a Seneca I can hold 8000 ft single engine, if you use the Rajay on the running engine. Coming from a Seneca II with KFC150 I found it a step backwards but not too badly so, the KFC 150 did not have altitude preselect for a start.
I really wonder how much of the spite which is often extended to such systems (or aeroplanes) come from real life experience or just hearsay.
Oh, and those who moan about a functionality of a Century or S-Tec are very welcome to try this one.
Not only does it work fine it is also CAT II approved. It took me a while longer to understand it than any GA AP, even though in some aspects the alticontrol III comes close.
Are rate based autopilots inherently inferior for the roll axis and if so why? Or is it just that certain rate based products aren’t very good?
This thread is not about the STEC 55X in particular which might just have poor software and poor tuning for a given aircraft. And it’s not about the pitch axis as I think it is a given than having a pitch attitude reference is advantageous over just a static pressure source.
A comparison of the KAP140 and the KFC225 would be most useful here as one is turn coordinator driven and one is AI driven, but otherwise a similar product from the same manufacturer. Does the KAP140 have all the same issues as the 55X or in reality is it closer to KFC225 performance?
And does the potential safety/redundancy advantage of a rate based autopilot offset any possible performance disadvantage? Which would you prefer to rely on in challenging IMC conditions?
ortac wrote:
Yes that’s what I was getting at earlier with regard to the 414. That airframe might fly really nicely in turbulence even with the 55X autopilot, simply because it’s a bigger airframe with a couple of engines a long way out from the roll axis. And a higher wing loading too.
I actually discussed it with someone who knows the S-Tecs really well following the reactions here and has done quite a few upgrades with Cirrus airplanes. What he said was, that the actual way the 55x is installed and programmed differs hugely between airframes. He has had very few complaints about it from Mooney owners, mainly because the way the flight controls are set up, which helps damping the fishtail effect some people complain about in other makes. Interestingly, most complaints came from the Cirrus installations, which of course is nice for Avidyne but also from some Cessnas, where apparently in larger airplanes it seems to work adequately again. I also heard the same from the guy who did the test flight on our Mooney after the upgrade, he owns 2 Cirrus planes, one of which recently got upgraded from a 55x to an Avidyne and one thing he said is that he was amazed how well it did on this installation after his rather negative impression in the SR22.
So I can really imagine that you are right. If I recall my own experience on the different Senecas, they fly rather stable and have the reputation of a truck like handling, mostly for the Seneca I that is pretty true! So I can well imagine that an AP which shows the characteristics some quote here may be less than ideal for agile planes like the Cirrus but fine for “heavier” controls.
ortac wrote:
Are rate based autopilots inherently inferior for the roll axis and if so why?
A clear yes. They can only sense a turn rate (and counteract it) which is the consequence of an attitude change. They are therefore always behind. Crude analog control loops like from your 1950s electronics 101 manual are used until today to “program” them, which is nothing but adjusting potentiometers. There simply is no comparison between the two systems, rate based APs are cheap entry level solutions. Already in the 1970s, airplanes like C210 and similar were shipped with attitude-based APs and so were all the twins.
Only with the downfall of GA, one vendor somehow managed to get STCs for hundreds of airframes (some people say that involved illegal business practises and has resulted in changes at the relevant FAA office). For lack of sensible alternatives, those APs got fitted into airframes they really should never have gone into.
Of course it hurts having spent $30k on what really is a technological POS but nothing is going to change that I went through all those iterations and spent the value of a nice car on AP upgrades. At one point I was intimately familiar with the logic board of a very widespread AP of a well known company from Mineral Springs, TX.