A TC is a GA thing, dating from the old days when a cheap and simple instrument was needed, and when one used to actually have to fly Rate 1 turns. Today, flying Rate 1 accurately is a rather obsolete exercise – used in IR training to separate real men from goats, with obscure holding pattern procedures – and in any case one can get really close to R1 by flying a given roll angle on the AI. On a TB20, 23 degrees is as close as dammit.
“Real planes” (turboprops and jets) have a proper backup AI, albeit it tends to be a small one and – on a multi pilot plane – positioned in the centre and thus hard to use.
STEC use a TC rather than an AI because they are crude autopilots, designed to work in lots of airframes, with a large number of STCs obtained by sneaking them in under the wire in the good old days without proper stability testing and consequently a lot of installations are unstable when anywhere near the corners of the envelope. King did this stuff properly but paid the price in much smaller sales in their later years.
Out of interest – I understand the KAP140 autopilot installation in a DA42 has a TC installed behind the G1000 PFD for it, you just can’t see it!
Same setup can be found on older SR20/22 without dual ADAHRS. A “blind” TC is located behind the copilot’s panel. You can see it by peeking through the RHS sidestick rod hole.
MattL wrote:
I understand the KAP140 autopilot installation in a DA42 has a TC installed behind the G1000 PFD for it, you just can’t see it!
Not only the KAP140. There are also others, like if a G1000 is used with any S-Tec AP.
If I could find the space on any of those, I’d have it moved to somewhere I can see it, possibly directly next to the back up AI. In a DA42 this should not be a problem.
Peter wrote:
STEC use a TC rather than an AI because they are crude autopilots, designed to work in lots of airframes, with a large number of STCs obtained by sneaking them in under the wire in the good old days without proper stability testing and consequently a lot of installations are unstable when anywhere near the corners of the envelope.
Well, I’ve been using a STEC 55X for a few years and I am reasonably happy with it. You get the gist right when you say they were designed to fit just about any airplane. They became popular because almost every airplane is on their list and because they sold them as retrofits as well. King made the decision to focus primarily on the factory installed market, I am not sure they even offered it as retrofit.
So for a long time, Stec were the only way to upgrade existing airframes.
If I was trying to fill a hole with a TC or an AI I would use an AI every time.
If I was trying to fill a hole with a TC or an AI I would use an AI every time
One friend used to parrot that it would make his aircraft illegal for IFR, pretty sure he never saw EFIS + AI backup 
he sucks at partial panel flying on TC, I had to cover him while he was on the hood once…
OK; we know that mechanical instruments are not vulnerable to loss of airdata etc, so what about the topic of this thread: failure modes of avionics which are affected?
We know that Aspen EFD was totally trashed by loss of airspeed data, but it is claimed they addressed this by using GPS data, but does that retain autopilot functionality?
We also know that Avidyne and probably Garmin have/had a similar issue.
Does anyone know of situations applicable to other systems, where the loss of airdata suppresses pitch/roll display on a PFD and/or stops the autopilot working?
I believe this incident caused a new software version for the GI-275 to be released. As a GI-275 can drive a GFC-500 autopilot, I’m pretty sure it would also be affected.
