RobertL18C wrote:
It had five attitude indicators, not sure if it met @Mooney_Driver level of redundancy but it was a nice aircraft.
Most probably 
It is not so important how many actual indicators but what source they use. If they all have the same source, the number is useless. It needs to be 3 indicators with 3 sources. Either 3 indepentent batteries, or vaccum or whatever.
I find the number 3 a good number, because it is very rare that more than one fail at one time. Therefore, you have 2 which show the same and one which is off. If you have two, it will take analysis and time to figure out which one is ok.
Malibuflyer wrote:
Related to Autopilots, I’m not a ware of a single “steam gauge” GA autopilot that can switch between two independent attitude references.
True, neither can the digital ones to my knowledge. AP’s however are not as critical as attitude and other basic instruments.
Malibuflyer wrote:
For all other instruments the “digital” solution is typically more reliable in the first place, an error is typically easier to identify (one of biggest risk with “steam gauges” is that you follow an indication that looks right but is wrong) and redundancy is typically cheaper to achieve.
No question. And I don’t really advocate throwing it all out, only to have “Steam” backups. Also, it is quite important that manufacturers reckognize those problems and act on them. Aspen for instance did: Loss of airspeed does no longer take away attitude on the Max series. I am not familiar with the others popular EFIS enough to know what they do and what not.
E.g. I do have the Aspen but still the analogue speed indicator. Both however hang on the same pitot, so both would fail if e.g. the pitot ices up. It would be nice to get a 2nd pitot, but that is quite an installation, but the critical thing there is the Aspen, which in it’s original version dumps everything if IAS fails. My answer to that will be to upgrade it to Max the moment I can afford it.
Of all the data we need to keep an airplane flying in IMC or night, Attitude is the most vital Therefore, it is for me the most vital information for which I want backups. The solutions to this are quite open, be it vaccum, be it a portable Dynon D1,2,3 or similar or even a AHARS displaying on an Ipad can in such a case be the third source.
The other bit is that installations with protections should not fight the pilots with false data. That goes for airliners as much (or worse) than for GA. I recall that there was a Vision jet which encountered similar problems with AOA as did the 737Max, but it was easily controllable. AF447 was mentioned above, the epitome of a cluster….. where a crew was overwhelmed by a “confused” cockpit. Learmouth (A333) come to mind. also QF32 to an extent.
Autopilot is just as important. Do you want to hand fly for 20hrs like I did years ago?
Also having several AIs doesn’t help if departing in OVC002 and the main one fails just after rotation. You won’t spot the discrepancy before you die.
Does anyone know the specifics here?
Peter wrote:
Autopilot is just as important. Do you want to hand fly for 20hrs like I did years ago?
It is not what I want to do, it is what is needed to survive. An Attitude indicator failure at low altitude is very often deadly. AP failure is a nuissance, but not really deadly normally. No, I don’t want to handfly anymore, but I’ve done it many times.
Peter wrote:
Also having several AIs doesn’t help if departing in OVC002 and the main one fails just after rotation. You won’t spot the discrepancy before you die.
That depends how they are arranged. If they are in your primary field of vision you should be able to check them all. It also depends which one goes pear-shaped. And of course you will always find a situation in which almost nothing really helps, but if you have a single ADI failure in flight, the comparison with the other two will help.If your primary EADI fails, it will go XX or blank, which is an advantage at this stage. In this situation, you need to have the 2ndary right in your field of vision. Most of the time 2 ADI’s will do the job here.
A clear case for a third one: Let’s say you are flying at night, normal cruise, and your primary fails gradually, starting with a small bank. You correct, but it won’t follow. You have one more, but due to spatial disorientation, you need time to figure which one is right, so a third one, which can verify that, is a big help. ADI failures at night without backups have in the past cost a lot of lives. Particularly but not specifically in the military. There have been cases even in airliners, where crews followed the wrong ADI into hell, because they did not compare the two others properly. So I would suggest, for GA they should be pretty close together, unless you fly dual crew, where it is very useful to have one in the copilots side.
Mooney_Driver wrote:
If your primary EADI fails, it will go XX or blank
Don’t be so sure. I’ve written previously about a crash in Sweden some years ago, where the EADI suddenly began showing a steady pitch-up rate of about 4°/second until it indicated 90° pitch up, without any relation to the actual aircraft attitude. The pilot flying responded by diving the aircraft into the ground.
LeSving wrote:
How exactly did that happen?
This is Russia and therefore the rumour mill is doing a couple of thousand RPM right now.
However, Rosavia has made some statements which give some idea what happened. Avherald states (my summary)
The airplane was de iced in 2 stages, first with Type one and then Type IV. However, only the wings and stabilizer were de-iced. The fusellage was not. After the windshield heat was turned on, the ice on the windshield melted and ran down the sides of the airplane. During taxi, it froze and impaired all 3 pitot and static ports to different degrees. So the airplane took off with large amounts of snow and ice on the fusellage.
If they are in your primary field of vision you should be able to check them all.
Nobody is going to install three AIs in the primary field of vision – there is no room.
Don’t be so sure. I’ve written previously about a crash in Sweden some years ago, where the EADI suddenly began showing a steady pitch-up rate of about 4°/second until it indicated 90° pitch up, without any relation to the actual aircraft attitude.
Completely possible. I wonder how they discovered it though? Was there a camera in the cockpit?
The first post in this thread is pretty scary too…
And an answer to my Garmin question above seems to be here. Not very good…
Peter wrote:
Completely possible. I wonder how they discovered it though? Was there a camera in the cockpit?
There was a FDR and a CVR. The actual pitch was computed using airspeed, ROD, AoA etc. recordings.
Not GA then, presumably. So they would have had multiple attitude references…
Airborne_Again wrote:
Don’t be so sure. I’ve written previously about a crash in Sweden some years ago, where the EADI suddenly began showing a steady pitch-up rate of about 4°/second until it indicated 90° pitch up
Yes, it happens. I had a mid-continent SAM that started to tilt in long descents (like when going down from FL220 to 3000ft at 1000ft/min you ended up with an indication of of 30 deg left bank).
I’d still argue that the undetected faults are order of magnitude more likely with mechanical instruments than with solid state ones.
Mooney_Driver wrote:
Full article here
Trivia: The remarks sections of some METARs include the QBB code which I’ve never seen before. According to Wikipedia it means the cloud height. It was used in METARs giving vertical visibility instead of cloud so it makes sense.
