I cannot understand the fixation on this, when all single engined aircraft rely on one engine which has one vacuum pump on it which is so unreliable it only lasts a few hundred hours. OK it’s probably backed up by an electric instrument of some sort, but there have been plenty of crashes due to failure to recognise vacuum instrument failure, and the single bus electrical systems are not exactly robust, with their antique voltage regulators etc.. All gyro instruments either electric or vacuum are in my opinion flaky. I’ll take the glass alternatives any day.
The failure modes of glass are a lot more complex than the old steam stuff. I have had a Garmin GPS antenna go crazy and jam everything else (including VHF and transponder) and I have had my Aspen disable the AI in flight. The CDI was gone as well but I have a separate “mechanical” CDI. The C210 video referenced above did not have a separate CDI. Had it been difficult weather and had they not been able to fix the Aspen issue, they would have had a serious problem.
A vacuum driven AI and an electric AI, both mechanical, are an extremely reliable combination with no single point of failure that they both have in common. That’s the gold standard and my concern is that glass systems don’t provide that level of safety.
Don’t forget that vacuum pumps rarely fail without warning. All current ones have an inspection window to see the vanes. They don’t last very long but they are certainly not unreliable.
alioth wrote:
at least suggests to me it’ll give a “good enough” indication to keep flying should you lose both GPS and air data at the same time.
That can’t be true for more than a few minutes in level flight. Without an erection mechanism, a solid state AI cannot display correct information for more than a short while. However, it suggests that even the G5 uses both GPS and air data for erection. Seems the Garmin implementation is superior to the Aspen implementation.
This article seems to agree with me.
There are fatalities every year due to the failure of mechanical gyros, particularly suction based instruments, often due to the loss of the suction that powers them.
I believe that to be an order of magnitude more likely to happen than a convoluted sequence of failures in modern instruments.
I am with Neil, I think the obsession with glass failures is just misplaced nostalgia.
A vacuum driven AI and an electric AI, both mechanical, are an extremely reliable combination with no single point of failure
In fact an AI alone is not suffcient for reliable IMC attitude control. As it works with gravity erection it needs a second instrument and a pilot or autopilot using that second instrument to avoid long term curved flying. An electronic AHRS could also be build with gravity erection but that would limit its use to the same flight profile. In many industry applications the movement profile is different so engineers probably choose magnetic erection as it is much more versatile.
Looking at specifics, I don’t see a first order difference between
Both would involve very bad luck, for sure.
I am sure the first would be easier to spot – because you get a huge great cross right across the screen. But knowing it won’t help you fly in IMC. You still need an independent AI. Big jets have much more redundancy but they all still have a little AI and if that is INOP, they cannot fly (my A&P works on Airbuses).
In fact an AI alone is not suffcient for reliable IMC attitude control
It is enough if you have a compass, or some other heading representation. Then you can keep wings level in the long term. And if you have altitude then you can maintain pitch in the long term. Which, unsurprisingly, is exactly the airdata inputs these boxes use for background erection (in the absence of GPS data)… but they also seem to need airspeed (pitot) which a human pilot doesn’t need (because he knows attitude+power → speed) an in piston GA you can’t ever end up going too fast if the heading is relatively stable and the altitude is relatively stable.
A vac or electric AI will give you reasonable long term stability unless you fool it by flying a coordinated turn, which is impossible in any long term without the compass going round. Certainly, a KFC225 etc fed from a KI256, in ROL mode, will fly a curved path, which is why one has the HDG or NAV modes. Even HDG will fly a curved path, usually.
I am sure that if everybody had two alternators, nobody would use vacuum pumps, but most don’t have the option.
Neil wrote:
I believe that to be an order of magnitude more likely to happen than a convoluted sequence of failures in modern instruments.
Me too. The problem is that once you (or the autopilot!) become aware that your attitude is totally wrong, the usual means of backup that most of us were trained to/instruct to use, namely turn&slip and altimeter/VSI will be so far off bounds that it becomes impossible to re-establish level flight before the aircraft disintegrates.
Personally I also much prefer all-electric avionics powered by different sources. Whether the standby AI is a “simple” electric gyro or a little glass cockpit of it’s own (like a Meggitt) makes no difference for me. In our school’s IFR trainers we have the opposite situation now: The main attitude instrument is a G500 installation with the old vacuum driven AI as a backup. A cheap and very reliable solution.
And BTW: I did some googling and found many solid state AHRS systems which do not require GPS input in order to provide attitude information. Most of them are not really cheap but not overly expensive either. They will probably not be able to provide accurate attitudes after a Red-Arrows display, but for flying home straight and level they should be good enough.
I did some googling and found many solid state AHRS systems which do not require GPS input in order to provide attitude information
My SG102 is one of these. However AFAIK these cannot get certified as a primary pitch/roll source, and it is possible that they also cannot be used as a backup to a glass cockpit like the G5 cannot. Does anyone know the reason for the G5 restriction?
I think, returning to the OP, this is a valid concern, no matter how unlikely it may seem. Does the TBM have dual pitots, fed from separate generators?
achimha wrote:
Now consider GPS jamming (…) I believe you would lose your G1000/G3000 and also your glass backup display
The backup instrument in TBM930 is an MD302
I don’t think it uses GPS signal.
Once I had a failure of G3000. Both PFDs lost GPS position and heading. (It happened in Russia. Pure coincidence :-) The standby instrument continued to work, as did the magnetic compass (it is located above autopilot panel, not visible on the picture you posted).
Peter wrote:
Does the TBM have dual pitots, fed from separate generators?
I had a look at the G1000 failure modes in the post from Dave_Phillips and it does not tell the whole story. If the magnetometer output is intermittent or differs significantly from the GPS direction then you lose the DI and attitude completely – apparently the G1000 decides it cannot work out whether the GPS or the AHRS is correct so it gives you neither. It also takes out the localizer and glideslope display on the PFD (and MFD, if switched to reversionary mode). So a magnetometer failure is a single point of failure (singe engine G1000). All you get is a big red X on the PFD.
I know this for a fact as I was in solid IMC, doing an ILS approach when exactly this happened. I had to use the GPS moving map, compass, standby AI (separately driven, electrical), standby altimeter to get down – see pic taken on the ground.
