50 NM for a typical tourer such as a TB20.
This shows why one would not want to end up in such a situation.
“ Even in Germany there are places where you’ll likely to be more than 35 NM away.”
It depends if you are looking for direct then 3nm final? or want to fly a fully procedural with few holds, 20nm dme arcs and full reversal turns with 10nm FAF? if you need the latter then yes you will need more than 30min
I am sure when flying one has to plan for en-route diversions and his options to go down on every point of the route for X reason, I am sure lot of emergencies in C172 require you to be on the ground in less than 30min (fire, icing, engine failure, no fuel…), electrical failure in IMC may not be one of them but you should have an answer to that before jumping in the aircraft?
In a FADEC DA40, I will probably be dead or stretching my legs in nearby airport or field 30min after I pressed the backup battery
It’s easy to select times & routes that minimise IMC exposure during planing & actual flying to less than 30min, one is usually aware before departing
Ibra wrote:
It’s easy to select times & routes that minimise IMC exposure during planing & actual flying to less than 30min, one is usually aware before departing
Sure it is easy if you make it a no-go item. Otherwise not so much.
Antonio wrote:
you can ask the late Pete Conrad and his crew….
Or try Al Bean who actually found the switch when it was called for in mid launch. He’s still alive.
Graham wrote:
However if you already have a vacuum system then adding one piece of electronics to it seems a better (not to mention cheaper) option than pulling out the vacuum system and adding 2x electronics.
I concur with this approach. The fault tree for cases of electrical faults, lighning , SW fautls, etc is so much better than purely electronic or even electric, no matter how independent your electrical systems are….
Graham wrote:
My point about the vacuum system is really that it’s a known quantity. It can fail, yes, but it only fails in one way
The system can fail in more than one way – the vacuum can continue to suck but the gyro fail, for instance. I’ve had that happen – the AI keeled over on an IFR flight in a C182, without warning, after working perfectly fine for 2.5 hours of the flight. I put one of those partial panel simulation rubber suckers on it so I wouldn’t look at it.
The idea is to have multiple independent systems, however they are powered. An independent electronic system is likely to be a lot more reliable than an independent mechanical system.
The big difference between
is that after a power failure you know that with the latter you have only x hours before your “time is up”, whereas with the former the chances of losing it are just the % chance of a random failure.
And I would bet the chance of a random failure of a piece of electronics is not clearly arguably better! This thread itself is a proof of that. Especially if you rip out the mechanical system and replace it with two identical electronic ones, whose installations, market experience, debugging, etc, are, shall we say, south of negligible!
I had no failures of the old KG102A gyro, while I’ve had two failures of the AHRS. I had to do one 600nm IFR flight with one of the latter, with the autopilot working purely off the (vacuum!) KI256, and from the KEA130A encoding altimeter, no HDG mode, only ROL + ALT modes, and me “flying a GPS track” for 3.5hrs with the rudder pedals 
On top of that, all GI275 installations will still be in warranty so chances of forum reports of problems are very low because everybody has to look after their dealer relationship! Of course some disagree. I bet you the only reason this one came to light is because the pilot was very nearly killed.
To be clear: I am not saying that using a vacuum pump to drive a gyro isn’t a crap engineering approach (if you can work out the double negatives) What I am saying is that
I’d say one thing for electronic devices in GA:
If they fail, they fail. Fully. Either blank screen, flag, or X out or something. At least the ones we are talking about here, Aspen, Garmin, Dynon, e.t.c.
Electromechanical and Vaccum ones are more ambiguous. They may fail fully or they may start to give false indications without a flag. And that is quite dangerous at times.
Hence the idea of 3 independent sources. That two of them fail similarily is quite difficult to achieve. And it is not a very expensive exercise either, there are plenty of options, be it an AHRS on an ipad, Dynon’s D1,2,3 series, e.t.c. all work quite ok for this purpose as the third AI.
