Almost 9 years behind a G1000 and no failures of any of its units themselves.
Did have some issues with external sensors though:
1. Traffic unit (TAS 600) is starting to show 180 degrees errors. Jesse, I may come and see you this summer..
2. A few years ago had a clearly erroneous and sometimes disappearing CDI indication. That was quite interesting because the cause was that one of the sensors that feeds it (the magnetometer) gave a wrong signal after starting up the aircraft while the wingtip (near where the magnetometer lives) was near a massive hangar door. Took a while to figure that one out.. Just to make sure we reproduced this effect on another DA42. Luckily there is a procedure to reset the magnetometer which the pilot can do himself. I now take care not to power up when there is a large chunk of metal within 25 m of my starboard wingtip!
I have had an instance where the map part on both the PFD and MFD in the Cirrus SR22T turned black in flight and while about to intercept the ILS for EHRD.
The Entegra is much easier if you know the 430s. However I do think it has more issues with interaction between different components. It is also less powerful but a very good system nonetheless.
Jason,
do you also share my view of the Avidyne Entegra? I felt pretty good for IFR after ten hours (having known the 430s for 12 years).
On The G1000, I now have nearly 1000 hours with G1000 all IFR and there are still times when it does something I don’t expect. Now that is almost always because of something I did wrong. But it shows that these are complex pieces of software.
I had an Airbus airline pilot in my Mustang type rating. He was an excellent pilot but had never used G1000. Using G1000 for serious IFR after about 25 sim and PTD hours he was just OK. Had he had some real pilot induced mode confusion in his checkride, I was not certain he could have worked his way out of it.
10 hours sure you can read the speed and altitude and set the transponder. But flying serious IFR in IMC I don’t think so. Not because any individual process is that hard but there are an awful lot of them.
And all of the above is in normal ops.
@RobertL18C : beautiful post.
Quoted from US Flyer on the North Sea SR20 thread
But if you can work an iPad then you can work a G1000.
The PadPilot crew have quite a good product for learning DA42 systems but suitably edited down. Nevertheless it is a good introduction to the ‘knobology’, (somewhat more appropriate than buttonology), of one of the leading GA glass cockpits. In particular some good common sense insights into managing the KAP140 autopilot interface.
Systems with more redundancy built in will typically have several hundred pages of description in section 7 of the AFM, plus the interface of the system with pressurisation, stick/shaker/pusher, aircraft configuration, warning systems, etc. The full Garmin 1000 manual may be several hundred pages.
Before reaching the cyan advisory items on Crew Alerting the warning and caution advisory elements alone, on a more advanced Honeywell system come to nearly fifty, a Shakespearean variation on how many ways might I fail you, and how when I fail you, you need to revert to manual on a host of systems. Ideally you would know what CB might be allowed a reset, only once, and whether it is a two second, five second or two minute re set. There are around a dozen and a half glass cockpit related CBs in various cockpit locations.
For a fixed gear, SEP, Garmin 1000/KAP140 or GFC700, may have less redundancy, and less systems interface, but still may have two display units, ADC, AHRS, AGM, GPS, TXPDR, DME, PFD and MFD Controllers, FMS, etc which can fail or overheat, with knock on effects on VNAV and LNAV, let alone the basic six pack read outs.
Unfortunately there are several relatively recent fatal accidents where the glass cockpit threw the toys out and the system, sometimes helpfully playing to crew and passengers Cavaleria Rusticana, disconnected close to the stall and the hapless iPad expert spent his last two minutes alive trying to reset his security blanket.
The bit of the AFM in bold saying step 1 : Aircraft Wheel Control GRASP FIRMLY regain aircraft control, perhaps, like a Shakespeare sonnet, left unread.
I did a training session on a simulator with a experienced instructor, going through different kinds of G1000 failures.
That was very useful. What I found worrying is that when the workload goes up, you won’t notice some failures.
I had a scenario where I could not get radio contact and I did not spot that COM1 was in a failed state. That was when handflying an approach.
Just like with instrument flying, you need to learn to actively scan for failures, and you need to be prepared on what to do when a failure pops up.
This is much neglected during PPL training, where hardly any time is spent on operating the G1000.
The SG102 Sandel AHRS (a superset of the KG102A and plug-compatible) supposedly won’t do an airborne restart but actually it does, in still air etc. Also it has a big capacitor inside which keeps it powered for something like 30 seconds, so if you shut down the avionics master during engine start, it will still be up when you turn it back on.
Whereas the old KG102A would restart anywhere so long as it was the right way up 
Being Apollo technology it would probably restart on the Moon…
It is necessary to understand these systems and their limitations, but the information is hard to find unless you are a total bloody anorak.
I’ve had one instance of losing the AHRS on a G1000 in flight. It restarted automatically and everything was back to normal after about a minute. I talked to an avionics guy about this and apparently it is a known problem with the software version the aircraft has. We’ll have an update soon…
Another time I lost the heading indication on the same G1000 during taxi. The flight was going to be VFR so the standby compass would be enough. The heading indication came back as I lined up on the runway..
