But you can't usefully fly IFR without a heading... so why get airborne with such a defect?
I had a heading (2 actually) but the two AHARs differed such that it required me to switch over to one exclusively before the autopilot would operate. It is common to see differences between the two computers. It wasn't a defect just a system design issue. So had to hand fly until I had time to swap the system. In level flight it realigned and was back to normal.
I had a heading (2 actually) but the two AHARs differed such that it required me to switch over to one exclusively before the autopilot would operate. It is common to see differences between the two computers. It wasn't a defect just a system design issue
This is off topic but I do think you should get it looked at.
A plane or a subsystem would never get certified with such behaviour.
An AHRS system cannot generate a heading. You have a fluxgate magnetometer which generates a heading from the magnetic field, and an AHRS directional gyro (DG) which stabilises the heading value and delivers it in various forms (ARINC429, XYZ, stepper maybe) to other avionics. Normally one has to wait, without moving the plane, for say 2 minutes, for the DG to power up and stabilise. If one doesn't wait, one never gets a valid heading displayed - until the aircraft happens to find itself in a stable orientation sometime later for the required 2 minutes.
If OTOH you simply did not wait for some specified time then that is OK, but personally I wouldn't do it because if that system failed you end up with no heading and loads of avionics functions disabled.
The ~2 min wait is one change I now have since the SG102 install. With the old KG102A one could start up and taxi right away; it would erect in no time at all. Due to this, Sandel recommend connecting the SG102 directly to the aircraft bus, before the avionics master, but one is then relying on the text in its IM saying it will withstand the starter motor transients 
I will monitor it and of course get it checked. Headings on PFDs are often different however. Happened with the Avidyne and G1000. I always wait until stabilised but the g1000 takes far less time than the Avidyne used to.
I find it very interesting how different pilots approach different risks.
Some of the things I am averse to:
Some of the pilots I fly with scoff at my aversion to those things, but then raise their eyebrows at my willingness to:
Really interesting how everyone has their own ideas about what constitutes risk.
I think a lot of it comes down to one's view of the risk of an engine failure at just the wrong moment.
Low flight over the sea is a bit pointless... but west of Italy you may have no choice
Short runways, yes, one needs to know one's aircraft performance etc
Congested airspace... risk of a mid-air enroute is very low
Flying over some mountainous areas is certainly a real risk with no escape, but what % of the airborne time is spent doing that, against a ~50000hr MTBF of a properly maintained and operated Lyco engine, assuming there is fuel in the tanks?
Same goes for approaches to many airports. Not naming any in case some twat from the Daily Mail is reading this.
Same goes for forests - plenty of those in Germany. I would not fly at say 1000ft over a huge forest. FL150 is a lot better.
One thing which I find of real concern, now that I have TCAS, is "ghost traffic" near airports, either stuff passing through non-radio, or people arriving but making false position reports in order to get an earlier clearance.
Previously I used to get the odd airprox with somebody flying through the ILS, etc. Once, at Shoreham, I got some twin flying through the 20 IAP inbound track at just the right height. Now I see a lot more of this kind of thing. The other day I was landing, and about 1nm out, on final, I had a return 200ft below, 12 o'clock, and about 0.5nm ahead so clearly a risk. Not seeing anything ahead I went around. No use spending £12k on some box which is telling you that you are about to hit something and then ignoring it. It was a 100% real return, though the azimuth could be up to ~ 20 degrees out. No known traffic... It might have been somebody landing non-radio or it might have been a helicopter hovering on the final approach path (had that before, at night!).
The risk of mid-airs is obviously much bigger near airports.
Of course appreciate the risk of an en-route mid-air is small, but finding oneself in such a bottleneck, having multiple contacts called to you and trying to spot them, is stressful and not much fun - hence why I try to avoid it.
Some pilots I fly with seem to have an astonishing "you're probably dead anyway so what does the height matter" attitude to flying SEP over the sea.
I just can't understand this. Even if you're not high enough to glide to land, the higher you are the longer you have to get yourself and your passengers prepared for ditching, make sure the raft is in the right place, secure the cabin, etc. You also have a lot better chance of being able to put down next to a ship/yacht, your MAYDAY call (which you have more time to make) is more likely to be heard by multiple stations, and you have more time to give a 100% accurate position report as well as your height and heading so that SAR units know where to look for you.
Ditching is not an attractive prospect, but I will do everything I possibly can (short of flying in an immersion suit) to maximise the chances of surviving.
I'm very willing to cancel a flight, and often do. It happens that the aeroplane I do most long trips in is IFR equipped, but without anti-icing, so likely icing conditions is a common reason, particularly in the winter.
I only fly about 100hrs pa, and only perhaps a third of that is IFR, so I also tend to make a judgement about my currency and competence on some trips.
I do some quite specific flying (testing, and so-on) so again, may need very specific weather conditions.
The other issue is "fun". If I'm flying for myself rather than to get somewhere at a particular time, or I can work where I am for half a day until the weather improves and still get everything done - why on earth shouldn't I cancel if it's not going to be an enjoyable flight.
G
Out of interest, what would the view on maintenance risk be? I have a near zero tolerance to not take an issue into the air other pilots I know seem less bothered OR if engineer says its ok then it seems to give some pilots comfort. For example, if on start up one of your mags failed, would you launch? If engineer, who sits miles away says its fine, would you launch? Would you launch with passengers? What would the chances of the 2nd mag failing if you have no idea why first failed? I have scrubbed a flight when one mag went rough so was I being silly? I use this as an example but there are lots of scenarios one could mention. Would be interested to hear others think about this. Ta.
The key to taking maintenance risks is knowledge about the aircraft. Unless you have a deep understanding of aircraft systems, you shouldn't take risk.
For example, if on start up one of your mags failed, would you launch?
Absolutely not, an aircraft with one mag not working is not airworthy.
I've flown my aircraft with issues. Once the starter adapter was broken but I had an introduction in how to handprop it. Another time the alternator was broken but I knew I had the battery and if I didn't use the avionics, it would be more than sufficient for operating the flaps (and even if it didn't I had trained no flap landings).
If the aircraft is not airworthy because of the mag issue, I presume the knock on would be insurance not paying out should something go wrong?
