Peter, does your EHSI have an ADAHRS? Where does it get its heading from?
I would have thought that given the magnetic track, and given reasonably light winds aloft, one is looking at second order effects due to the wind or the TAS.
If you are prepared to guesstimate the wind, it becomes easy. Flying North or South:
GS x sin(Track-Hdg) = compass-error-drift + wind-drift
Or, using the clock rule:
GS x (Track-Hdg)/60 = compass-error-drift + wind-drift
My original point was that flying North and South gives two data points, which are better than one, but in principle flying twice as far North would be just as accurate :-)
does your EHSI have an ADAHRS? Where does it get its heading from?
From a KMT112, connected to a KG102A for heading stabilisation. One enters heading calibration values into the SN3500, and this can be done airborne.
If you are prepared to guesstimate the wind, it becomes easy. Flying North or South:
In reality one always can guess the wind, using TAS-GS and doing that on a couple of headings.
I just "feel" there must be a formula for extracting the current heading from the indicated heading, the TAS and the GS, having collected data from 3 headings previously.
There must be. The Avidyne computes a heading, TAS, wind vector and GS. It is just a triangle problem is it not?
All GA systems use a fluxgate magnetometer to get the magnetic heading, and use a directional gyro to stabilise that against the noise of the fluxgate and aircraft roll/pitch etc.
The KMT112+KG102A is an old (but very solid) implementation of that. The heading adjustments are done using the KA51B slaving accessory.
One can also use the Sandel MT102 with the SG102 AHRS directional gyro (which I will be installing) to do the same job. In that case the adjustments are done in the SN3500. The other systems like the G1000 use the same principles.
The point is that the heading comes from the above fluxgate system in a single-ended manner. Computing the heading does not involve GPS or airdata. It can't be done that way because there is obviously no way to get the heading from any data you have when just by flying in a straight line.
One might use the thus obtained heading to get the other stuff e.g. the wind.
What I am thinking is that there must be a simple way to (STEP 1) fly 3 uncalibrated headings (120 apart) and knowing the GS on each, and knowing the TAS on each, one can work out the wind aloft at that time. The headings will be uncalibrated at that point but assuming they are not grossly out (say within 20 degrees) it should be fine because these are all 2nd order effects.
Then, you have the wind, and you have the TAS and GS, and working out a correction for the actual heading from the GPS track would be STEP 2, which you would perform on each of the cardinals i.e. 4 times.
Then one could do STEP 1 again just to make sure the wind has not changed.
This should work within about 1 degree which would be damn good and a damn lot better than is possible on the ground, given electrical (alternator charging etc) and metallic-proximity effects on the compass heading.
Once the EHSI is calibrated, one would calibrate the liquid compass from that. Sacrilege, I know, but it would be accurate 
Definitely up for that :)
Once I've got my book back, I'll drop you a note. I'm not bored enough of life to derive the maths from basic principles myself; putting a test plan and test cards together is enough fun.
G
I do now think it definitely is possible. The North/South tracks idea threw me.
There probably are more efficient ways, but the following must work, for constant wind and modest wind correction angles:
1) Fly a magnetic East track using the GPS only. Record the GS, either from a readout or analysing GPS tracks later.
2) Repeat flying West.
3) The two airspeeds in 1 and 2 should be identical, and the East-West airspeeds should also be identical, (same wind correction angle), so the difference in the two Ground Speeds is twice the East-West wind speed. Let's say that's 'w'. Calculate the wind correction angle for a North/South flight in the usual way as 60w/TAS.
4) Fly North using the compass. The GPS track, minus the wind correction angle is the compass error.
5) Fly South using the compass. The GPS track-180, minus the wind correction angle is the compass error.
6) Average 4 and 5. That's it! (Though you will have to pay more attention to signs than I just did ;-) In theory either North or South alone should do, but doing both feels better.
If somebody has thought about it enough to include it in a book, they have undoubtedly thought about it more than I have, so do please share if you find it!
PS I quite enjoyed trying to do it from scratch :-(
Peter, I don't honestly understand what you are trying to achieve.
Peter, I don't honestly understand what you are trying to achieve.
An in-flight compass swing using GPS ground-track truth data, and some clever maths. Also an excuse to do something a bit different with an hour or two's flying?
G
OK you know your ground track. You know your TAS. Without a reliable heading, how do you know you your wind components.
