Mks5 I found when checking my gear this morning that the guide cam was a little loose in the guide scope - this wasn't the case earlier
Yep, that could cause the simultaneous RA and declination spikes. I.e., the scope is moving in some random direction that is not RA only and not declination only.
Something like this will also cause differential flexure. I.e., notice from your guide graph that they both eventually recovered. However, if you were imaging (or had done a plate solve before and after the spike), you will find that the main camera has not recovered. The guide recovered because the mount was autoguided to recenter the guide star; but that shift between the OTA and the guide scope had remained.
It is possible that the declination-only jumps will still re-appear after you fix this, i.e., you may have had two unrelated problems last night, and one is simply masking the other.
Usually, a loose guide scope will not cause a declination only shift, unless by some luck the guide scope is loose only in the declination direction.
I did a jump test last night and it had remarkably less of an effect than I thought it would, it barley registered on the guide trace. This was on soft ground though, when I have it on the patio the trace goes like an earthquake if I walk anywhere near it.
That's interesting, and may indicate that the problem is not from the tripod moving. The tripod moving will not likely cause your original declination-only jump. I only suggested jumping around the tripod when I saw the RA-and-declination coupling. We are at least getting rid of yet one more source of possible error.
When I've been using the PA I've generally been running it when I get to my target, hadn't realised that it needs to be done at the CE,
No, no, Polar alignment need not be done at the equator. Guide calibration is the one that needs to be done at near the equator.
Guide calibration tries to measure the relationship between the x-y axis of the camera and the RA-Declination movement of the mount. Both is direction and in amplitude. It really is trying to measure the transformation matrix.
You will notice from a star chart that when you are nearer to the pole, you can move the RA by a lot and yet the x-y position of the guide star would hardly move. It would be easy to demonstrate if we had a star at the pole (Polaris is close enough to demo this, though) ti use as a guide star and calibrate against -- PHD2 would have complained that the star has not moved enough during calibration :-).
The other thing that you will find is that when you calibrate guiding near the pole, those two read and blue vectors are no longer orthogonal.
PHD2 will account for it, but the accuracy is worse as you calibrate closer to the pole -- this is why the recommendation to calibrate at the equator, where vectors are equal and orthogonal.
I have to say I'm sure it's the calibration - it's just doing odd things with the same settings v1.9. After the four steps have run, W E N and S and its cleared the backlash and nudged around a bit, the green box and the two lines have always landed up back where they started, at the star the calibration started at.
Yeah, let's reset the debugging, and do the proper guide calibrations near the equator first after locking the guide scope down, to see where it leads us.
Chen
