Drift alignment - 1st time

[malc-c]

Hi Chris, and thanks for your input.

I was under the impression that every time you used the sync option and confirmed that the scope was indeed pointing at the target star, that EQMod added that point to the list and thus the more points you add the more accuracy you got as EQmod some how could calculate the small errors and make any required adjustments.

What would be the best way in EQmod to replicate the way the handset performs a 3point alignment so I can compare like for like.

[chrisshillito]

If you are in the append on sync mode then yes, every time you sync you will add a new point to the list. How EQASCOM then uses those points will depend on the number of points in the list and the "alignment behavior" setting.

If one point is in the list then all corrections will be made based upon the measured error of that one point.

If two points are in the list then corrections are made based upon the measured error of the point closest to your target.

If three or more points are in the list and the alignment behaviour setting is set to nearest point then a search is made through the list to find the point nearest to your target (in terms of RA/DEC axis movement) and corrections are made based upon the measured error of that point.

If three or more points are in the list and the alignment behaviour setting is set to N-Point + nearest point then things start to get complicated! First the points list is searched to find the 50 nearest points to your target. Then all possible combination of triangles are evaluated to find the triangle that has a centre point closest to your target. If a suitable triangle is found then the pointing correction transformation is made that takes into account the data from all three points forming the triangle. If your target falls outside of all possible triangles then a nearest point transformation is performed (as above).

In addition to thee above you can force the goto transformations to only consider points that are on the same side of the meridian, or in the same quadrant, as your target.

It may be a reasonable assumption that 3-Star alignment is equivalent to using an EQASCOM N-point+Nearest transformation with three points. However, I really can't say this for sure as I've absolutely no idea what algorithms the handcontroller uses when providing its goto corrections (just because data from three points has been entered doesn't mean it is always using it.). Also there is a quite a fundamental difference in alignment strategies between EQASCOM and the Synscan. The handcontroller is quite restrictive in it's choice of suitable alignment stars and I suspect the approach usually taken is to cover the entire sky with just three points. With EQASCOM you are not restricted in your choice of points and you can enclose your targets within a much tighter alignment triangle if you wish. An "all sky model" in EQASCOM is not restricted to 3-points and most will go for six (three each side of the meridian) if going down that route.

I guess what I'm saying is EQASCOM isn't intended as an SYNSCAN copy it has its own algorithms and workflows etc. An alignment strategy that works for the synscan may not represent the optimum strategy for EQASCOM (and vice verca).

If you want to know a little more about EQASCOM alignment take a look at this document http://eq-mod.sourceforge.net/docs/eqmod_alignment_models.pdf

Chris

[malc-c]

Thanks Chris... I'll check my settings the next chance I get

[malc-c]

Well somthing is really screwy when running the mount from the PC with EQMOD and the EQDIR cable built from the EQ website. I've just taken some test shots of M31. The longest exposure I got before lots of trailing was just 80 seconds. 60 seconds was the longest time I could expose for without any drifting visible in the image.

I then removed the EQ lead and connected the handset, repositioned for the slight drift whilst doing so and the proceeded to do the same test, increasing the exposure until I got similar trails... I got to 300 seconds before I had noticable drift, and went a full 12 minutes before I could replicate the same amount of trailing. I'll post up images later this morning when I've had a chance to resize them for the forum.

[malc-c]

Well after retiring at 3:30am this morning (didn't want to, it was such a nice night) and had a few hours sleep, here are the two images. There is a slight orientation difference as the camera was moved at the end of the evening when I used the handset, but you can see the difference.