Paramount MyT - Day 2 - Polar Alignment, First try TPoint

Now, where the mount is connected, I wanted to start with doing some imaging. First, I needed to Polar Align the scope. Boy, am I used to the simple RAPAS alignment routine. Centering stars is tricky if you only have cameras and no red dot finder or such!

After trying to get at least a rough polar alignment done with the cameras I almost gave up and thought about mounting my red dot finder back on the scope. But then I remembered by DSLR connector that I have on top of the telescope - and its dovetail can be screwed into the red dot finder!!! So, I mounted it
<need picture>

Now, I could do a rough polar alignment. One thing I noticed is that my Azimuth alignment screws have a lot of backlash. Takes half a turn or so that I can change direction!!! :-(

In order to improve my polar alignment, I used good, old AlignMaster - with the red dot finder, it wasn't a problem at all. Yei!

From my posts on the PHD2 and Bisque forums I was guided towards the Ascom driver settings and that I need to enable PulseGuide and GetPositionFromMount.
<screenshot>

I enabled these, connected to PHD2 and tried to move the mount around - worked!!! Then I wanted to do run calibration. But every time the mount would go into the East direction. But then when PHD2 wanted to move the mount back West to the origin, it moved it orthogonal to the initial direction. When it then moved South, it moved almost exactly into the same direction as the East. And the final North movement didn't really do anything.
<screenshot of directions>

And after the calibration ran, PHD2 would warn that the RA and DEC movement look suspiciously non-orthogonal.
<error screenshot>

Asked about that again.

Next, I wanted to check how the mount works in SGPro. Connected everything and tried to run a sequence. But the center command didn't work. For some reasons, the mount never moved closer to the initial slew to the object
<screenshot>

Posted on the SGPro forum.

Finally, I wanted to try out TPoint. As it relies heavily on Plate Solving (which it calls ImageLink), I first tried that. When it works it's wickedly fast! But too often, it wouldn't work at all (doesn't find enough or any stars). I tried a number of things, but couldn't get it to work reliable. I even set the exposure time REALLY high, but still not. Now, it was really late again - time for bed.