Tpoint adjustment question

[Paul Haese]

Using SkyX, and Tpoint add on. Did several runs last night on PA. First did a 20 point run and then did the adjustment. Next did 50 points and clicked on supermodel and adjustment said raise the alt 1.2ticks and Az loosen left and tighten right by 3.3 ticks.
Then do another model 30 points and use the super model function again and I get exactly the same requirement. The Alt does not need to be moved though.

The first run I had to move the AZ by 7.7 ticks and the alt by 3.6 ticks lower.

Why would I get the same result again for the second and third runs? Super model says I am pointing at 40 arc seconds but I would like this around 15-20. Not sure if that is achievable with the RC12.

[PRejto (Peter)]

Originally Posted by Paul Haese

Using SkyX, and Tpoint add on. Did several runs last night on PA. First did a 20 point run and then did the adjustment. Next did 50 points and clicked on supermodel and adjustment said raise the alt 1.2ticks and Az loosen left and tighten right by 3.3 ticks.
Then do another model 30 points and use the super model function again and I get exactly the same requirement. The Alt does not need to be moved though.

The first run I had to move the AZ by 7.7 ticks and the alt by 3.6 ticks lower.

Why would I get the same result again for the second and third runs? Super model says I am pointing at 40 arc seconds but I would like this around 15-20. Not sure if that is achievable with the RC12.

Paul,

We are typing at the same time. Please see my answer in your other thread.

As for this question, you really need 50 points to be accurate for the super model. 20 points isn't enough. That may be why you are seeing the same result....or, when you moved azi a small amount it actually may not have moved! That is why I now move (when close) photographically to confirm the actual move and the amount of the move.

T-Point is not iterative. i.e.,if you have enough samples (50+), and you move the mount exactly as indicated in the report, you can be done in one go. But if you cannot be totally sure of the physical move you will need to repeat to confirm. This is where it gets frustrating and slow. If you do it photographically you can be extremely close in one move.

Peter

[Paul Haese]

Originally Posted by PRejto

Paul,


T-Point is not iterative. i.e.,if you have enough samples (50+), and you move the mount exactly as indicated in the report, you can be done in one go. But if you cannot be totally sure of the physical move you will need to repeat to confirm. This is where it gets frustrating and slow. If you do it photographically you can be extremely close in one move.

Peter

Yes I have experienced this before with Tpoint for windows and Sky6. In fact it can go from one side to the other really easily before you get PA right. I had thought that this might have been resolved in the latest version and there was something I was missing.

The manual seems to indicate anything from 30-50 sample points works well for PA adjustment. Though over 20 is what is required.

I might use that photographic method, but had hoped Tpoint modelling would work easier. For now it is a good pointing to PA, and maybe tonight I will do a 300 point run if the sky clears. I will check after 50 points though as you suggest.

Thanks Peter.

[PRejto (Peter)]

Originally Posted by Paul Haese

Yes I have experienced this before with Tpoint for windows and Sky6. In fact it can go from one side to the other really easily before you get PA right. I had thought that this might have been resolved in the latest version and there was something I was missing.

The manual seems to indicate anything from 30-50 sample points works well for PA adjustment. Though over 20 is what is required.

I might use that photographic method, but had hoped Tpoint modelling would work easier. For now it is a good pointing to PA, and maybe tonight I will do a 300 point run if the sky clears. I will check after 50 points though as you suggest.

Thanks Peter.

Hi Paul,

Well, yes, there are real differences between what the manual says and what gets discussed at length on the forum at SB. The latest advice re T-Point is if you want an accurate PA report you need 50+ samples. But, the real issue I think is that making small moves, like 1 tic, or .5 tics, is quite difficult even with these calibrated Paramounts - especially in azi. I think T-Point works well and is very accurate but moving the mount isn't.

Peter

PS You might know this already, but it was kind of (good) news to me. If you do a very large model, but then subsequently need to redo PA, you can save the model by recalibrating it as a "portable" mount. The recalibration just figures out new terms for MA and ME but preserves all the other terms of the large model. The confusing part of this is that the results after recalibration look worse in terms of RMS pointing accuracy. But, a very recent discussion at SB assures that behind the scenes the original model is still doing the job.