Sub Pixel Accuracy?

[Stub Mandrel]

As a rule of thumb, it seems that for imaging with a decent scope and a typical setup is ~ 1 arc-second per pixel.

I was wondering what sort of step resolution you need to achieve sub-pixel guiding at such a resolution. Is it enough to be roughly 1:1, or do you need the resolution of the drive train to be significantly better?

Whjat sort of micro-step resolution does something like an HEQ5 have?

I doubt that I'm working at the level where-subpixel guiding is significant yet, but the goto setup I'm making will have 0.75 arc-second resolution for the microsteps and 2.45 for declination. I'm aware that microstepping is not perfectly even, but I assume that's one of the things guiding addresses!

[Pompey Monkey]

In the UK, a good pixel scale to aim for would be approximately 2" per pixel on the imaging train. This is small enough to not get square stars, but big enough to not waste pixels due to scintillation in the average seeing conditions.

For the guiding set-up, the star centroid location accuracy is regularly in the order of 1/8th or 1/10th of a pixel on the guide camera. Hence the guiding system can have a fl of 1/4* of the imaging train and still achieve guiding errors of only 1/2 a pixel in the imaging train, assuming that both the guide and imaging cameras have the same size pixels.

*There is a good argument for even shorter focal lengths in the guiding system: http://www.sbig.com/site/assets/files/1359/autoguidingversion3_mb.pdf

 

 

[Stub Mandrel]

7 minutes ago, Pompey Monkey said:

In the UK, a good pixel scale to aim for would be approximately 2" per pixel on the imaging train. This is small enough to not get square stars, but big enough to not waste pixels due to scintillation in the average seeing conditions.

For the guiding set-up, the star centroid location accuracy is regularly in the order of 1/8th or 1/10th of a pixel on the guide camera. Hence the guiding system can have a fl of 1/4* of the imaging train and still achieve guiding errors of only 1/2 a pixel in the imaging train, assuming that both the guide and imaging cameras have the same size pixels.

That still leaves the question 'what resolution should the mount drive have' unanswered, though.

I understand enough about sampling theory that having guider accuracy and precisions that is greater than the precision of the mount can still have advantages (simply because it will make its more limited moves at the 'optimal' moments) but i am interested in what 'poiting resolution' typical systems have in practice as this has implications for whether or not I need to build a new declination gearbox!

[Pompey Monkey]

42 minutes ago, Stub Mandrel said:

That still leaves the question 'what resolution should the mount drive have' unanswered, though.

I understand enough about sampling theory that having guider accuracy and precisions that is greater than the precision of the mount can still have advantages (simply because it will make its more limited moves at the 'optimal' moments) but i am interested in what 'poiting resolution' typical systems have in practice as this has implications for whether or not I need to build a new declination gearbox!

Do you mean for the "GoTo" function?

http://www.firstlightoptics.com/user/AZ_EQ6-GT_instruction_manual.pdf Page 19.

The GoTo functions of such mounts are absolute and, therefore, only as accurate as the encoders or stepper motor step.

For my mount (as in the manual) to get better accuracy, plate-solving is used. This is a closed loop system and will regularly get to within 10-15" of a specified point after two or three iterations.

Other mounts may use other techniquies, but close loop, negative feedback, is a tried and tested method.

 

For the guiding correction pulses, there seems to be a lot more going on in my mount/PHD combo than just a pre-defined accuracy - again it is a closed loop system. Hopefully a proper SME will be along at some ponit to elaborate

 

[Stub Mandrel]

36 minutes ago, Pompey Monkey said:

Do you mean for the "GoTo" function?

http://www.firstlightoptics.com/user/AZ_EQ6-GT_instruction_manual.pdf Page 19.

The GoTo functions of such mounts are absolute and, therefore, only as accurate as the encoders or stepper motor step.

No - I want to know that basic reolution - how big the 'motor steps' are on the sort of mounts most folk are using, such as HEQ5.

<edit> that manual you linked tells me for the EQ6 its 0.14 arc-seconds, but Skywatcher don't seem to want to tell me the resolution of the EQ5.