Ideal focal length to guide

[Ken82]

Is there an ideal focal length to guide ? I mean if your guide scope has a much smaller focal length (predominantly ) then surely the guider can't detect the movement before the main imaging camera with a longer focal length ? Obviously not ! I appreciate guiding with a long focal length scope you would get "chasing the seeing " but potentially it's going to be within the pixel margin of the short focal length anyway ??? So it wouldn't matter if your chasing seeing or not ?? 

[vlaiv]

All sorts of things come into play when choosing guiding setup.

It is not only focal length that counts, it is also the size of pixels both in imaging and guiding equipment. With increased focal length of guider scope you get more precise guiding provided your mount is up to it and the skies allow but trade offs are: Smaller field of view to pick guide stars from, longer exposure needed to get good SNR on guide star. Many people use off axis guiders that operate on same focal length as imaging train but use guider cameras with smaller pixels.

I have read somewhere that you should aim for 1/3 - 1/4 in arcsec/pixel ratio between imaging and guiding setup if using guide scope. For example if using scope of 800mm focal length and have same pixel size on both imaging and guiding camera you would like to stay above or equal to 200mm focal length for guide scope. You can guide 1000mm focal length scope with 160mm guider provided that your imaging camera has twice as large pixels as guide camera.

Sensitivity of guide camera and level of read noise also play part in this equation because of way that guide programs work.

I'm in favor of fast guide scope with large FOV. This can have benefit of having multiple guide stars to guide in the same time. Not sure if this is implemented in any of guiding programs but I feel that it would be of great help in beating the seeing - perturbation of star position due to seeing is rather local so if average drift across multiple stars is taken - one expect for seeing effects to be smaller.

[Andy Mitchell]

The guide star will be spread over multiple pixels in the guide camera, and by analysing the relative brightness of all the pixels over multiple frames the software can determine the centre to an accuracy in the fraction of a pixel range. The guide camera I use with an OAG has pixels almost exactly the same size as the main camera, but it can still guide.

[StuartJPP]

Here's an interesting write-up, though that really pushes it to the limits:
http://www.wilmslowastro.com/tips/autoguiding.htm

I found that the biggest factor when I first started guiding was differential flexure...I now use a 130mm focal length guide scope with 3.75um pixel size (6.24 arcseconds per pixel) and image with a 500mm focal length and pixel size of 6.25um (2.71 arcseconds per pixel) so a ratio of ~4/10 well within spec.

[kens]

The most scientific write up I've seen on the topic is this one: http://www.sbig.com/site/assets/files/1359/autoguidingversion3_mb.pdf although the data is somewhat specific to the SBIG guiders. There are several factors involved, including focal ratio, that determine how accurately the guider can determine the centroid of the guide star.

[Owmuchonomy]

Craig Stark of PHD fame suggests not exceeding a 4:1 ratio guider to imager in arc seconds per pixel. This concurs with comments above. I am at around 6.2: 1.7 with my set up.

[iapa]

I found this in an old spread sheet that I had used to guage the ratio

 

 

[michael8554]

From the PHD2 Forum (edit - might have been the LX200GPS Forum) :

0.2 x Guide Pixel Size(microns) x Imager Focal Length (mm)

Divided by Imager Pixel Size (microns)

In my case,:

0.2 x 3.75 x 1280

Divided by 4.7 = 204mm

I use a 200mm Telephoto Lens.

Assumption is that PHD2 can measure Centroid to 0.2 of a pixel, a conservative assumption.

This happens to be a ratio in arcsecs/pixel of 1.05 to 3.87.

Michael

Edited January 24, 2020 by michael8554