PHD2 Calibration and Focal Length

[wave...@talktalk.net]

I have a 10" LX200 (F10 ) fitted with a 0.63 focal reducer directly on the back, then an OAG. The guide camera (ASI120 mono) has 3.75 micron pixels and is binned x2 for guiding. The imaging camera (ASI224) also has 3.75 pixels. My question is about the focal length I should insert in the Profile Wizard. So far, I've used 1600mm  (2540mm x 0.63) and manage to guide below 1 arcsec total RMS. PHD2 reports a guide image scale of 0.97 arcsec/px. I assume this would mean the imaging camera has a scale of twice this at 1.94 arcsec/px (?)

Anyway, I submitted to Astrometry.net a test image of the M3 cluster obtained from this setup to determine the image scale produced by the ASI224. This came back as 1.04 arcsec/px. Calculating the scope's effective focal length from this image scale gives a result of 744mm, less than half the 1600mm 'theoretical' figure. So, am I using the wrong focal length or am I missing something obvious?

A previous Astronometry.net test using the same scope/FR setup but without the OAG gave an image scale of 1.14 arcsec/px and thus a focal length of 678 mm. Small variations in the results are likely due to varying spacing between the FR and the camera chip, but why are they so very different from 'theory? Am I over-thinking this stuff?

Cheers,

- Jack

[mj.w...@gmail.com]

Hi Jack

I use this tool for Image Scale calculations:

https://astronomy.tools/calculators/ccd

With the theoretical 1600mm FL:

ASI120MM with Binned 3.75um pixels (=7.5um) at 1600mm FL = 0.97arcsecs/pixel, surprisingly the same as PHD2 reports.

Surprisingly, because SCT FL depends on the position of the primary mirror after focusing, so 1600mm would usually be thereabouts.

Now an ASI124 with 3.75um pixels at 1600mm FL = 0.48arcsec/pixel, worth Binning that if it's the Mono version.

So 1600mm is the correct FL, and Astronomy net has been given an incorrect setting ?

Michael

Wiltshire UK

[Bryan]

Jack

A puzzle!  Sort of thinking on the fly here.  No promises about answers!

My sense from your description is that, for whatever reason, AN got an image scale (1.04) for the 224 that is off by about 2x; hence the calculated f.l. is low by 2x.

Astrometry.net (AN) can blind solve images.  It does not need anything other than the image itself.  It does not know or need to know about OAGs, sensors, pixel size.  It will get sensor  dimensions (x total pixels by y total pixels) from the image, e.g. FITS header, and FOV from its solution.  It calculates image scale from those.  As you know, it does not know about individual pixel size; hence cannot calc f.l.  That, of course, is why we have to do that calculation manually.

Is there any chance you gave AN a parameter somewhere?  If you did not let AN do a blind solve, you might try that.  Other possibility that comes to mind is that the image given to AN was cropped or somehow changed.

Bryan

[wave...@talktalk.net]

Hello Michael, many thanks for your comments. I basically used the same equation you linked to. The guide image scale reported by PHD2 is simply from the same calculation. It's using the FL(1600) that I enter in the Profile Wizrd. So, everything it reports about my guiding is only as good as the data I give it.

Cheers,

- Jack

[wave...@talktalk.net]

Hi Bryan, Yes I noted the 2-fold discrepency, yet I simply uploaded the FITS image (2 stacked using Deep Sky Stacker without any cropping) and with no additional parameters of any kind. I'm mystified.

Cheers,

- Jack

[mj.w...@gmail.com]

"The guide image scale reported by PHD2 is simply from the same calculation. It's using the FL(1600) that I enter in the Profile Wizrd. "

Doh ! Slaps forehead.....

Michael

Wiltshire UK