Using a Bahtinov Mask to measure sensor alignment.

[avandonk]

I have been busy tweaking my new system to iron out all the bugs. One problem I had was image train flexure. This is very sensitive at F3!

This has been solved by the construction of an image train stabiliser. See image one below.

Here are two images about 5MB each.

http://d1355990.i49.quadrahosting.co..._08/41000_.jpg



http://d1355990.i49.quadrahosting.co..._08/42000_.jpg

They were 60s exposures in green with a Bahtinov Mask 1000 steps of the Atlas focuser apart. 1000 steps is 85 micron.

The image 42000_ is at focus in the centre brighter stars and at the bottom centre are also in focus. This means the sensor is orthogonal to the optic axis in the vertical.

But if you look at the RHS centre there are two brighter stars which are out of focus.

If you then look at these same two stars in image 41000_ they are then in focus! This means the RHS of the image is 85 micron closer to the front of the optic.

This of course means that the sensor is 85 micron closer on the LHS. So I need to shift the ITS to the right.

This is a very simple way to see what is happening visually to align any sensor. Like all simple elegant methods they are blindingly obvious in hindsight!

In fact if you suspect image train flexure this is a simple way to see if it varies depending on mount orientation with your system.


Here is deeep image.
This is 10x16 minutes of 3nm NII of NGC6334 and NGC 6357.

http://d1355990.i49.quadrahosting.com.au/2012_07/ngc6334&6357_ITS.jpg


The image is at the native resolution of the PL16803 camera.


Bert

[troypiggo (Troy)]

Great method for working it out!

I'm curious why you're testing using green filters? Wouldn't Luminance give quicker results with stronger signal to noise ratio?

[Geoff45 (Geoff)]

I'd been thinking along these lines myself, but it never occurred to me to use the BM. Good move!
Geoff

[avandonk]

I did a final? alignment and check last night of the ITS.

This time I did 60s exposures and jogged the mount by 30' seven times about seven seconds apart.

This gives a line of diffraction patterns of the same bright star that are far easier to compare.

Horizontal image 2MB

http://d1355990.i49.quadrahosting.co...2_10/A_hor.jpg


Vertical image 2MB

http://d1355990.i49.quadrahosting.co..._10/A_vert.jpg


Horizontal alignment is almost perfect ie within about 5 micron. This is less than a 9 micron pixel! In angular terms about an arc second misalignment horizontally.

Vertical needs to come up a tiny tad. The sensor is about 15 micron closer at the bottom compared to the top.

The last image is a single 8minute exposure in 3nm HA of a bit of the Vela SNR.

Full size image here 10MB

http://d1355990.i49.quadrahosting.co..._10/VelaHA.jpg

Bert