Sensor tilt

[Jezphil]

Any advice is super welcome on how my sensor tilt correction is developing...

I made some slow but slightly encouraging progress with correcting the tilt last night in UK skies affected by US smoke. It took around four hours to go through multiple iterations, using maths and feeler gauges in a bid to remove the guesswork. I found it pretty hard, but tried to be as methodical as possible. The comparison image shows the ASTAP 'CCD Inspector' analysis of the factory set tilt on the left and my adjusted tilt on the right. I think the single bottom screw of the screw triangle can now be adjusted to reduce the top to bottom discrepancy. Does it need to be made shorter or longer? The current spacing of screws according to my feeler gauge are:

TOP LEFT  +0.68mm

TOP RIGHT + 0.71mm

BOTTOM MIDDLE +0.63mm

The other image here is the right hand (post adjustment) frame from above. This shows my back focus is slightly too close (do you agree?) but that should be an easy enough fix. I think the vignetting emphasises the need to get the top to bottom adjustment dialled in more?

What do you think?

Edited August 19 by Jezphil
title typo

[symmetal]

First quick check of your image shows that there is no tilt top to bottom, and slight tilt left to right shown by the elongated stars on the left edge. I ran it through CCD Inspector and it confirms this.

Vertical tilt is perfect. 🙂 The best area of focus (lowest FWHM) is just right of centre. The darker the colour the better the focus here.

I should ignore the ASTAP results as they are rather suspect here. I don't know what weighting various areas of the image are given to produce its results but they are wrong in this case. It's much better to just take an image and look at the corner stars, which will indicate where any tilt is with more accuracy.

To determine which direction you need to adjust the tilt corrector, make a note of the focuser position and then take another image with the focuser moved in a small amount. Look at the left corner stars and see if they are better or worse than before. If they are better (sharper and more circular) then it indicates that the left side of the sensor is further away than the right side. If they are worse, then the left side is closer than the right side.

If you happen to have tilt in both axis and it's hard to know where to begin, then just do the above, moving the focuser in and out by small amounts, examining the corners each time and making a note of the focuser position at all four corners which gives the best stars in that corner. This will indicate which corner is behind or in front of the central area and by how much.

Alan

[Jezphil]

55 minutes ago, symmetal said:

First quick check of your image shows that there is no tilt top to bottom, and slight tilt left to right shown by the elongated stars on the left edge. I ran it through CCD Inspector and it confirms this.

Vertical tilt is perfect. 🙂 The best area of focus (lowest FWHM) is just right of centre. The darker the colour the better the focus here.

I should ignore the ASTAP results as they are rather suspect here. I don't know what weighting various areas of the image are given to produce its results but they are wrong in this case. It's much better to just take an image and look at the corner stars, which will indicate where any tilt is with more accuracy.

To determine which direction you need to adjust the tilt corrector, make a note of the focuser position and then take another image with the focuser moved in a small amount. Look at the left corner stars and see if they are better or worse than before. If they are better (sharper and more circular) then it indicates that the left side of the sensor is further away than the right side. If they are worse, then the left side is closer than the right side.

If you happen to have tilt in both axis and it's hard to know where to begin, then just do the above, moving the focuser in and out by small amounts, examining the corners each time and making a note of the focuser position at all four corners which gives the best stars in that corner. This will indicate which corner is behind or in front of the central area and by how much.

Alan

Alan - thanks so much for spending your time on this. The above results are a partial relief because I was up until 4am this morning correcting a severe tilt whilst feeling unwell. Looks like the left/right tilt is an issue which I will examine. Thanks for the tips on using the focuser - will take those on board and try the method on those left hand stars. I know my back focus was slightly out due to the extra distance created by expanding the tilt sensor and I have removed a small 0.5 mm ring which hopefully will help sort that that. 

Edited August 19 by Jezphil

[Jezphil]

I bought a Gerd Neumann tilt sensor after spending several frustrating hours attempting to perfect the above. This was a game changer - so much better that the ZWO tilt sensor, especially because it can be adjusted from the side without having to disassemble the optical train each time. With the Gerd Neumann tilter I was able to get to this stage with 0% tilt in around 30 mins and can highly recommend it. Here's the result. Still not perfect, but the stars look okay...

Edited August 29 by Jezphil

[phantom199]

To balance tilt, try adjusting the bottom screw to match the top better. For back focus, it does seem a bit off but should be an easy fix.

[Jezphil]

27 minutes ago, phantom199 said:

To balance tilt, try adjusting the bottom screw to match the top better. For back focus, it does seem a bit off but should be an easy fix.

Thank you. But the tilt is now classed as 0%, so should be fine I think.  What are you using from the image to asses the back focus here? i.e. by looking at the CCD inspector analysis rather than looking at the stars themselves? 

Edited August 29 by Jezphil

[symmetal]

Good choice on getting the Gerd Neumann tilt adjuster. They're expensive but as you say save many hours of messing around. 🙂 Result is looking pretty good now too. If your corner stars are round then the sensor spacing is fine. They will almost always exhibit a higher FWHM compared to the centre, unless you have a small sensor or very expensive scope. 😃

Alan

[Jezphil]

16 minutes ago, symmetal said:

Good choice on getting the Gerd Neumann tilt adjuster. They're expensive but as you say save many hours of messing around. 🙂 Result is looking pretty good now too. If your corner stars are round then the sensor spacing is fine. They will almost always exhibit a higher FWHM compared to the centre, unless you have a small sensor or very expensive scope. 😃

Alan

Thanks Alan. Yes, big sensor and not very expensive scope. The stars are slightly oblong in the corners, but don't conform to either of the patterns below. I am tempted to tweak the tilt a little more tonight. 

Edited August 29 by Jezphil

[symmetal]

I haven't found that spacing guide useful either. I tend to get more triangular looking stars when it's out in one direction and elongated and coma like in the other.

A quick check to see if it's spacing, is just as I mentioned on my previous post about determining tilt direction. After focusing normally, make a note of your focuser position and then move the focuser in a very small amount and take another image to see if the corner stars get better or worse.

If you move the focuser in slightly and all the corner stars look better then it means the sensor was too far away initially, so you need to reduce the spacing distance.

Likewise, if they all look worse then the sensor was too close so need to increase the spacing distance.

Alan

[Jezphil]

23 hours ago, symmetal said:

I haven't found that spacing guide useful either. I tend to get more triangular looking stars when it's out in one direction and elongated and coma like in the other.

A quick check to see if it's spacing, is just as I mentioned on my previous post about determining tilt direction. After focusing normally, make a note of your focuser position and then move the focuser in a very small amount and take another image to see if the corner stars get better or worse.

If you move the focuser in slightly and all the corner stars look better then it means the sensor was too far away initially, so you need to reduce the spacing distance.

Likewise, if they all look worse then the sensor was too close so need to increase the spacing distance.

Alan

Thank you. Yes, I did take your advice and tried that. The stars wren't in better focus after the focus move and I do think the issue has been tilt. The stars are looking very good in three corners and I am perhaps pixel peeing the other a little too much, but will tweak again anyway!