Dear Andy and all,
I've looked at all your messages. I think that you all already pointed out the different potential causes.
It looks you've an astigmatism issue, I confirm : depending on the focusing position, the star image is either vertical or horizontal. This is why star spectrum (horizontal) and Neon spectrum (vertical) are not fcused at same position.
This astigmatism is inherent of the Lhires III, because we're working in a 3D scheme - it is perfectly shown by the calculation (see Christian Buil's pages, as mentionned by Robin, I think).
When the Lhires III is properly tuned, the astigmatism is minimal, and is almost not visible. This is not your case, for sure.
To my knowlege, there are three potential root cause for increasing the astigmatism effect (two of them have been mentionned in this thread) :
- if the acquisition camera adapter has not the right thickness ; then, the distances slit/collimator and CCD/collimator are not the same.
- working with "high" off-axis angles (astigmatism appears when you work out of the optical axis). This can happen for instance if the grating os rotated on its holder (rotation in the plane of the grating surface). This may make the off-axis angle higher than nominal. If you're in this case, you may be at (close to) maximum position of the main mirror tuning (to compensate the off-axis). Not sure I'm clear... tell me.
- grating surface deformation (not flat enough). It can happen if there is some dust between grating and holder. We've seen that sometime (not often). In the first generation of Lhires III, the grating was glued on the holder, and lot of people had problem with temperature changes- this is now totally solved (grating are not glued anymore for a long time).
It is also possible that you've a combination of these phenomena.
My feeling tells me that you should look at the grating first (not flat surface), because this is the most important effect - and your effect seems important.
The best way to further investigate is to change some parts ; grating (as you did with the 600l/mm), or use a hole instead of the slit (a very efficient way to see the astigmatism effect). Of course, you must first tune as much as possible the instrument ; but it looks you didi it already (many times !).
I think it will be useful to send you some of the above pare parts, to make the exchange and see what happens.
We see the details by mail (
francois.cochard@shelyak.com) ?
Sure we'll find the bug !
Friendly,
François