eShel (II) progress, with problems

[Bernard Heathcote]

Over the last two or three years I have upgraded, at considerable cost, my 11 year old original eShel to effectively the latest II version. This exercise involved installing a new collimator lens, Canon 135mm f/2 L-series lens, OH fiber optic cables, QHY269M camera and a 250mm f/4 Newtonian scope, replacing a C11 SCT.

Overall I am happy with what this has achieved, extending the wavelength range with better SNR into the UV and IR, going from typically 22 orders (32 to 53) to 32 orders (29 to 60), a 45% gain!

The first problem I've found is that there is a small gap, typically only a few Angstroms, between most orders. See last image below.
The only explanation I've found is that the FOV of the QHY268M camera (with IMX571, 28.43mm diag CMOS), similar to the ZWO ASI2600MM, may be cutting off the ends of the orders (see middle thorium image below). If this is the reason then I am to blame in trying to save money by not getting a camera with the larger IMX455 (36.24mm diag), as in the ZWO ASI2600MM or QHY600M and as used by Christian.

The next problem is the merged orders spectrum made by ISIS. As the first image below shows, there is a significant wavelength shift or error, clearly shown by the Balmer lines. I've tried various software option but without success. Could it be caused by the 'cramped' spectrum image, or is it a bug?

(Sorry, but the three images are in the reverse order I intended and I don't know how to change that)

Any/all comments welcome as to how to fix these problems.

Cheers,
Bernard

[Bernard Heathcote]

Adding a comment to my post above, comparing my present 2D Th-Ar image (Cannon 135mm with QHY268M) to previous original (Cannon 85mm with QHY22), I see that there is in fact a larger FOV with the present setup, but shifted to the left, so it is unlikely to be the cause of the order gaps.

As the second image shows, there is in fact a good overlap (set to 7%) between a couple of example orders, so it doesn't appear that the orders are too short. So what is the reason for the large short negative spikes?

Cheers,
Bernard

[Joan Guarro Flo]

Hi Bernard,

How are things in Melbourne? You exposed here some questions that in my opinion claim step to step solutions, and I'll try to explain honestly what I'm thinking about them and what I would do if I were you.

You can use perfectly  your echelle with any problem in these both telescopes, it's better that you bring the C11 at  F7/6 through a focal reducer, but if you want to use the Newton at F/4, first you to be really sure if on the optical fiber, the F/D light conus in its output is the same as its input, - sorry, I unknown it -, in case that it were true, you need to have in account that the condenser focal length lens of your echelle is 120mm., so 120mm / F4 = 30mm. So that would mean there is a loss of light when the light arrives at the width 25mm. echelle grating, the problem increases when the light goes out from this grating because it's not a cylindrical shape so is a bit  conical. This conus light bumps over the prisma surface inclined around 35º making an elliptical shape that it's possible doesn't fit on the prism surface, and you should need to carry up your Newton to F6/6.5 through a good Barlow lens.   All of that is necessary if the input and output light cones have the same angle values, - in it case F4 -, in your optical fiber.  Another way is to directly ask for it at Shelyak, they always serve everyone very well.

Don't worry, your QHY268M camera fits very well with your 135mm. objective, and to clarify a bit more these problems in your graphics, at the moment  I would need to see their CHECK2.fits where previously you have pointed to the "Reference line length", and a screenshot of your ISIS, Seput window.

Cheers, Joan.

[Bernard Heathcote]

Just a quick update on this problem. I finally got around to digging deeper into what was causing the merged orders wavelength shift, over 20A at Halpha, and it was a faulty def_order.lst file!

Cheers,
Bernard

[Berthold Stober]

Dr. Daniel Sablowski has calculated a Petzval lens from two achromats with 1:2/109, which shows practically no error of the chromatic longitudinal aberration and due to the slightly shorter focal length than 135 mm still allows to use smaller chips (KAF 3200). I built the lens. Highly recommended

[Olivier GARDE]

Yes Berthold, this new Petzval lens looks very good and better than the classical 85mm Canon lens with better focus à the same time in the blue and red parts of the echelle spectrum.

We do some test actually with the eShel and also the whoppshel.

A new solution for echelle spectrograph because we can see that lens manufacturers (Nikon, Canon, etc..) are abandoning more and more the sale of fixed focal lengths: the Canon 135mm f/2 has disappeared from their catalog since few months.

So we have to find other solutions to continue to propose echelle spectrographs.

[Bernard Heathcote]

Hi Daniel, Berthold, Olivier,

The news about the Sablowski Petzvall lens is most interesting and encouraging.
When can we expect to learn more about its performance, availability and cost or, if is a DIY project, can we have the details on how to build our own?

Cheers,
Bernrd