PU Vul (symbiotic) and consideration about sampling

[Christian Buil]

Observation of PU Vul the 8.05/06/2017 by using Alpy600 on a C11 telescope and Atik460EX in bin 2x2 mode:



Note: the Alpy600 is here equipped with a very narrow slit (13 microns width) for
maximize spectral resolution (here R = 800 at Halpha line). The guiding camera is an ASI178MM-Cooled (ZWO) model
with 2.6 microns pixels only. But if I use a Atik460EX camera for science data, in binning 2x2 mode
the spectrum is significantly under sampled, here 1.8 pixel/FWHM. The spectral resolution
is seriously limited by the pixel sampling (the Nyquist criteria is not respected, i.e more than 2 pixel/FWHM).
A binning 1x1 mode is better. Sampling is now 3.4 pixel/FWHM and R = 900 @ Ha (nearly the maximum possible
with the Alpy600 spectrograph).

Here a comparison of mu Cep spectrum taken in Bin 2x2 and Bin 1x1 mode (and a 13 microns silt):





Christian Buil

[Peter Somogyi]

Hello Christian,

It's good that somebody raises Alpy has switcheable slit. (Fairly easy I think, except you need to screw down the adapter to reach it .)

To join you, I have to confess I'm using the 18 micron slit (ordered via Shelyak's support, guess anybody can do) with the Alpy since 2017.03.31 with the 300/1200 Newton scope.
However - maybe I'm clumsy - seen a max of R~550 vs R~506 reported by ISIS (depending on which lines include in calib.)
Where I've seen benefit is the blue end somehow (but I did focus the grism there intendedly - as it's the Alpy's somewhat unique capability - hence my recent R~476 ISIS report maybe, using Ar/Ne lines for calib only + per-exposure calibration that method makes me ref.stars to exclude).
Also haven't seen any noticeable loss of SNR (most of the star still fully lost visually in my f/4 case, requires one of the best Lodestar X2 guide camera).
I was able to guide SN 2017eaw at 1 second (unsafe with star losses: 0.5 second) with the X2.

The larger pixel of ATIK 414 I have, suggests not to go below 6.45 * 2.5=16.125 micron - offered options I've got were 13 and 18 micron, or higher -and the 2.5x was the safe limit I felt (Gauss fit still to apply).

What I've been thinking about for a while, why not someone designs an "Alpy 1200" core element. (I know not as efficient as LISA, but having the blue end uniquely and as a cheap option would still make it usable.)

Cheers,
Peter

[Christian Buil]

When using a low resolution spectrograph (working at order 1, this is not the case for a échelle spectrograph), it is necessary to consider that the spectral resolution power varies greatly as a function of the wavelength. For this reason it is necessary to specify the wavelength calculation applies. Often this type of information is not provided, and this is also one of the reasons why the information returned by ISIS is really approximate.

My recommendation is to make a "manual" estimate by processing the calibration spectrum itself (Relco, neon ...) and to measure the FWHM of the profile of a selected line (for this, it reminds that it is necessary (1) not remove sky background (2) not recenter spectra trace along Y-axis, (3) choose a relatively large binning area for a good SNR).

I agree that an Alpy300, Alpy1200, ... would be welcome.

Christian

[Robin Leadbeater]

Different ALPY slit widths? When did Shelyak start offering these? They are not on the website. What widths are available ? I want to try a wider one for my faint object ALPY 200

Robin

[Peter Somogyi]

It turns out the Alpy got defocused when I've put it out (month later than focused in room).

Details (ATIK 414, Alpy 600, 18 micron slit):

[wavelength A] => [FWHM] => R estimate:

Room, focusing time:
3945.7 (3946?) => 10.622 R~371
6716.9611 => 10.824 R~621

During SN2017eaw session, one cal. (8 sec):

4764.7 => 15.1 => R~315.5 (lower lines blended, can't measure)
6717.1166: 12.157 => R~552.52
Seems like my focus got poor, need to look for the resaon.

But, I have a habit focusing the target to give a sharp horizontal line profile (1 pixel visually, 5 pixel saturated), so measuring AG Dra (asdb_agdra_20170528_868) spectrum itself directly (on the sharp lines):

4099.3358 ((H I 4102)=> 10.126 => R~405
4683.1766 (HeII 4686) => 10.323 => R~453
6674.0483 (HeI 6675) => 10.228 => R~652

So when focusing the scope (that I always do), seems these lines come back sharp.

Robin: I'd suggest asking Shelyak support.

Cheers,
Peter

[Christian Buil]

Oops, I made a mistake during identification of PU Vul field (I had not yet in the eye the characteristics of my new guidance camera...) !!! Sorry. The real spectrum of PU Vul is this one, made last night:



A low SNR spectrum because the relatively short observation time, the presence of cirrus cloud and the use of a very special configuration : a small D=65 mm, F = 420 mm refractor (TSO optics quadruplet apochromatic) - for test... ! Here the situation during observation:



("natural" light of my observatory pollution - I use a Sony A7S for the photography - note cirrus).

I have added a 1.6X Takahashi Barlow lens for increase the focal length (here F = 670 mm and a F/10 setup).

Two reasons: (1) minimize the optical aberrations of the Alpy 600 spectrograph (because the f/10 focal ratio), (2) nearly optimal sample of the stellar image by the 13 micron slit (because the F=670 mm focal length).

Below the Vega image on the 13 micron slit (the image quality is almost limited by diffraction, the effect of turbulence is negligible - it is one of the interest of small diameter telescopes) (display with two distioncts contrasts):



(note the core of the disk image visible because the 2% reflection of slit glass - It is a good thing to observe this detail.).

It's important to work in binning 1x1 mode (4.54 microns pixel) with such a configuration for not undersample. Comparison between binning 2x2 and 1x1 modes on Vega spectra:





Many details are visible in bin 1x1 and the cosmetic aspect is better. The measured spectral resolution power is R = 920 at 656 nm, a good result for this spectrograph.

Christian Buil