This is a preliminary review. Once I have had a chance to use the Modules at the telescope I will write more. I think it very important to experiment and make adjustments on the bench before going to the telescope.
I received my ALPY 600 Basic ($760) and Guiding Modules ($850).
Basic Module
The first thing was to couple a CCD camera to the Basic Modules Core Element (which contains the slit plate, lens and grism. Next, is to focus the Core Element and set the rotation for a horizonatal spectrum with the shorter wavelengths (blue to the left. I use an Orion StarShoot G3 monochrome camera for spectrum imaging. It has 19 mm back focus. This is supposed to work, but I had a problem. I used a single neon lamp for the light source for the spectrum focusing. It turns out the only way I could get proper focus on the spectrum was to remove the locking ring on the Core Element. There is a lip on the CCD Camera Coupler that is o.160” high. By milling off 0.060” I was able to get good focus and use the locking ring. The next item was with the Guiding Module. After focus is achieved the locking ring locks the Core Element travel. There are 3 hex screws on the Core Element body. These must be loosened to allow the slit to be rotated so the spectrum is horizontal and with the shorter wavelengths to the left. Using the neon lamp the neon spectrum should be all on the right side for proper orientation.
Guiding Module
When using the Guiding Module the slit plate on the Basic Module must be removed and the 23 μm reflective slit (supplied with the Guiding Module) put in its place. The connection to the guiding camera is a C-Mount. Most cameras (all that I have) use a T-thread. An adapter can be purchased from ScopeStuff for $24, item #T2CA. This comes with a washer, but may not be needed. Make sure you get a male T-thread to female C-mount adapter. I use a DSI Pro II monochrome camera for the guiding. It has a low profile adapter from ScopeStuff and the focus was good.
The Guiding Module is a considerable expense, but works well. Shelyak recommends using the Basic Module without the Guiding Module with a flip mirror. While this will work, it will be a challenge. Once the mirror is flipped up there is no easy way to keep the star on the slit. At HPO we are experimenting with a fiber optic coupler and star diagonal. If this works out it will be described in detail in my new books sue out later this summer or early fall from Springer. The complete fiber optic coupler can be made for under $50.
Note: While the Calibration Module can be handy, the expense is a bit extreme. A simple neon ring for under $5.00 can be made that provides an excellent wavelength calibration. In addition once a good wavelength calibration has been obtained it should remain stable for future observation. The neon ring can provide periodic checks. I do not recommend the purchase of the Calibration Module unless you really have the extra money. I do not feel it is needed.
Next is a report from the telescope.
Jeff
Hopkins Phoenix Observtory
Last night was the first chance I had to try the ALPY 600 with the Guiding Module on my 8" LX90 telescope.
It was close to 90 degrees out, but clear. As I expected I learned a lot and will be fine tuning the ALPY 600 and my technique. The following is a 30 second exposure of Vega. I used the Guiding Module with the 23 micron slit. It shows the Hydrogen Balmer Lines nicely. On the bench I adjusted the ALPY for spectral lines to be vertical, but the star's spectrum was tilted. Unlike the Star Analyser, one cannot easily readjust that with the ALPY as it requires taking it apart. I was able to correct the tilt with RSpec, but I am not happy with the spectrum. One thing I notices is at F/10 the star is very big. With my Lhires III I could always tell when the star dropped in part into the slit. I could see very little change with the ALPY. My next try will use a 6.3 reducer.
I do feel it was a good initial try, however. Attached is a line profile of that spectrum.
Jeff
Hopkins Phoenix Observatory
(187283)
Counting Photons
Phoenix, Arizona USA
www.hposoft.com/Astro/astro.html
I have been able to get a couple more nights worth of spectroscopy using the ALPY 600.
As noted earlier, my first spectrum was tilted slightly. I corrected that for the 2nd and third night's observations. I also refined the guiding camera focus and was able to cut the observation times in half for the same ADU counts. One thing I wanted to check was the stability of the wavelength calibration. I used Vega for all these observations as it has easily identifiable Hydrogen Balmer lines.
Between night 1 and two the calibration shifted around 25Å. I attribute this to the fact I had to disassemble the unit to get the tilt corrected. The calibration between night 2 and 3 remained very close. There was a slight shift of about 5Å. This is close to the resolution of the unit. What I recommend is to determine a good calibration using an A type star and save it. Then use this calibration for new observations, but fine tuning the calibration for the new observation. The initial calibration should be good enough to allow easy identification of spectral lines for further fine tuning. I found using a neon ring for calibration was not needed and as such I do not think the Calibration Module is worth the cost.
Next is to do some experimenting with a low cost fiber optic interface and just the Basic Module.
Jeff
Hopkins Phoenix Observatory
(187283)
Counting Photons
Phoenix, Arizona USA
www.hposoft.com/Astro/astro.html
Jeff, I'm sure that it is not fun to try to get the spectrum lined up so that the software is happy. So I'd like to point out that you can still extract and calibrate that spectrum that you got the first night. You can even calibrate it! See the short tutorial I put together for the SAS last year:
http://edocs.uis.edu/jmart5/www/SpectroscopyWorkshop/ExtractTutorial/index.html
[quote=HPO]
The calibration between night 2 and 3 remained very close. There was a slight shift of about 5Å.
[/quote]
Hi Jeff,
Shifts of this magnitude (more than a pixel) are actually quite large and unexpected. I ran a series of tests on my Alpy setup last night looking for flexure effects in different parts of the sky and thermal efects over ~20C temperature range. I found nothing greater than 1/10 pixel (0.6 um) - An excellent result.
Cheers
Robin
Thanks for sharing your results. It is very useful for future users of Alpy600 as me.
Hi Robin,
I'm a total novice, don't even have a scope yet just some binoculars, but did wonder how you measured the 1/10 pixel shift? I expect its a very standard technique but haven't found it in the few text books I've read and could only think of three possibilities:
1. A calibration source with two spectral lines spaced across ten pixels
2. A means of finely adjusting the distance to the camera array so that you could magnify the spectra in steps of 1/10 pixel
3. A calibration source with a line tunable to within 1/10 of a pixel
No idea if any of these methods are conventional or practical but would love to know how you did it and whether the technique is adaptable to a measurement or data processing technique that improves the apparent resolution of a spectroscope.
Cheers, Peter
Hi Peter,
Completely missed this one at the time, sorry but here is the answer anyway. Measuring the wavelength of a line to much higher precision than the resolution of the spectrograph is straightforward though unfortunately it does not improve the resolution itself (ie the ability to separate two closely spaced features) This is also not the same as absolute accuracy as here we are just looking for shifts between measurements. High absolute accuracy is much harder to obtain.
A line covers several pixels to give typically a gaussian shape line profile but for simplicity we can consider a simple rectangular line covering exactly 2 pixels, with equal intensity (say 100) in each and zero elsewhere. ie reading across the line we will see pixel values 0,0,100,100,0,0
If that line then moves by 1/10 pixel then we will see pixel intensities 0,0,90,100,10,0 which is easily detectable provided we have low enough noise to detect the change.
In practise the line centre is calculated using all the data in the line profile, for example by fitting a gaussian profile or calculating the centroid.
If you have a number of lines to work with then even smaller shifts can be detected by cross correlating the spectra. This is how shifts many hundreds of times smaller than the resolution of the spectrograph are measured when detecting exoplanets for example.
Cheers
Robin
Thank you HPO!
I have an Orion G3 and read your comment about the locking ring.
Jupe.