Affiliation
Variable Stars South (VSS)
Wed, 07/04/2018 - 00:29
Hi All,
I'm looking to buy a light weight CCD camera for an 80mm f6 refractor and the Atik One 6.0 with built-in 5 position filter wheel looks attractive. However it doesn't have a mechanical shutter and uses interline transfer instead of full frame transfer. Are these a problem for photometry?
The QSI 660ws-8 uses the same Sony ICX694 chip but also has a mechanical shutter, deeper cooling (delta T of -45 deg C as opposed to -38 deg C) and 8 position filter wheel.It's a little heaver and more expensive. Are these features worth the extra $1000?
Is anyone using either of these cameras for photometry, if so are you happy with the results?
Cheers,
Mark
I'm using an interline transfer chip (Kodak 11002) for photometry, and -- after a period of figuring out the quirks -- it's working out pretty well. The lack of a mechanical shutter can be a problem, especially if there are particularly bright objects in your field. Be very careful taking dark frames; make sure that repeated series of exposures of the same time yield identical results.
The best precision I've managed with this camera is only about 0.006 mag, which is a bit less good than the 0.004 or 0.003 I used to get with an old camera which had a traditional CCD; but that might be due to the larger pixel size on the old chip, or other factors. I think you shouldn't let the lack of a shutter stop you ... but if all other things were equal, I might choose the camera with a shutter.
Thanks Michael,
can you ellaborate on the "quirks" you mentioned in your post? Cheers,
Mark
I'm not sure if these are properties of every ATIK 11000, or if there might be something funny going on with our unit. In any case, the properties that I don't understand (but can work around) are:
a) the mean level of dark frames does not rise monotonically with exposure time. Instead, it rises until about 2.5 seconds, then drops drastically, then starts rising again.
b) there is a low-level (< 10 ADU) pattern of vertical stripes across the chip, with a period of 32 columns. This is a known issue with the timing of the readout; others have found it, too. I'm not sure if it occurs in all of these chips, or only an unlucky few.
Hello
These two quirks sound like the things we are seeing with the CMOS cameras.
1. Its reported that some processing is being done by the camera or driver, and it subtracts off a dark like image on the fly. This is why the mean level changes and does not rise monotomically.
2. The pattern of vertical stripes is also a known artfact. I believe its caused by the increase indark current as the chip is being read.
The good news is that if you use a Master Dark of the same exposure as your science frame, these effects are greatly minimized. In fact on my sCMOS camera, the residual is less than 1 e- in most cases after the dark subtractions. You cannot scale darks with these cameras.
Hope this helps. I have lots of PT on NSVS1444........ in the AID, and the cmos results are every bit as good as the best there.
Gary
Hi Gary,
yes, the dark current behavior sounds similar to what I see with my DSLR dark current. As you say, using darks of the same exposure time as the science frames is necessary to overcome this issue.
I haven't seen the pattern of vertical stripes before. If they occur in the same places on all the frames then that is easily calibrated out. Michael, have you checked the pattern is consistently in the same columns and of the same amplitude?
Gary, I couldn't find NSVS1444 in VSX, is that the full description? Cheers,
Mark
Alas, the stripes shift position from night to night.
Hi Michael,
that's unfortunate, but I guess the stripes would only be an issue if working at the faint limit of your camera+telescope where SNR is very low. High SNR targets should not be affected significantly. Cheers,
Mark
Hello Mark
Sorry about that, I was just being lazy. The complete name of the object is the symbiotic variable J1444107-074451. Its in the AID. It was a campaign for Jeno Sokoloski. I am still observing it.
Gary
One of the most significant issues, IMO, often over looked by those selecting a CCD or similar device is the importance of Sampling. The observer should try to spread the seeing over 2-3 pixels, otherwise under sampling may occur if less than 2 pixels (can result in spurious data) See this topic in Chapter 3 of the CCD Observing Manual.
Sampling value (pixels) = local seeing (arcsec)/image scale(arcsec/pixel)
One of the best ways to help you evaluate scope and camera options is the free CCD Calculator pgm provided by Ron Wodaski: http://www.newastro.com/book_new/camera_app.html which will let you calculate the image scale of your system.
Hi Tim,
I agree, sampling is important and Ron Wodaski's CCD Calculator is an excellent tool for checking various combinations of detectors and optics.
The telescope I'll be using with the new CCD is an 80mm f6 with fl=480mm and will be for bright (<mag 10) variable star imaging. Seeing is typically around 3 arcsec here so the CCD pixels need to be small to avoid under sampling when tightly focused. However very small pixels have correspondingly low full well depth, so there needs to be a compromise.
The Atik One 9.0 has 3.69 micron pixels (full well of ~12,000 electrons) which will give an image scale of 1.58 arcsec/pixel, i.e. 1.9 pixels FWHM which is on the low side of the 2 to 3 pixels usually recommended so a little defocus would probably be required.
The Atik One 6.0 has 4.54 micron pixels (full well of ~18,000 electrons) which will give an image scale of 1.95 arcsec/pixel, i.e. 1.5 pixels FWHM so a some defocus would definitely be required. I figure if I need to defocus anyway I'd rather have the greater full well depth of the Atik One 6.0 (or equivalent in the QSI or other manufacturer).
I don't want to use a Barlow to increase focal length as that will reduce the FOV. Cheers,
Mark
Hi Mark;
I have used the QSI 690ws8 for several years (Sony ICX814), and it is the CCD camera I have used for the CCD1, CCD2, and VPhot Choice classes. Of course I need to bin (103 mm refractor), but I have been very pleased with its performance. Both the Sony ICX814 and the QSI660 which uses the ICX694 chip have the same spectral response. Note: All QSI CCD cameras implement anti-blooming.
What software do you plan to use to control the CCD? QSI only provides drivers and suggests several vendors: http://www.qsimaging.com/software.html
I hope this helps.
Clear Skies
Joe
Hi Joe,
Thanks for the info. I'll be using MaxIm DL Pro for camera control which should work with the driver from QSI. Cheers,
Mark
Hi Mark
I have been using the Atik one for sometime now and note the following...
1) It is linear almost through to saturation (but you would need to confirm this for your camera) which is good as the well depth is small.
2) the low read noise means that the dynamic range is still OK despite the small well size
3) Despite 1 and 2 above the results are never as good as I achieved with a ST-7 however many other factors may be at play (like I am pushing this camera more than I did the ST7)
4) You can 'simulate' a shutter by putting a blank in place of one of the filters. Maxim has a setting for this where it uses the blank as a shutter for Darks. When I have done this in the past I ran a test showing that some light still leaks in from somewhere so I did darks at night so not to worry about the leak. I do not do this anymore as I use all five filter slots for filters.
Best of luck
Simon
Hi Simon,
sorry for the long delay in responding, I had to make a trip to Sydney over the last few days and didn't have a computer with me.
Thanks for the feedback, looks like the Atik One would be perfectly fine for the bright star photometry I intend doing. I'll make a decision in the next week after getting quotes for a couple of cameras from Bintel. Cheers,
Mark