Magnitude limit for CCDs?

[lukebl]

Hi all,

Call me weird, but I'm slightly obsessed with capturing small dots, i.e. asteroids, faint planetary satellites, plutoids and the like.

With my Canon 450d and f/4.8 Newtonian, I can capture objects down to around magnitude 17 or 18 but would like to go deeper. I can't stretch to a larger-format cooled CCD, but I just wondered if one of the smaller (cheaper) cooled CCDs like the Atik Titan would be able to get me further magnitude-wise? I'm not worried about the small field, but would the Titan, for example, get me down to say magnitude 20?

As an aside, If anyone has a budget cooled CCD up for grabs, then I'd be interested.

[Trull]

Any of the pukka CCDs will have sensitivities at least 4 times that of your Canon, particularly if you go for a mono sensor. This is because they have larger pixels, and no RGB filter infront. Add in the cooling and you then really see the sensitivity rise. Have you got a pal you can borrow a CCD from?

[FraserClarke]

It'll help, but I'm not sure it will get you another 3 magnitudes of sensitivity -- that's a factor of 16...

I think you'd need to combine it with longer total exposure times, good tracking/image quality (make the stars as sharp as possible, and make the pixel is match, so that you spread out the sky background as much as possible), and -- afraid to say -- more aperture...

Odd as it seems, *more* focal length might help you -- making the 'scope slower. If you can keep the image quality up, push the focal length out until your pixel scale is ~half the seeing size (typically ~1--1.5"/pixel, depending on your site). That means you're keeping the sky background per pixel as small as possible, whilst keeping the object flux within a few pixels. That will maximise sensitivity for point sources like asteroids. As soon as the image quality starts to go >2 pixels though, it's a losing battle.

How dark is your sky?

[lukebl]

...How dark is your sky?

I think about mag 5.2 on a good night.

Thanks for the comments guys. The simple question is 'would a cheaper cooled CCD record fainter stars than my DSLR?' I think the answer is probably yes, but it would be nice to be able to quantify it. i.e., has anyone with, say an Atik Titan or a DMK, managed to image stars to 19th mag?

[ollypenrice]

Interesting thread. I really don't know but will have a go at finding out because I am planning some deep exposures in Leo with the Atik 4000 mono. I will try to see how deep is deep from a stellar mag POV. We have a mag 7 zenith for those with young eyes.

Olly

[rfdesigner]

If it's any help I'm using an uncooled Lodestar on a 12" on fairly good skys, so that's a b&w CCD but with some dark current. (This is a temporary situation until I purchase a cooled CCD for imaging)

I can reach a sky background of ~19th mag in 10 minutes with 5second frames... I'm 90% convinced I've recorded a 20th mag object in a half hour shot, again with 5sec subs.

If I had taken the same shot with 3x10minute shots on this camera I would expect to have suffered some 2.8 magnitudes less camera noise, so no doubt skyglow would have begun to dominate.

To go deep for asteroids I would:

1. focus. If you can tighten up your stars you'll buy a bit of depth. (bahtinov mask)

2. tracking. as per focus, smaller stars means more depth.

3. long subs (10 minutes).. means you rid yourself of being dominated by readout noise in the camera.

4. yes get a cooled CCD.

5. If you have much light pollution concider a LPR filter.

6. Remember NIR photons are more numberous per Watt that Visual or UV, so a DSLR with a IR filter will not be helping you.

Any cooling will help, but if you go for setpoint cooling it makes processing easier. In my case I have noticable dark current AND no setpoint, so when I try and take away dark current I have to compare the apparent dark current in the image and then scale the dark frame to match before subtracting.

One thing that bites me continuously with a small format camera is targetting... you will waste quite a lot of time trying to hit your target unless you have fairly accurate GOTO or some parfocal way of swapping eyepieces and camera without disturbing the scope (this is what I do)

Hope this is some use to you.

Derek

Additional:

check out the lodestar product info page: http://www.starlight-xpress.co.uk/Lodestar.htm

1x 600s shot on a 12" scope; Centaur Eris at 18.75Mag

[lukebl]

Thanks Derek, that's really informative.

To achieve 20th magnitude with 5 second subs sounds amazing. Did I understand that correctly, as you later say that Eris was captured with a single 600s shot? I also don't quite understand what you meant by long subs ridding yourself of being dominated by readout noise in the camera. I thought that the longer your subs were, the more noise you got.

Anyway, I've just picked up an old MX716 which is basically a cooled version of your Lodestar, so I'll see how I get on (tiny chip and and field of view, I know!). Do you do much imaging with the Lodestar, as I assumed they're primarily for guiding?

[rfdesigner]

lukebl

Take a close look at the eris image.. eris isn't the faintest object, I'm guessing the limiting magnitude is somewhat fainter than 18.75

Also I was at 20th mag.. only maybe, after half an hour.

Noise:

10 images of 1 minute will have 10 minutes of dark current & 10 lots of readout noise. 1 image of 10 minutes will also have 10 minutes of dark current but only 1 lot of readout noise. Both will have the same signal level.

and in maths:

Lodestar Readout noise = 15e

Dark current at +10c = 0.1 e/s

When integration time = 600s

Dark current = 600*0.1e = 60e

Dark current noise = sqrt(60e) ~=8e

Total Noise = sqrt(15e^2 + 8e^2) = 17e

When integration time is 120 * 5s

Dark current = 600*0.1e = 60e

Dark current noise = sqrt(60e) ~=8e

Total Noise = sqrt(120*(15e^2) + 8e^2) = 164e (~2.5Magnitudes more noise than the single frame.. doesn't mean it will see 2.5 more magnitudes as there's skyglow to concider)

Take a quick glance at my album for some pics I've taken.

I bought this camera when I did as I'm using it to help get my mount development correct. Once that's working properly and I've got good tracking I can go to long exposures and it's worth buying a cooled CCD.

Hope this helps

Derek