Rethink on the Canon 400D

[MarkS]

I forgot to say I did a 3-way batch test of the Canon 300D, 350D and 400D under identical conditions at the weekend. I included a 2-hour "thermal" test where I had all 3 cameras simultaneously taking successive 5 minute exposures for 2 hours i.e. a typical imaging scenario.  I haven't had a chance to write the results up because  I've been so busy at work.

However, the 400D didn't come out as bad as I expected, given Christain Buil's review.  I came up with a higher gain figure than he does.  This matters quite a lot because a higher gain figure indicates that the CCD is catching more photons. 

It is true that 400D frames appear noisier than the 350D frames because the 400D CCD has a wider variability between the pixels.  This increases noise estimates and in turn lowers the estimation of gain.  My approach (the same used by Carl Stark) uses the difference between 2 frames at every stage of the gain calculation to cancel out the extra "noise" caused by the pixel variations.  This increases the gain estimate and, I believe, is more accurate.

The upshot is that Christian Buil reckons the QE of the 400D is worse than the QE of the 350D and then then smaller pixel size of the 400D reduces per pixel photon collection still further.  If I am right, the QE of the 400D is better than the 350D and the resulting improvment in photon collection partly compensates for the smaller pixels. 

I'm now beginning to think that as long as proper dark frame subtraction is used for 400D images then the results will be, in practice, virtually indistinguishable from 350D astro-images.  Also the 400D has no amp-glow. And they're going for £200 secondhand on eBay.

Proof of the pudding will be once I've H-alpha modded it.  But I'm not going to touch it until I've finally got the Tak collimation and CCD tilt problems completely sorted.

Mark

[Rocket Pooch]

I do admire your persistance with the DSLR, but then again I keen persisting with sub £400 telescopes.  If only I have another £4k for a proper telescope

Will be good to see your results though.

Just as an asside, I connected up the QSI last night put the cooler on full and tried to get it too -20c to do some referance bias and darks, it could not reach -20c, got stuck at -18c.  So unless I get the water pump for it it seems the £3k CCD camera (regulated) is still ambient critical.

But I have a spare fridge, so the camera is going in that tonight, I'm positive it will get to -20c then, after all I know the beer was cool before I drunk it all.

Actually the effort of drinking a whole fridge of beer to take some dark frames is not recommended....

[Mac]

Just as an asside, I connected up the QSI last night put the cooler on full and tried to get it too -20c to do some referance bias and darks, it could not reach -20c, got stuck at -18c.  So unless I get the water pump for it it seems the £3k CCD camera (regulated) is still ambient critical.

[just a quick Hijack]
How about a peltier mod on the back of the QSI. Just to give it that extra cooling.
1 12v Peltier + 1 12v Fan. on top of the cooling already offered.[/just a quick Hijack]

On a though Mark, before you mod the 400D, could you not stick it in the fridge and take your frames, that way you simulate the camera at 3C and then the freezer at -28C.
for a comparison on the noise issue

[RobertM]

That's a very interesting comparison Mark.  From memory the 400D has larger pixels than my 450D so should be more efficient.  Frame to frame dithering should offset most of that variability you're experiencing.

Robert

[Ian]

Just to reply to Mac, that would effectively be a stack of pelts, which in general doesn't work too well as the additional thermal load of the second pelt would make for a huge heatsink and fan. Better to find another way of removing heat. How about immersion in mineral oil    works for arc welders.

[RobertM]

How about immersion in mineral oil    works for arc welders.

Yes but it seems to have a mind of it's own like Transformer oil 

All that's needed is some form of supplying either more or cooled air to the air intakes, it could even be that a fan would give the edge to get that extra few degrees.

Robert

[mickw]

CO2 fire extinguisher would do it 

[Ian]

Yes but it seems to have a mind of it's own like Transformer oil 

You mean it looks like innocent oil but then becomes a giant megan fox eating killing machine?

But, you're right, shift more heat from the existing hot side, or lower the thermal load on the cold side. Smaller ccd? I've got an old toucam ccd you can swap yours for chris.

Any more thread hijack and this thread will be split

[Rocket Pooch]

How do you hijack a Thread?  Also what ransom would you get from the Thread's family?

Also if you watched Transformers you would have seen Transformer oil is in face  Bumble Bee wee wee.

[RobertM]

I've never seen Bumble Bee wee but is that way Honey tastes the way it does ?

I was thinking more along the lines of the creaping gunk from Quatermass experiment...

[Fay]

Well, it is all done in the honey!!!!!So is it that healthy?