The graph below was generated by taking a ten second dark at 400 ISO straight after turning on the camera after the camera had equilibrated at a range of fridge temperatures. The raw frames were converted to linear tiffs and digitally developed with a break point of 1268 in Images Plus. IP was used to measure the value of the same hot pixel with the photometric function under point.
I fitted a third order polynomial to the data with the graphing software to get a best curve to fit the data.
We can now use this graph to measure sensor temperature directly. It is just a matter of taking a ten second dark at 400ISO and measuring the value of the same hot pixel.
With far more data I got quite reproducible results.
The sensors temperature is consistently 3.5C higher over the full range of fridge temperatures when the camera is on and at idle (not exposing) compared to camera off. This takes about twenty minutes to equilibrate from off.
If you set the camera taking exposures continuously the sensor temperature rises by 16.5 C from the off state. This of course is 13C above the idle state. It takes about forty minutes to reach equilibrium.
What all this means is that you should take at least 30 to 40 minutes of dummy exposures before taking any lights or darks. Then you have some sort of hope of them matching in temperature. This really depends how far your ambient temperature drops over the imaging run.
I now realise why my sensor does not get any condensation. Last night with the fridge Peltiers set to -15.0C the fridge temperature was at -10.5C and the sensor was at 6.5C. Ambient temperature got down to 5.0C!
I have a heating strap around the lens just outside the fridge which keeps the lens at 24C. This is where the drop in filter is. This keeps the air around the filter above the dew point (I hope).
Now we all also know why nothing beats ICNR!
Any questions, and I will attempt to answer.
Bert
Also see this thread
http://www.iceinspace.com.au/forum/s...ad.php?t=43552
