|
Gary the ideal would be to bleed in pure dry nitrogen or dry air from a compressed 'bottle'.
I would get one of those fish tank airators and pass the flow over a cold Peltier surface to get rid of the majority of the water vapour and then through a container with silica gel or molecular sieve pellets. Both of these can be regenerated by heating in an oven. A vacuum oven works even better. The Peltier cold trap can be regenerated by allowing it to warm up and drain any water.
The flow rate does not need to be high only high enough to produce a dry positive pressure inside the camera. A focal reducer or filter immediately in front of the camera would really help to stop 'warm' moiste air from getting to the sensor. Any slight leaks would be easily compensated by the internal positive pressure and low flow.
I would try at first to keep your sensor temperature above the dew point.
I have measured the temperature of my sensor in the 5DH and it is 17C above ambient when continously exposing.
When my fridge Peltier heat exchanger is set at -10C the fridge interior is at -5C and the sensor is at 12C. Without the fridge and at 20C ambient say the sensor would be at 37C!
Since you are cooling your sensor directly I would start off at about 15C for the sensor and slowly work down until you get condensation.
You will be amazed how much the thermal noise is suppressed even when running your sensor at 15C.
Bert
|