We are working on a spectrometer, which uses many LEDs as light source and CCD as detector. We are worried that temperature variations might make things tough for us. For eg. spectrum shift in LEDs, dark noise in CCD and other other electronic components.
We want to make the device portable and cost-friendly. Hence, peltier and water cooling are not an option. I once heard of a technique where they choose electronic components based on their effects on increasing temperature. So as we increase temperature, if the contribution of component A is +x, contribution of component B will be -x, thus we still get the same result. But still this method is not very clear to us.
Please help in this regard.
Temperature effects on CCD sensors are common issues and one can compensate for it in software. Dark current and hot pixels are things you can measure in functrion of temperature. So take dark, flat and bias frames and do some image processing. Up to a certain level this will work. Also for the emission of LEDs I am afraid that you can do some preparatory work to model and compensate in function of measured temperature.
I agree with the issue of Peltier. Peltier requires a lot of current. If your device is supposed to collect data for hours and one needs to be mobile (no power net connection but a battery) this is heavy. In my own project a mobile telescope set up should allow autonomous image recording for 7hours in the field. If my Peltier requires 1 or 2 A at 12V for 7 hours plus the telescope mount plus some credit card computers plus a network router and don't forget the camera. You need a good car battery or even more.
- Badges
- Science topic