photodiode noise

I have been thinking of try to cool down some op-amps and see how
much this helps noise. Some of the lowered noise amps
tend also to get hotter. Anybody try this?

l also work with some Hamamatsu photodiode cameras. These things
have bias circuits which also tend to have more noise with biasing,
however the newer cameras have a new bias circuit suppy thats putting
a 440 hz noise spike. I wanted to try cooling this camera, maybe
also get rid of the bias circuit. A little scary working wwith $15K stuff.

greg

 

I might add, I'm guessing it would be better to cool the photodiode rather
than the amp if one had the option, but there is also that gigohm resistor
around the amp.

grge

 

Have a look at some of the astronomy sites. They seem to cool CCD cameras to
reduce noise AFIAR.

Wim

 

Hi Greg

May be this one is your main noise source. Try to cool everyone alone and
you will see, wich one you have to cool.

Marte

 

Yes. The 'simpler' units, have Peltier coolers (often dual stage), giving
temperatures like -30C, while the professional units, at extreme, get into
things like liquid nitrogen cooling. It is the thermal _signal_ (dark
current), that is reduced. Most CCD's, will gain electrons from thermal
effects, at a rate proportional to the temperature. This will be specified
with a figure like 0.5e/p/s at 0C (an average of 0.5 electrons per pixel,
per second at 0C). This signal then (of course) has a 'noise' component
itself. Reducing the temperature, reduces the rate at which this signal
accrues, and with it the contribution from it's noise component. The
typical rate of change, is about a 50% drop in the signal, for every 6 to
7C drop in temperature. The noise, is approximately proportional to the
square root of the dark current. So dropping the temperature by (say) 30C
from ambient, gives about a twenty fold decrease in the signal, and about
a 4.5* decrease in the thermal noise.

Best Wishes

 

I will get a chance when I build a new project. The camera system has 512
individual channels of amplification and filtering. A little difficult to
check there.

greg