>
>
> As I understand it, by providing a threshhold against which pulses are
> judged, the idea of photon counting is to reject the thermal counts
> (low energy) and accept genuine photon events (high energy).  The
> consequence of this is inevitably that some genuine photons are in
> fact thrown away - total signal will be reduced - but even more noise
> is thrown away so that S/N ratio is improved.
>
The only problem is that the PMT cannot distinguish between a
photoelectron and a thermally generated electron which is emitted from
the photocathode. Photon counting will eliminate noise from the
electronics, provided that the total noise of this kind is less than
the threshold for pulse counting. It can also eliminate most noise
from electrons which are thermally generated off the dynode (or at
least most of the dynodes) provided the PMT gain (ie electron
amplification) is set correctly.

When next at my scope, I will try to measure this by accumulating a
dark image in the photon counting mode.

--aryeh
Aryeh Weiss                          | email: [log in to unmask]
Department of Electronics            | URL:   http://optics.jct.ac.il/~aryeh
Jerusalem College of Technology      | phone: 972-2-6751146
POB 16031                            | FAX:   972-2-6422075
Jerusalem, Israel                    | ham radio: 4X1PB/KA1PB