[Halld-pid] PMT test results question

[Alexander Ostrovidov]

Hi Beni,

I think you are right. Indeed, our Am241 source does not produce
that many photons, and I do see a Landau distribution which is cut
by threshold instead of Gaussian far from threshold. And yes, I'm
getting somewhat higher efficiency values for lower gain tubes
due to the effect of threshold cut. Nevertheless, I still consider this
to be a useful method of finding any tubes which significantly deviate 
from general trend of other tubes but not as a very precise 
measurement of quantum efficiency.

Paul is looking into getting optical filter or diffuser so
we can apply normal voltage to LED instead of suffering
from instability of its pulse near threshold which we
observed in the single photo electron gain measurements,

Sasha


On Wednesday, October 31, 2012, Beni Zihlmann wrote:
> Hi Sasha,
> upon further discussion with Elton I think I now understand the
> problem better.
> the main issue is that the Am241 source produces only a few photo
> electrons. So the ADC peak you see is a Landau distribution close
> to the pedestal. Because you have to trigger on the the same signal
> you are looking at the threshold relative to the signal hight is important.
> So when a PMT has a much larger gain then a constant threshold will
> lead to a different result as compared to a PMT with a low gain because
> the threshold will cut into the rising edge of the landau distribution. That
> is must likely the reasons for the still rather large variation you see in
> the quantum efficiency of the different PMTs.
> 
> Generally I think the Am241 source is not a good tool to measure the
> photo electron efficiency because of the too small light output. I think
> just using an LED with sufficient light output and monitoring the stability
> of it will do a much better job. Then you can fit the ADC peak with a 
> Gaussian
> and the width divided by position will give a much better result.
> 
> cheers,
> Beni
> 
>