Re: temperature test

[Andrei Semenov <semenov@jlab.org>]

An average anode current of 1 uA looks as slight underestimation... If you
have 1V PMT output pulse (10 ns FWHM) on 50 Ohm load, you will completely
draw out an average current of 1 uA at the frequency of 50 kHz. To
provide close-to-be-linear responce of PMT at the peak current load (that
is 20 mA), the base should be supplied with significantly higher (at least
10 times?) current. And (of course) higher desired frequency means higher
current draw. (Do not forget that the PMT noise current should be
included.)

What kind of PMT base you discussing? If it's simple resistor base, the
average consumed base current is

U(HV)/R(base) PLUS (average anode current)

and the value is not small probably.

Talking about the test, Elton is right. You can flash LED on PMT
photocathode with the desired frequency, and set the amplitude of the
LED pulse on desired value (to have, lets say, 1V output PMT signal). And
you should see the rise of the consumed current from the HV supply as well
as temperature rise at high frequences (compared with stabilized
low-frequency case). The control of PMT anode current should be done via
measuring of the amplitude and shape of output PMT signals or might be
measured directly with precise ammetter. And at high frequencies, LED
might rise its own temperature, so it should be separated from PMT with
some distance (via optical link).

Andrei





On Thu, 31 Jan 2008, Matthew Shepherd wrote:

>
> Elton and Beni,
>
> Paul and I were talking about these bases -- see below.
>
> -Matt
>
> from Paul:
>
> One thing Beni probably isn't doing is putting them into the "power
> down" mode which puts the processor and CAN transceiver to sleep
> (while maintaining the HV).  Although he already isn't seeing a
> temperature rise, and this will make them produce even less heat.  I
> saw an email from Elton suggesting Beni provide a pulsed light source
> to increase the current draw, but I've always found that with
> realistic light levels and rates this makes no detectable difference
> in the power consumption.  I believe the average anode current is only
> something like 1 microAmp.  This is into 50 ohms, so the base is only
> supplying 50 microWatts to the outside world.  The incremental
> efficiency is something like 50%, so the base consumes an extra 100
> microWatts; insignificant compared to 100 milliWatts.
>
> In any case, this temperature rise seems like a very difficult
> measurement for such a small # of bases; he would have to thermally
> insulate the setup extremely well to see any temperature rise.  I'd
> just use the worst case power consumption of say 1/4 W per base,
> multiply by 2800 bases = 700 W that has to go somewhere.
>
>
> On Jan 31, 2008, at 9:12 AM, Elton Smith wrote:
>
> >
> > HI Beni,
> >
> > You should have a source (anything will do) that runs the pmts at the
> > rates expected for the experiment.
> >
> > Cheers, Elton.
> >
> >
> >
> >
> > Elton Smith
> > Jefferson Lab MS 12H5
> > 12000 Jefferson Ave
> > Suite # 16
> > Newport News, VA 23606
> > elton@jlab.org
> > (757) 269-7625
> > (757) 269-6331 fax
> >
> > On Thu, 31 Jan 2008, beni zihlmann wrote:
> >
> >> Hi Tim,
> >> here some pictures of the temperature test setup. I hope that gives
> >> a better idea what this test is worth (or not).
> >> I checked on Elke's concern that the PMTs might not really draw
> >> that much power because if there are no signals then the PMT does
> >> not have to do any "work" hence less power consumption.
> >> I do see signals because there is enough light leakage that the PMT
> >> does some "work". However I found out that exactly the PMT/BASE where
> >> I put the temperature probe in stopped working at some point. So
> >> I put the temperature probe into the hole of a pmt that works for a
> >> fast test and also there I see no  obvious temperature change.
> >> Also the test with putting a  cover  (garbage bin) over the whole
> >> assembly
> >> is not as good a check as I thouth because that shields also the
> >> light
> >> and the PMTs have much less to do.
> >> Any suggestions what I might "improve" to get a kine of "worst case
> >> szenario"
> >> are appriciated.
> >>
> >>
> >> cheers,
> >> Beni
> >>
> >
>
>