[Halld-pid] PMT test results question

Fri Oct 26 21:41:09 EDT 2012

Thanks to Beni and Sasha for those detailed comments/reports. Looks like 
the situation is getting clearer. :-)

On 10/26/2012 07:14 PM, Alexander Ostrovidov wrote:
>> Hi Paul,
>>
>> sorry that I missed the meeting. Mark explained to me the situation and
>> I have some
>> remarks and questions because I think something is completely wrong with
>> the ADC
>> spectrum using the Am source.
> Hi Beni,
>
> Thanks for taking a look into this test. Let me start by saying that things
> are better today than yesterday when we were getting a factor of 4
> inconsistency in anode sensitivity between tubes. What apparently has
> happened is we overloaded NIM crate power supply with too many modules.
> For whatever reason, this led to variations of about 20-30mV in baseline (offset)
> of the signal which I didn't notice initially. This, obviously, didn't go well with
> a 25mV discriminator which is used to form ADC gate. We pulled out a few
> unused modules and the offset went back to zero. Now all tubes measured
> so far show relative consistency within 30% when their readings are normalized
> to their listed anode sensitivity. Check the last column here (standard gluex
> user/pass): http://hadron.physics.fsu.edu/TOF/database/PMT_am.html
>
> I absolutely agree with you that one would expect a relatively narrow gaussian
> from alpha source. Unfortunately, that's not what I see  with the scope. Signals
> (straight from PMT into the scope with 50 Ohm terminator, and I put the scope
> into 5sec accumulation mode) are spread from about 50mV to 300mV with
> maximum closer to the lower values. Their FWHH visually varies from 5ns to 15ns.
> It's no wonder with such variations that ADC spectrum is nowhere near to being
> monochromatic.
>
> I couldn't find anything wrong with ADC setup. ADC gate is about 160ns.
> Signal comes about 80ns after the start of the gate and nicely fits inside it.
> Signal visual spread on the scope roughly corresponds to the spread of
> the ADC spectrum. The most probable signal corresponds approximately
> to about 10 initial photo electrons at the cathode.
>
> At this point, I'm  ready to suspect that something is wrong with Americium
> source itself. I can see a small silver metallic button on the other side of
> the scintillator but I'm not sure if there is something that stops (or slowdowns)
> alpha particles before they enter the scintillator. May be, what I see is such
> slowed down alphas. Or a conversion of 60keV gammas instead of alphas
> which are also emitted by Am241.  Or betas from its decay products.
> At least, what I do know is:
>
> a) The rate, from whatever it is, does go up to 2000Hz (with 30mV threshold)
> comparing with 10-30Hz rate from PMT with nothing inserted, and 20-40Hz
> rate when a comparable size scintillator (with no source) is attached to PMT.
> Just from the probability, it's very unlikely that we trigger on noise or cosmics.
>
> b) ADC spectra do look differently in these 3 cases Take a look here
> http://hadron.physics.fsu.edu/TOF/pmt_am/ZQ2263/
> File noscint* is with nothing attached to PMT, file nosource* is for
> a passive scintillator, and am10x* is for a scintillator with Am source.
> >From the difference of these spectra as well, it also doesn' t look like we trigger
> on cosmics or PMT noise that much.
>
> We'll take a closer look at the Am source with Paul on Monday. Meanwhile,
> even if this is not a pure Americium source, I guess, I can still treat it
> just as a consistent, albeit puzzling, source of light. Therefore, relative
> comparison of PMT anode/cathode sensitivity can still proceed as long
> as light illumination is stable from tube to tube. At least, current 30% is
> much better and almost acceptable comparing with factor of 4 (or even
> factor of 10 from   UV LED rates) I had before.
>
> Sasha
>
>
>> 1) the Am 241 generates 5.6 MeV alphas. This alphas are mono-energetic
>> and should
>>        produce a lot of light in the scintillator and therefore a very
>> large signal in the PMT
>>        if you put this little scintillator directly in front of the photo
>> cathode.
>>        -> so have a look on the scope how the signals look like. They
>> should be huge if you
>>        use 1750V
>>
>> 2) Because the alphas are mono-energetic the resulting energy spectrum
>> in the ADC
>>        should be a peak. Or maybe two peaks depending of the pureness of
>> the Am241
>>        source and the resolution of the PMT/scintillator system. This
>> peak should be rather
>>        sharp, high above the pedestal and Gaussian. If any tail this tail
>> should go to lower
>>        energy not higher energy. These are the main reasons why I think
>> the ADC spectra
>>        you see are completely off and something in the setup must be off.
>> My first thinking
>>        is that you look a noise and the actual signal is at much much
>> higher ADC values.
>>        -> again that is why looking on the scope at the signals and in
>> particular large signals
>>        should be done.
>>
>> 3) Because I say that the Am241/scintillator source produces so much
>> light it might be that
>>       1750 is far to high Voltage, and for sure you do not need
>> amplification.
>>       Note: 5.6 MeV from the alpha particle is deposited in the first 1mm
>> of the scintillator and
>>       the scintillator in total is not that think so there is hardly any
>> attenuation and all the light
>>       is seen by the photo cathode. => lots and lots of light => huge signal!
>>
>> 4) you trigger on the same signal as you are looking at. That means you
>> should never see
>>       a pedestal. However the data show a pedestal => that does not make
>> sense.
>>
>> 5) the peak in the ADC is at a similar location above pedestal as the
>> single photo electron peak
>>       is at the same HV and amplification. Again that does not make sense
>> because from the
>>       alpha source you should have lots and lots of light.
>>
>>
>> I hope this gives you some ideas of what you can look at in your setup
>> to see
>> what might be going wrong.
>>
>>
>> cheers,
>> Beni
>>
>>
>> _______________________________________________
>> Halld-pid mailing list
>> Halld-pid at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/halld-pid
>>
>>
>
> _______________________________________________
> Halld-pid mailing list
> Halld-pid at jlab.org
> https://mailman.jlab.org/mailman/listinfo/halld-pid

-- 
Mark M. Ito
Jefferson Lab
marki at jlab.org
(757)269-5295