Page 1 of 2

finding plateau voltage

Posted: 19 Jun 2018, 21:13
by keuleeule
Hi guys,
beeing forced to stay at home after an operation means time to play around with my tools since a long time.
I got a beautiful 3" detector by Luuk, thanks so much. Now its time to play with it.
But honestly I strugle or beeing not sure why my plateau voltage is that different to the suggested one...

Luuk stated fortunately different working voltages to his detector NaITl around 550V to 920V, his plateau voltage for Cs137 was around 730V.
Trying to find the best plateau voltage with Cs137 now with my GS2000-Pro brings me to 1550V...
BUT I realy cant find a plateau, as the voltage can rise to 1888 and not more. There is no significant changes above 1580V
WHY is it that high?
Does it all depends on the divider? -> for sure, but I tried two configurations,
first the PMT Steven is offering ... -base-14-2 with 1M each resistors in total 12M

then the recommended from Luuk, with 470k each, so just half of it.

Stevens configuration brings me to 1550V plateau at max 1888 is adjustable
Luuks to 1550 whereas 1690 is the maximum achievable voltage.

How can I proceed? There must be something wrong...
in both ways I uses two wire connection.
Thanks Kolja

Re: finding plateau voltage

Posted: 20 Jun 2018, 05:37
by luuk
Hi Kolja,

I am not sure hat you want to achieve with this experiment but the detector you have cannot work with an high voltage of >1400Volt it will destroy the pmt if you do!
With every detector I would recommend to use the lowest possible voltage, so it gives enough signal to do your measurements with and gives a linear signal.
A plateau voltage is an old way of testing. used for gm counters or detectors that just are use for counting, I don’t see any purpose in using it with a scintillation detector used for spectroscopy.
The high voltages I used in the test-sheets are giving an output from around 500mV in to an oscilloscope with an impedance of ±1MΩ.
You do not get a better resolution by raising the high voltage! only thing you will get is aging of the pmt, lots of noise and finally destruction of the pmt.
You have to use an high voltage that gives a linear output, and that is different for every pmt/detector I would strongly advice to use an high voltage close the ones on my test sheets if used than you will have an optimal resolution and linearity.


Re: finding plateau voltage

Posted: 20 Jun 2018, 06:27
by keuleeule
Ok, Luuk,
thing is I do not get any pulses registered by PRA if I use lower voltages...
So I try to set up audio again.

I just did it like I used to in my work several years ago, in which I had a Scint counter and plotted count rate during increasing voltage to find the plateau...

I will use your voltage now and try to get pulses out of the GS2000pro.
thanks, Kolja

Re: finding plateau voltage

Posted: 20 Jun 2018, 08:08
by keuleeule
well, I´m not able to get a signal out of the GS2000-pro if I use the recommended voltage of 730V.
PRA does not get anything, Theremino tells me that the audio signal is too low.
So I remember that there was a adjustment possible inside the GS2000, so I will search for that now.

Re: finding plateau voltage

Posted: 20 Jun 2018, 08:19
by keuleeule
Luuk, you stated that you just de or increase the voltage to that point you got 500mV signal output, correct? So I will hook up my osci and check what I get, maybe that is the easy way, then changing the pre-amp in GS2000 to get good audio signal.

Re: finding plateau voltage

Posted: 20 Jun 2018, 14:32
by luuk
Hi Kolja,

If you got an oscilloscope that is always good to use, at least you know what kind of signal is coming out of the detector or amplifier, that gives you a lot of information.
That would be also the first thing I would do if a detector is not working, I don’t know what the software is expecting to see as enough signal to work with, but I guess that you can adjust the gain somewhere with the software.
I don’t work with software mca’s so I have no experience with them, I cannot help you with that, I have only hardware mca’s.


Re: finding plateau voltage

Posted: 20 Jun 2018, 14:59
by Steven Sesselmann

A good starting point is to check the PMT manufacturers recommended voltage. A common voltage range is 650-750 volts.

Next step is to check the total resistance of your voltage divider, using a multimeter you can usually just measure across the BNC connector, but if the impedance is very high you might have to measure each resistor. A resistance higher than 10 M Ohm is best. If you have a divider with very low resistance the GS-PRO will read incorrect voltage.

The digital voltage display on the GS-PRO measures the voltage before the 1M load resistor, and has been calibrated with a 1 Gohm load, so the voltage on the anode will always be a little less than indicated.

Follow these steps for best results;

1) Calculate correct voltage for the pmt and set GS-PRO to this voltage
2) Set sound card to Line In level 100% volume
3) Adjust preamplifier on the GS-PRO board so Cs-137 (662 keV) sits at or around 30 arb.u (using PRA)
(If you don't have Cs137, just look for the K40 peak and set it to 50 arb.u)

That's about all..

Remember V(real) = V(display) * (a)/(a+b)

Where a = divider resistance and b = load resistor resistance


Re: finding plateau voltage

Posted: 20 Jun 2018, 23:00
by keuleeule
Hi all,
as I got my detector from Luuk last year, and his Cs137 measurement was taken at 730V, I used 730V now.
As divider I used the one from Steven (the assembled one) I have a Cs137 source.
Using the GS2000-pro
I have two wire configuration (in the divider and in the GS2000-pro)
Sound Blaster X-Fi HD USB Line in at 100%

I got lets say 1 klick per second, but a lot of background noise (no crackling, hissing) at very low volume. This background noise disapears if the Cs137 source is away. The klicks are still present.

In PRA I used:
0.3 height threshold
tolerance threshold 5
correction 0
but I have to use boost 10 to count , otherwise PRA does not count any.
use shape tolerance is enabled

with that I got 112 pulses in 60sec (pulse shape aquis). Compared to your video Steven, in which you got 26655 in 53 sec, its nothing...
ok ok, its all about the activity of the source...

If I run now the configuration like in your video, I do not get a spectrum...

If I increase the voltage to 1350 I will.

What I did not performed so far is an manual adjustment of the GS2000-pro, as ALL my detectors show the same results.
Even the 30y old ones.

Are there some testpoints I can follow in the GS2000-pro to check if its the box, or should I try to adjust the volume pre-amplifier first?
Thanks, Kolja

Re: finding plateau voltage

Posted: 21 Jun 2018, 03:53
by luuk
Hi kolja,
Did you already looked with the ocsilloscope at the output of the detector, what do you see?
can you show me an picture of the screen?

Re: finding plateau voltage

Posted: 21 Jun 2018, 06:02
by keuleeule
Well, as Stevens PMT socket is quite small I used my own like Luuk recommended to me, with "Cath - 1M - 470k - 470k - 470k - 470k - 470k - 470k - 470k - 470k - 470k - 100k - AN"
With adjusted 730V on GS2000p I got AN-CATH measured 583,7V
"Cath - 1M - 470k - 470k - 470k - 470k - 470k - 470k - 470k - 470k - 470k - 100k - AN"
Cath----109V - 158V - 205V - 252V - 300V - 349V - 401V - 456V - 516V - 583V

At adjusted 730V on the GS2000 I get randomly a pulse as I told, I tried to catch one.
The normal signal is 200mV and peaks like described around 550mV, quite randomly even with source.

I will try to catch the signal in a better way.

BTW: I had still the Audio Plug in my Soundcard !
and will change that now. :)