Help setting up PRA

[tnielsen]

Hello everyone,
I've just finished building a scintillation probe using an Phillips 3inch round photomultiplier tube and a 63mm round NaI(Tl) crystal. I've hooked it up the GS-USB-PRO and I'm currently feeding it around 650 volts. When I run the pulse shape acquisition I get these results, which to my eye seem good.

After about 1000 pulses I saved the pulse shape, changed "Use shape method" to True, and set "Gaussian correlation" to true, and then I ran a data capture where I placed a Thorium source next to the detector. I got a histogram that is compacted into the first 5 arbitrary units on the horizontal axis and I'm not sure why this is the case.

Anyone can help me or know why it's so compacted? I couldn't figure out a way to zoom horizontally into the histogram, not sure that's possible.

[Sesselmann]

Trevor,

Thanks for posting the image...

Sample Rate
Change your audio input to 384 kHz sample rate 16 Bit left Channel, this will give you the best resolution. Note: the pulse shape will look more stretched out.

Use Shape Method
The use shape method box must always be ticked when you are collecting a pulse height spectrum.

Pulse Shape Acquisition
Before you collect your spectrum you need to train the software by doing a pulse shape acquisition. Run it until you have collected around 1000-5000 pulses and then stop. This will bring up a window with 64 samples being the sample points for the mean pulse shape. - Just click okay.

Theory says pulse shape in invariant regardless of pulse height, so the software will use the mean pulse shape to discriminate pulse pile up (PPU)

Ready set Go!
Now you are ready to Start Data Acquisition.

Best of luck....share the result here.

[tnielsen]

Hi Steven,
I've changed the sample rate to 384kHz and redid the pulse shape acquisition, I collected about 4000 pulses before stopping and saving the shape data. Just tried to take another spectrum of my Thorium sample and got the same sort of heavily compressed histogram. Thoughts?

Edit: I also noticed that during the pulse shape acquisition the pulse widths are around 100-140 microseconds versus when I take a data spectrum its a very tight cluster around 50 microseconds, do you know why that would be?

[tnielsen]

I've figured out one of my issues, I have a lot of ground loop noise on the audio input that I need to get rid of. Looking at USB isolator ICs I see the vast majority only support USB 1.1 Full Speed (12Mbit/s), is the GS-USB-PRO a USB Full Speed device? I would assume so, since 384000 samples per second at 16 bits per sample is only around 6Mbit/s. However, the cable it shipped with is certified for USB 2 Hi-speed (480Mbit/s), but I assume the device doesn't actually need that much bandwidth?

Thanks,

[Sesselmann]

Trevor,

It looks like your gain is way too low, you need to boost the voltage and up the gain.

You have multiple adjustments which all increase pulse height.

1) Bias voltage
2) Preamplifier gain
3) Pulse shape
4) Boost gain

The optimum gain is when the K40 peak at 1460 appears around 50 arb.u (middle of the range).

The signal to noise ratio will automatically improve as you increase the bias.

Good luck..

Steven

[tnielsen]

I seem to have identified an issue when I look at the incoming audio in Audacity, I see a very noisy spectrum with a few large pulses, but when I put my fingers on the bayonet of the SHV connector the noise disappears.

noisy audio.png (14.25 KiB) Viewed 1185 times

This is my schematic for the PMT wiring, I wired a panel mount SHV connector to the anode and cathode and have a SHV-SHV RG-58C/U cable connecting it to the GS-USB-PRO.

scintillator hv chain.png (18.35 KiB) Viewed 1185 times

Any advice on how to get rid of this noise?

[tnielsen]

Ah, my mistake, the noise is actually from a light leak from the room light onto the photomultiplier tube. I will add more electrical tape and see if that fixes the issue.