So I made a detector from a R7400U + CsI(Tl) from ebay

[stamasd]

These little things have intrigued me for a while
https://www.ebay.com/itm/196522968379

They have a subminiature PMT Hamamatsu R7400U coupled with a small CsI(Tl) crystal, reportedly 10x10x40mm and were originally intended for use in "radiation pagers" for the US government I think.

Eventually curiosity took me over and I bought a couple of them
This is the pinout on the small PCB attached to the tube


I made an 81M voltage divider for them



Attached it to the green PCB after taking the wires out and replacing them with female pin sockets, and a small PCB on top with matched male pins onto which I built the divider. It's not pretty but it works.




(I've since reinforced the assembly and properly insulated it)

And here is what it does in PRA, using a GS-USB-PRO at 800V, source is a 12x12x22mm block of LYSO; you see the 2 peaks of the Lu-176 at 202 and 307keV. Resolution is not great, but for the price I'll take it.

[Conor Whyte]

good greif that is small ! Nice work :)

[stamasd]

One problem with this detector is that every couple of seconds it generates some very intense pulses, see picture below. This happens with or without a source nearby, i.e. when measuring background.



It actually happens very regularly at a rate of 30 of these pulses per 60 seconds. I'm wondering if it's something internal to the detector itself.

[Madmax]

I tried this detectors, remember resolution around 13%. No any generation of unwanted signals noticed. I can assume this is some interference coming from outside. Try decreasing voltage to, say, 500 V, and if unwanted pulses still present - it have external nature.
One more thing: glass insulation on R7400 leads is small, breakdown possible. Try washing carefully in isopropyl alcohol and drying - all the system including divider.

[stamasd]

Thank you for this suggestion.
I was actually suspecting that it had something to do with insulation/dielectric breakdown since this only started after I had applied insulating tape over the assembly. Foolishly enough I put a layer of kapton tape directly over the exposed contacts on the green PCB connecting to the PMT's pins. Not thinking that the dielectric strength of the tape's adhesive may not be adequate for hundreds of volts. :)
I have since cleaned both PCBs (the green one and my homemade one) first with 99% IPA, then with a chlorinated hydrocarbon-based contact cleaner, and dried them thoroughly. The rogue pulses are gone now. :)

[stamasd]

I reworked the voltage divider (actually I made a whole new divider for the other R7400 detector I have), this time with 3Mohm resistors x8 plus 1Mohm final, so total 25Mohm. It seems to work better this way.

This was acquired at 800V with a LYSO source


And this with a Cs-137 source


The calculated FWHM for the 662keV peak of the Cs-137 is 10.2%; the FWHM for the 307keV of Lu-176 is as expected higher at 13.3%. Overall seems improved over the 81M divider variant. It does seem to favor the lower energy gammas over the higher ones - see how small the 662keV peak is compared to the others. Not surprising with such a small scintillator. But it seems quite usable nonetheless.

I have to find a way to pack the divider better because I built it in air, spider-style. :)

[Rolin]

Hello. What is background CPM for this detector with 800 V ?