PMT to PC audio

Been typing some code, the guys in the yahoo group for spectrometry have a MS windows program to display spectra with the PMT connected to the PC soundcard.

So, as I burned that MS disk, I wrote some Linux code.

The interface is really simple, just an emitter follower from the PMT anode (cathode is at -1284 V in this test): +9V 470n |-----||------- c | PMT anode |-------||----------------------------- b NPN /// | 4n7 | e BC547-B 10 m screened audio cable 1k 3M9 |----------------------------===============//============== PC soundcard mike input | | 1k /// +9 - 100k ---- 100k --| | | /// ///

Why these values? That is sort of what came up when I grabbed in the box. It is always an art to make something with what you get :-)

Anyways, this was a first test, to see what the pulses would look like. I added the scope display as I had no clue (tm) what was going on IN that soundcard, and to write the rest of the soft I needed this. This was still the old xy plot version with only 256 points (for 8 bits via RS232).

formatting link

See how nice the pulse gets stretched, and even keeps polarity (negative at the anode). That spectrum is bogus, just background from a large BGO crystal for testing.

After typing some more code adding features, one of the things I added is the yellow line in the scope display, that is the slice level for the pulses, also this shows it at 10 x gain (150 mVpp fsd). I did not need the mike boost on the soundcard, just mike slider full open.

formatting link
it is again the same background, but now using the full 16 bits from the soundcard, so more resolution, not that BGO has that. The 'blankout level', is set just above the thermal?? noise. remember without the crystal the line is flat (no signal) no PMT noise, all light flashes.

If anybody is interested in the source code, they should ask here. This code will probably change constantly anyways... Oh, the above thing runs from a 9 V battery... of course it should be build in somewhere, it now sits an a piece of A4 paper on the table.

So that was the coding exercise for today. We can write anything you want in Jan Panteltje's restaurant.

Reply to
Jan Panteltje
Loading thread data ...

the yellow line in the scope display,



ll light flashes.

The "dark current" in photomultiplier tubes comes - in large part - from the photo-cathode. With blue-sensitive photocathodes it's mostly from cosmic rays hitting the photocathode - twenty or thirty a second is a figure I've seen - while with red-sensitive photocathodes more of the electroncs are generated by thermal energy in the photocathode and can be made less numerous by cooling the photocathode (which can be tricky, particularly if you want to avoid condensation on the photo- multiplier window and nearby optical glass).

The dark current coming from the photocathode is single electrons, just like the photo-generated signal, and they get amplified by the dynode chain in exactly the same way as the signal electrons, and produce exactly the same pulse height distribution at the anode.

There can be some spontaneous electron emission from the dynodes, and that can produce smaller pulses, but they are scarcely worth rejecting.

-- Bill Sloman, Nijmegen

Reply to
Bill Sloman

On a sunny day (Sun, 18 Dec 2011 15:43:36 -0800 (PST)) it happened Bill Sloman wrote in :

Yes, but I am still way below where the PMT dark current becomes an issue. I tried to find that point by running the PMT in absolute darkness (ehh well 'properly shielded'), and it starts to give signal in dark at about 1750 V or higher. This test is running near 1250 V, and all you see in that picture is light flashes from the BGO crystal Although the Thorium fuel rod shipment for my fast breeder arrived, and may contribute some to the background radiation, the spectrogram shown is still bogus (that is why I used that word), because I have not implemented the peak detection in the software yet (except for that dB meter). What happens now is this, in the pulse drawn below there are samples taken of the PMT pulse at each 'dot', that means for ONE pulse you get a lot of samples with a lower amplitude, these now fill up the bins too, in fact many samples with lower amplitude for one correct sample with the right amplitude. So I have to add those software lines that look for the peak and use that sample only.

That will clear up the display and then I hope to make a nice Thorium background spectrum. It was 12 o'clock last night so I was happy the proof of concept worked in the software / hardware. I will add the peak detection later, and if it works publish some spectra BTW there is also a 10 nF cap in series with the soundcard mike input I forgot to draw, this is needed as the mike input from soundcards output a few volts (3.3 in my case) so electret mikes are automagically powered. Of course you could use that 3.3V to also power a JFET preamp for the PMT, but PMT has already all that power supply. Could not find any BF245 FETS.. cannot find any of my nice dual gates either, no idea where those that is why a BJT as emitter follower. Hope I did not accidently throw my FET supply away :-) BF245 are nice to have around.

... .............................. | . | . | . | . | . | . |. |

By 'thermal' I ment thermal neutrons hitting the crystal lattice. Or thermal motion in the crystal lattice perhaps, whatever. So I wonder if cooling the BGO crystal (not the PMT) will reduce pulses, Many more experiments are there for the curious.

Reply to
Jan Panteltje

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.