Using Geiger tube in proportional mode

A proportional counter is a fairly heartbreaking thing to use for spectroscopy--even a sodium iodide scintillator is much much better.

OTOH for a hobby project, it would probably work OK.

ISTM that you'd need to prevent ions from migrating around inside somehow, unless you're okay with a fairly long dead time after each Geiger pulse.

Cheers

Phil Hobbs

Yes, this is for an entirely hobby project. No great spectral resolution is required; in fact, anything resembling the right spectrum would be very rewarding. It is an off-label application driven by curiosity.

By necessity I am going to use only the readily available weak sources with pretty well-defined chemical composition: uranium and thorium, perhaps americium too. So the dead time should not be a problem.

Best regards, Piotr

Hi Phil, just related - have you experimented with NaI and some of the photodiodes etc. you seem to do a lot of work with? My current understanding is that PMT-s are still better by quite some margin but then I don't give NaI much thought (although our netMCA is HPGe grade we do have customers who use it with NaI though).

====================================================== Dimiter Popoff, TGI

Geiger tubes have a really tiny wire, to make very high e-fields close to the wire for lots of multiplication. I recall that ion chambers have more, bigger wires, or flat electrodes. That difference probably affects pulse-height spectroscopy.

SBT10A is interesting, a multi-wire geiger tube. It would probably work OK in proportional mode, with maybe a little multiplier gain.

Are the wires brought out to the connector individually? Looks like it.

"Lite salt" sold in the supermarket is half KCl, which is a decent test source too.

Cheers

Phil Hobbs

Nothing can touch a PMT for dark counts per unit area--they're five or six orders of magnitude better than an APD or SiPM / MPPC.

A typical 100-um diameter APD has about the same dark count rate as a _four_inch_ PMT.

Plus you gain another factor of at least 2 by using coupling gel between the crystal and the PMT face.

You'd think it was right around 2 because you win etendue by the square of the refractive index, but it's actually better than that--the extra light at high angles gets totally internally reflected at the photocathod/vacuum surface, so it gets at least two passes through the PC, which raises the quantum efficiency considerably.

There are tricks to do even better, e.g. by using prism coupling to make light rattle round inside the faceplate till it gets absorbed. With a negative electron affinity (NEA) photocathode, that'll get you almost

100% QE over a wide bandwidth.

Cheers

Phil Hobbs

Thanks Phil.

There are liquid scintillators too.

Yes, they are. It opens several interesting possibilities, e.g. shielding half of it to get beta and gamma signals separately or running some part of it in avalanche mode and the rest in proportional mode. Spectral sensitivity is also much better than typical metal tubes -- some specs say it should react to gamma as low as 15keV.

The digital part would be boring, just a bunch of counters. But it is the analog front-end that could make it possible to go way beyond the specs. Looks like a lot of fun in my spare time.

Best regards, Piotr

There's gas recipe differences, but that's because you want fat (time-duration) pulses for proportional but not Geiger. The big irritation, is that the proportional region is relatively small, it takes regulated V source, and only part of the (usually cylinder-around-wire) field region has best sensitivity.

But, the construction, if not the gas mix and bias, are about the same. There's a lot of literature available online, like

and spectroscopy is possible... as long as your gas mix is dense enough to actually stop some of the rays in question. Some variant of multichannel analyzer is needed, of course, and a test source for calibration.

Not sure about beta spectroscopy, unless you put the beta emitter inside the gas-containing tube.

I designed some wire chamber electronics for CERN. Each chamber was a big thing made of parallel sheets of aluminized mylar or something, with a zillion collector wires inside. I think there was some ion multiplication but not geiger mode. The wires were close enough that we could measure the pulses and interpolate the track locations between wires

It was a massive flood of data from multiple planes, the object being to tease out particle tracks. We did "progressive enrichment" of the data: ecl gates, then FPGAs, then software. The final output was a single bit, PUBLISH.