Making a Scint Probe

Under construction . . .

First Test:

Thanks to Andre I have 2 Hamamatsu R1166 PMTs. Thanks to Luuk I have a 27x27x45mm CsI(Na) scintillation crystal. Thank you both.

I made a voltage divider for the PMT from a Hamamatsu schematic and attached it to the PMT socket. (long wires temporary)

It was a "3 wire" configuration meaning that there is a separate signal wire instead of putting the pulses on the HV line.

(Later it was modified to a "2 wire configuration" with the signal on the HV line.)

Attached the crystal to the PMT and put it in a temporary box to test it. Go Juicy Fruit!

After struggling a bit, I eventually got pluses - yippee!

This scope shot shows the output with a thorite sample around 3cm from the probe.

Pulses range from 25 to 250 mV.

Occasionally there is a huge pulse. I assume these are muons. At least I hope so.

The PMT is running at 1000V.

Note: This scope shot and the following spectra were actually taken after the crystal was cut as described below. It's size was 27x27x38mm.

Some sample spectra (click for larger image). These compare well with those from the Bicron probe on the Gamma Spectroscopy page..

(above) I calibrated the PRA to the 5 peaks of thorite (see the Gamma Spectroscopy page for more on this.)).

Just for fun the thorite spectra above with the GGV2 on the DSO Nano.

(I may need to use a different splitter or set some parameters to improve this.)

On the left is a more recent spectrum of a uranium ore sample taken on the 38mm probe with the 110M voltage divider. It's curious how similar it is to the Ra test disk spectrum above.

Cutting the Crystal:

Since I have two PMTs and one rather large crystal, I decided to cut off a 5mm slice to make a low energy gamma probe for possible XRF work in the future. I thought a Japanese pull saw might work best, in part due to it's thin kerf.

If you do this take precautions! CsI is not something you run into every day and that makes it more likely to be a bad thing in my opinion.

.

I wore rubber gloves and a respirator and carefully removed all waste.

The crystal cut easily but slowly. I decided to round the 5mm slice to fit the PMT. I had a big sheet of emery paper with an adhesive back - the kind used by floor sanders and available where they rent them. When the paper clogged I'd wash it off and let it dry. I didn't think it was a good idea to wet sand it.

Then I followed Luuks guide to matting and polishing the sides of the crystals.

From the original 27x27x45mm crystal, I wound up with a 27x27x38mm and a 20x5mm disk (only 2mm lost) . . . (second photo with a UV penlight)

Not perfect but good enough to get me to the next stage.

Building a probe

First I rewired the original dynode string to a "2 wire configuration" to make it easy to interchange this probe with my others. I also wanted a voltage divider with a much higher total resistance.The resistor values used in Hamamatsu's dynode string are only 330kΩ. This makes the resistance of the whole thing only about 3.8MΩ. So 263uA at 1000V. Although my bench power supply can handle that load, I needed something like 110MΩ (9uA at 1000V) so it will work with my Bicron Analyst for counting. My first tries with a 110MΩ string had a greatly reduced output until a friend pointed out that R1 should be about 1.5x the other resistors. A schematic of the modified dynode string I used is here.

This time the voltage divider was mounted directly to the PMT socket. The 5mm CsI(Na) slice was mated to the R1166 PMT with 20,000 wt silicone grease, wrapped on Teflon tape, and electrical tape and tested in a cardboard tube.

Below are the first spectra I'm getting using the 5mm crystal. Since the same PMT is also used for the 38mm crystal it is interesting to compare the results with the spectra above.

(compare to 38mm thorite above)

(compare to 38mm Ra-226 above)

Pre-Made Voltage Dividers

I stumbled on some Hamamatsu voltage divider sockets on eBay. The were not a stock Hamamatsu product, but they do work nicely with R1166 PMTs. They are "3 wire" configuration types with a separate signal wire and have a fairly low total resistance of 4.2MΩ. It's not that I don't enjoy building dynode resistor strings, but . . .

Using the same crystals, PMT, and VD for both the 5mm and 38mm size crystals allows for some nice comparisons between the two sizes as shown above.

All that was left to do was to hose both probes in a cardboard tube. I found the BNC to screw terminal adapters handy for this type of construction.

At some point I may make a nicer housing and also post some more spectra, but this should wrap up this project for now.

John