Gamma spectrometry is a tough discipline

Instead of updating the gamma mega blogpost with all the latest work, we’ll recap the last few points made and bring you up to date here. The problem was that it was getting a bit big. Due to the potentially profound importance of finding repeatable spectra connected to this phenomena, much time has been invested in reverse-engineering the gamma detector probe that Jean-Paul donated.

The probe is a 4 inch diameter, 4 inch in length, thallium doped sodium iodide crystal, this is coupled to a photo-cathode and then a photomultiplier. With high quality advice coming from Steven Sesselmann of Gamma Spectacular, we re-engineered the photomultiplier head connector in order to make it work with his excellent hardware.

The first job was to deduce the circuit diagram of the previous connector which used a more classical sampling system. After some iterations in the interpretation and analysis of the existing system, came the following schematic:

Generally speaking, a photomultiplier connector is voltage divider, it uses using identical resistance from the anode up to the last stage. Because a photocathode requires more tension to work properly, the tension is usually greater and requires larger resistance.

What is interesting is that this very photomultiplier seems to use a different resistance not only for the photocathode but also a little bit before in order to allow adjustment of the gain/tension applied to the first stages.

After a quick expert analysis, Steven came-up with the following schematic.

Because Steven felt that the connector was part of the order that Earth-Tech paid and placed for us, he kindly donated the resistors that are used in this assembly. This leads us to the last update (#6) where Bob explained our need for $800 to make the lead well that would provide the appropriate shielding required to discern any gamma radiation signal coming out of the experiment above background.

This requires to dive into learning the software and manage to get better prices for the lead flashing and the copper tube required. Using a local lead flashing supplier, he has been offered the lead needed for 466€ which is a good price for this material. For the copper, a provider in Paris is able to supply the copper tube for 225€. The total for the whole well is 691€. Subtracting the €146 that Jacques Ruer sent, we only need 545€ ($750) for that part of the project.

The software is another fascinating part of the whole system. Firstly, the receiver converts the electronic signal from the the photomultiplier electron flux into sound that is then captured at high frequency by the computers sound card. From the modulations and tone of the sound interpreted and following a spot of CPU processing, the software provides a spectrum of the gamma detected!

There are few programmes serving this niche market. PRA and Thermino MCA are the most used. PRA is provided by the School of Physics in the University of Sydney Australia and Thermino MCA is programmed by another amazing team of passionate people in Italy. Fortunately both applications are cost free.

These two are very much complementary, since Thermino MCA is full of amazing features like comparison database and a simple frequency tuner, PRA has less eye-candy but has essential features like flux vs. time recording and a better pulse filtering and calibration scheme than T-MCA (thanks again to Steven for pointing that out!).

The first task was to get the whole measurement pipeline working. In doing so, we realised there were a small hole where light was able to get into the photomultiplier and mess-up the data. He re-engineered this part more carefully, just to make sure it won’t happen again.

Apparently, compared to classical gamma-metry, this is “piece of cake”, although Mathieu says that for a rookie like him it is still a steep learning curve that he is now getting to grips with. Take a look at the following two spectra, the first one is blank and the second is made with tungsten rods containing 2% of thorium that I use with Jean-Paul in the high pressure Mizuno experiment.

You should be able to clearly see two peaks emerging following the two blue vertical lines shown at low energy coming from the thorium. From the tables we guessed when the measurement started with the thorium that a 2.6MeV peak form Tl-208 would emerge, but it also shows up in the blank test. Please discuss this in the blog comments, if you are knowledgeable in this field you might understand what is going on.

Mathieu had advice from Larry Forsley who is clearly an expert in gamma spectroscopy. There may be problems resulting from the way the NaI crystal had previously been treated. It was mentionned that even a small variation of temperature tends to be break the crystal. Because they are so fragile, it should have been good procedure to carry it insulated from external temperature variation during the adventure it went through before ending up in the lab. For this reason Mathieu suggested to Jean-Paul that having a backup probe might be a necessary, if only for comparison. I just hope any second probe we may see won’t be completely different in shape because of the well design ... The good news is he said there may be one available and that he thinks it might be our detectors twin brother which should simplify building another re-fit!

The final advice received from Larry was to put an intermediate layer of tin between the lead and the copper, because tin is heavy enough (high-Z) to catch the X-Ray emitted by lead it would greatly improve the quality of the shielding making it almost completely opaque to any kind of rays.

He also offered a control source of 0.25 µCurie of Cs-137 for calibrations. This will be very helpful to calibrate the detector, it will enable quantitative and qualitative analysis of the complete system! We still have to figure out how to get it here because of legal constraints in France for this type of material, but his offer is highly valuable. We’ll keep you updated on this evolving story.