High purity germanium detectors are the highest resolution gamma detectors available. The only disadvantage is the need for cryogenic cooling, which is traditionally accomplished using a liquid nitrogen Dewar or more recently using a mechanical cooler such as Ortec X Cooler or Ortec ICS. In this article I show how to use a Princeton Gamma-Tech (PGT) Intrinsic Germanium detector with an MCA-PRO 16 and PulseCounter Pro setup to capture extremely sharp gamma spectra.

Dewar Evacuation

For proper operation germanium detectors require cryogenic cooling. For cooling to be effective, the detector system must be thoroughly evacuated. These systems come well evacuated from the factory, but with time vacuum quality degrades. If your detector develops excessive moisture condensation or builds up ice – this is a telltale sign of poor vacuum. There is usually an evacuation port on the detector, which requires a manufacturer-specific valve operator. For the PGT Intrinsic Germanium system I had to build my own valve operator using a rotary actuator and a 1.33″ ConFlat tee, which I mounted on the evacuation port of the detector and connected to a mechanical vacuum pump – Fig. 1.

If you want to be more sophisticated you can add a vacuum gauge and a gauge controller to monitor pressure in the Dewar. Expect to spend several hours pumping as some detectors have sorbent inside, which may take a very long time to evacuate. In my particular case I was pumping for about 8 hours until I reached 1E-5 Torr in the Dewar (instead of the mechanical pump pictured on Fig. 1 I switched to a turbo-vacuum system to attain lower pressure).

Cooling

In most metro areas liquid nitrogen can be purchased from NexAir and it is quite cheap – $300 in my case. NexAir delivers it in a gigantic Dewar, which requires a cryogenic hose for LN₂ transfer – Fig. 2.

The cryogenic hose is typically not supplied and therefore must be purchased separately. Also, it is a good idea to have mittens as the transfer valve will likely freeze over during filling.

Once you fill the detector Dewar wait for ~8 hours for the HPGe system to cool. The label on the system will tell you the required cooling time, which could be as long as 24 hours on some systems. If your detector ices over or your LN₂ boils out of the Dewar completely – you have a vacuum problem and you must evacuate your detector.

Liquid nitrogen in the detector Dewar should last for many days if the system is properly evacuated. Perhaps as long as a week or even longer.

Setting Up A Spectroscopy System

To turn your HPGe detector into a spectroscopy system you need the following:

• high voltage bias supply: many germanium detectors (like NIGC-1519) require negative polarity (the detector should have the recommended bias voltage value written on it), continuously adjustable power supply is best as PGT recommends adjusting the bias at a rate of no more than 100 volts per second, I used Tennelec TC-941;
• shaping amp: to provide +/-12V power to the detector preamp, amplify the detector signal (which is a step) and convert it to a triangular or a Gaussian pulse suitable for an MCA, I used Canberra 2026;
• MCA: I used PicoScope 4262 (which is a part of MCA-PRO16), high-resolution / low-noise MCA is necessary for high-resolution spectroscopy, therefore I chose a 16-bit ADC;
• NIM rack, to host the shaping amp and the bias supply modules, I used Berkley Nucleonics PortaNIM AP-1;
• software: I used PulseCounter Pro to capture and process gamma spectra;
• cables: I used high-quality / low-noise RG-400 BNC and SHV cables to connect the detector to the shaping amp and the MCA.

The complete system is shown on Fig. 3.

Acquiring Spectra

I used PulseCounter Pro to capture and process gamma spectra. I set the sampling rate to 5 MHz, signal range to 5V (to match the max range of the Canberra 2026 shaping amp), pulse processing mode to PEAK, baseline samples to 0, and spectrum channels to 8192 (although 16384 channels might be even better). These preset settings produced the best results.

Because maximum expected resolution of a high-purity germanium detector is 1.8 keV @ 662 keV (which corresponds to the astonishing FWHM of just 0.3%), extreme care must be taken in tuning the system to reduce noise and improve signal to noise ratio. Here are some tips:

• High-quality / low-noise linear bias supply is a must;
• high-quality cabling is a must (e.g. RG400);
• Canberra / Mirion loop-buster accessories such as ground loop isolator LB1504, HV bias isolation box LB1503 and ferrite choke for the preamp signal cable may be necessary to eliminate ground loops if any;
• high-quality / low-noise 16-bit ADC is a must;
• 8000 to 16000 spectrum channels is a must;
• pole zero (PZ) cancellation must be well adjusted on the shaping preamp to eliminate baseline drift;
• longer shaping time (e.g. 6 to 12 us) might be necessary.

The captured Cs-137 and Ba-133 spectra are shown on Fig. 4 and 5.

The captured spectra are nothing short of astonishing: I was able to achieve 0.4% FWHM at 662 keV without the loop-buster accessories, which is pretty close to the highest theoretically possible value of 0.3%.

P.S. It is possible to feed the preamp output directly into the MCA-PRO16 bypassing the shaping amp. In this case the PulseCounter preset should be configured to use 50 mV or 20 mV range with the pulse processing mode set to STEP. It will be possible to capture decent spectra with such a setup although the resolution won’t be as high as with the shaping amp: 50 mV preamp signal is quite weak and the reduced SNR will broaden peaks on a gamma spectrum.

Conclusion

MCA-PRO16 and PulseCounter Pro software work very well with high-purity germanium detectors. The complete high resolution HPGe spectroscopy system is available on eBay.