To better study the neutron flux originating from my bubble fusion reactor I decided to employ an additional LND 25169 neutron detector to supplement my 6x LND 251106 neutron detector bank. The LND 25169 detector has two unique features compared to LND 251006: it is smaller (3.5″ long and 0.5″ in diameter) and has much higher 3He pressure, 20-bar vs. 8-bar in 251106.
Strictly speaking, for thermal neutron detection high 3He pressure is not necessary. Detector sensitivity to thermal neutrons is almost entirely determined by the detector geometry: a larger detector would have higher sensitivity compared to a smaller detector and the difference in sensitivity is proportional to the ratio of the detector volumes. For example, thermal neutron sensitivity of LND 25169 and 251106 is 5.1 cps/nv amd 18 cps/nv respectively, which follows precisely the ratio of the detector effective volumes of 6.5 and 23.4 cm3.
Pressure has little effect on thermal neutron sensitivity, but sensitivity to non-thermal neutrons is directly proportional to 3He pressure.
Ortec 142PC Preamp and Ortec 570 Shaping Amplifier
To capture a neutron spectrum I decided to go ‘old school’ using an Ortec 142PC proportional counter preamp – Fig. 1 coupled to an Ortec 570 shaping amplifier – Fig. 2.
When selecting a preamp, the most important criterion is to match the detector capacitance to the preamp. E.g. Ortec 142PC, Tennelec TC170 and Canberra 2006 are designed for proportional counters with capacitance under 100 pF. Using a preamp designed for higher-capacitance detectors with a low-capacitance counter will result in a noisier / distorted spectrum.
Ortec 570 shaping amplifier is pretty much equivalent to Canberra 2026. Although the Canberra 2026 is newer and has more features (like a potentiometer for pole zero adjustment, triangular vs. Gaussian shaped pulse output), the performance of the two is very similar.
By why do we need a preamp to begin with? Because a proportional counter produces a very weak signal, on the order of a few mV in amplitude. This signal gets even weaker if we use any sort of cabling between the detector and the amplifier (due to RF attenuation). In my NEUTRON-LITE and NEUTRON-PRO systems I do not use a preamp and I capture the detector signal directly. I can get away with this because the cable from the detector to the data acquisition system is very short and the intrinsic noise of the electronics is very low. Also, I maximize the detector bias to get the highest amplitude signal.
Thermal Neutron Spectrum
The job of a shaping amplifier is to convert the pre-amplified detector pulse (which looks like a step) to a Gaussian or triangular pulse proportional to the pulse’s energy – Fig. 3.
We can use the shaping time as a trade-off between the energy resolution and the count rate. For highest counting rate it is best to use the timing output of the preamp.
Setting the shaping time to 3 us and bias to 1000V I acquired the following thermal neutron spectrum using the HDPE moderated Po-Be check source – Fig. 4.
The shaped output of the Ortec 570 was captured using the PicoScope 4262 and the spectrum was built using the PulseCounter software.
Compare this to the spectrum captured using the Tennelec TC171 preamp, which is not ideal for the LND 25169 detector because it is designed for detectors having the capacitance larger than 100 pF – Fig. 5.
The spectrum acquired using the Tennelec TC171 preamp is significantly distorted compared to the spectrum acquired using the Ortec 142PC preamp due to the detector capacitance mismatch.
Neutron Spectrum Without HDPE Moderator
Maximum detection efficiency requires using a moderator to slow down neutrons to thermal energy in order to maximize the detection probability. However, in the process we loose information about neutron energy and obtain a thermal neutron spectrum regardless of the neutron source. Because the LND 25169 detector has such a high 3He pressure (20 bar), it is interesting to see what the Po-Be neutron energy spectrum would look like if I do not use any HDPE moderator and place the detector right next to the source. The resulting spectrum is shown on Fig. 6.
Because non-thermal neutron detection probability is ~10,000 lower than thermal neutron detection probability the resulting count rate was just 0.02 cps, which is 20 times lower than when using the HDPE moderator.
The resulting neutron spectrum (Fig. 6) still has a thermal neutron peak. This happens because even without the moderator some neutrons thermalize due to their interaction with the environment (e.g. moisture in the air, hydrogen in the wood of the desk, etc.). These thermal neutrons are readily detected while vast majority of the non-thermal neutrons will pass through the detector without being registered.
However, thanks to the long 10 hour acquisition time, some non-thermal neutrons did register forming a tail that spans all the way to 4 MeV energy range. Compared to the thermal neutron spectrum captured using the HDPE moderator (Fig. 4), this is a significant and interesting difference as such tail is just about absent on the spectrum captured using the moderated detector.
Therefore, the LND 25169 detector can be used to detect non-thermal neutrons if the acquisition time is long enough or the neutron source is sufficiently intense. Still one would expect to see a dominant thermal neutron peak unless special care is taken to prevent thermalization, e.g. the source and the detector are located in a vacuum chamber away from hydrogen rich materials, and the environmental neutrons are screened using a Borax castle.
Conclusion
‘Old school’ NIM BIN setups are still useful and the components are still being manufactured. Proportional counter setup comprising an Ortec 142PC preamp coupled to an Ortec 570 shaping amplifier is a ‘golden standard’ for 3He-based neutron detector characterization. A similar setup can be built using the Mirion components such as a Canberra 2006 preamp coupled to a Canberra 2026 shaping amplifier.