Posted on by maxfomitchev

Detector Signal Processing

Most x-ray and gamma detectors produce a current pulse that has to be converted to a voltage pulse that can be measured and analyzed in order to build a spectrum.

Classic signal processing approach involves charge integration using a charge-sensitive preamp followed by pulse-shaping using a shaping amplifier, followed by analysis via a multi-channel analyzer.

Charge integration and pulse shaping steps are common steps in detector signal processing. These are the steps that one would need to take when building a spectrometer around an SDD, SiPIN, SiPM, or CZT detector.

Cremat Charge-Sensitive Preamp

Fortunately, there is no need to reinvent the wheel. Cremat offers a range of charge-sensitive preamplifiers that you can quickly evaluate using a CSP evaluation board. In fact, Cremat CSP board offers housing and +24V DC power supply that gives you a fully modular CSP to experiment with – Fig. 1.

Fig. 1. Cremat charge-sensitive preamp evaluation kit.

All and all, the CSP evaluation board is a good starting point for developing your own design. However, for most intends and purposes the CSP evaluation kit is already good enough and compares favorably to $1000-dollar third-party preamps. And it does not break the bank – you can buy all the components for under $300.

Cremat Shaping Amplifier

To shape the amplified detector signal into gaussian pulses suitable for multi-channel analysis one can use Cremat shaping amplifier chips, which support a variety of shaping time constants – Fig. 2.

Fig. 2. Creamt CR-200 shaping amplifier chip, 4 us shaping time.

Once again, to make your life easy you can go with Cremat shaping amplifier evaluation board, which can be purchased separately or as a part of a kit that gives you a fully-functional shaping amplifier complete with a power supply – Fig. 3.

Fig. 3. Cremat shaping amplifier evaluation kit.

The entire kit complete with a baseline restorer is also under $300. All parts are available for purchase online at Cremat’s Amazon store with the next day delivery via Amazon Prime! Wow.

Performance Evaluation

I have evaluated Cremat CSP and shaping amplifier performance using a Kromek EV-180 CZT detector. The objective was to see if I can reach the EV-180 resolution spec using Cremat’s very inexpensive hardware without any modifications or optimizations. The other objective was to compare Cremat CSP performance to Oxford / Tennelec TC 170 preamp.

To start, I have mounted the EV-180 directly on Cremat CSP – Fig. 4.

Fig. 4. Cremat CSP with Kromek EV-180 directly mounted.

I have used 57Co, 133Ba and 137Cs sources to evaluate the detector performance.

To build a complete CZT spectrometer system I have connected the Cremat CSP to the Ortec 456 power supply to provide 900V of bias to the Kromek EV-180 CZT detector. Then I connected the output of the Cremat CSP to the input of the Cremat shaping amplifier. Finally, I connected the output of the shaping amplifier to the channel A of the PicoScope 4262, which captured the shaped output for analysis using the PulseCounter software.

The entire system (except for the Ortec 456) is shown on Fig. 5.

Fig. 5. The complete CZT spectrometer system (except for the Ortec 456 power supply).

Results

According to Kromek, the resolution of my EV-180 CZT detector is 3.7% FWHM @ 122 keV and 3.4% FWHM @ 662 keV.

57Co spectrum captured using the Cremat CSP compared against the same spectrum captured using Tennelec TC 170 is shown on Fig. 6.

Fig. 6. 57Co spectrum captured using the Cremat CSP (blue) and Tennelec TC 170 (red). In both cases Cremat shaping amplifier was used.

With the Cremat CSP the resolution is 4.4% FWHM @ 122 keV, which is somewhat worse than the manufacturer’s specification of 3.7%. Using the Tennelec TC 170 I was able to get 3.9% FWHM @ 122 keV, which is inline with the manufacturer’s spec.

What does this result mean? It means that the Cremat CSP evaluation board has higher than optimal capacitance necessary to get the best resolution out of the CZT detector. Matching the preamp capacitance to the detector capacitance is a must in order to achieve full resolution. Still, 4.4% vs 3.7% right off the bat is not bad.

137Cs spectrum is shown on Fig. 7.

Fig. 7. 137Cs spectrum (logarithmic plot) captured using the Cremat CSP and Cremat shaping amplifier.

The resolution was 3.3% FWHM @ 662 keV, which is inline with the manufacturer’s spec of 3.4%.

Conclusion

Cremat offers excellent products for professionals and for enthusiasts alike. These products are inexpensive ($55 for a CSP chip) and ship next day via Amazon Prime. Cremat evaluation kits are awesome and enable quick experimentation with various detector types. Cremat’s performance is on par with much more expensive (e.g. $1,000+) third party amplifiers.