FIDLER stands for Field Instrument for the Detection of Low Energy Radiation. Typical FIDLER detector contains a 5″ diameter x 0.06″ thick NaI(Tl) “pancake” scintillator mounted on a 5″ photomultiplier tube protected by a thin (0.01″) beryllium window – Fig. 1.

The job of a FIDLER is to register low-energy gammas / x-rays. Typical energy range is 10 to 200 keV, average efficiency is about 10%.

In my devices I use FIDLER detectors made by Alpha Spectra.

Due to large area of the crystal FIDLER offers exceptional sensitivity with background counts on the order of few hundred CPS, depending on environment. Also, because the crystal is extremely narrow, FIDLER possesses some directionality: most x-rays that are not more or less orthogonal to the it’s surface will likely miss the detector.

The upper bound of the useful energy range is determined by the scintillator thickness, and the detection efficiency drops exponentially with gamma ray energy. But what about the low energy limit?

I have captured background counts using my 5″ FIDLER-PRO system by setting trigger to -7 mV, just 1 mV above the electronic noise. The spectrum (Fig. 2) contains a curious low-energy peak, which is made up of genuine photomultiplier pulses (Fig. 3). I specifically chose AREA mode when I acquired this spectrum so I could integrate the pulse area to accurately determine energy even for the lowest-magnitude pulses.

Now, is this pulse a genuine background signal or a strange artifact? The easy way to tell is by cover the detector window with a lead sheet, which will block virtually all low-energy background x-rays. So, I have recorded another background spectrum with a lead brick shielding the detector’s window – Fig. 4.

Mean counts with lead screen dropped from 376 to 223 but the low-energy peak has lost less than 20% of its height. So, the conclusion is that this low-energy peak is caused by device’s intrinsic noise. This low-energy peak corresponds PMT’s thermal noise, which can be mitigated by cooling the photomultiplier.

To eliminate this peak one must increase the absolute magnitude of the trigger level and sacrifice low energy resolution as a result. So I did just that and captured a ⁵⁵Fe spectrum, Fig. 5. I can still see massive 5.9 keV peak, although I lost some of the peak height due to the increased trigger level in order to cut out the PMT noise.

So, I have captured some more spectra: ¹³³Ba (Fig. 6), ⁵⁷Co (Fig. 7), and ¹³⁷Cs (Fig. 8). Calibration linearity is very good, and ⁵⁷Co 122 keV line FWHM is about 14%.