Generating a spectrum from the PIN diode X-ray detector.
(Electronics) (Ionizing radiation)The pulse shaping amplifier outputs short pulses with around 10 mV/keV:
.2 ms/50 mV per division
These pulses have a similar amplitude to audio from a microphone preamp, so they can be digitized by feeding them into a computer’s soundcard. If the computer has dedicated audio input, connect the circuit as the headset microphone using a TRRS plug. Generally, the sleeve is ground, and the ring next to it is the microphone signal. The soundcard will apply a bias voltage to power the microphone, so add a 1 uF MLCC DC blocking capacitor.
Make sure to disable any automatic gain/level control (AGC), and record uncompressed audio when capturing the signal.
Once captured, a simple script can measure the pulse heights, and plot a histogram of how many a given energy was seen:
The bremsstrahlung peak from 6-14 keV is clearly present, but not the characteristic X-rays from zinc at 8.6 and 9.5 keV.
The peak around 4 keV is just noise being interpreted as photons.
Nothing is shown before 3 keV becuase of a cut of in the processing program, to avoid millions of random counts.
With a setup that at least sort of works, it will be possible to optimize the circuit to get noise down to a usable level in future posts.
Raw data from spectrometer: Wave file, Raw mono signed 16 bit little endian 44100 Hz data.
Python script used for plotting: Source.