How I've turned a Picoscope USB oscilloscope into an MCA

[madexp]

A few months ago I purchased a PicoScope 2204A USB oscilloscope purely for my teaching duties. I would plug it into my laptop, project live waveforms onto the classroom screen, and use it to illustrate analog signals in real time. One day I came across Dr. Max Fomitchev-Zamilov's MCA kits on eBay, which turn a PicoScope into a multi-channel analyzer via a .NET interface and the PicoSDK. I thought it would be a fun experiment to reproduce that functionality in Python.




In practice it proved fairly straightforward, but I discovered two important hardware limitations:

• 8-bit ADC – with only 256 quantization levels, the spectral resolution is coarse.
• Quirky hardware trigger – setting a threshold that works at one voltage range (e.g. 40 mV in ±500 mV) may fail at another (±1 V), because the internal DAC or comparator resolution isn't fine enough to generate the trigger voltage accurately. In such cases raising the threshold (e.g. to 80 mV) restores reliable triggering.

Below I sketch the core acquisition algorithm for pulse spectroscopy, and then describe the Python GUI implementation.

Acquisition Algorithm

• Initialize
  Record the start time and zero the event counter.
• Read User Settings
  • Full-scale input range (volts)
  • Lower-level discriminator (LLD) threshold (volts)
  • Polarity ("Positive" or "Negative" pulses)
  • Time-per-division (s/div)
  • Number of samples per block
• Configure Scope
  Set the selected channel A range and coupling (AC for MCA), then apply the hardware trigger at the LLD threshold with the chosen edge.
• Acquire Waveform
  Call run_block(), which:
  
  • Queries the timebase to get the sampling interval
  • Arms the scope and waits for the block to complete
  • Reads back a buffer of raw ADC counts
  • Converts counts to millivolts and then to volts
  • Returns time (t) and voltage (v) arrays
• Detect Pulse Peak
  If polarity is positive, take peak = max(v); if negative, take peak = –min(v).
• Compute Histogram Bin
  Normalize: raw_idx = (peak / full_scale_range) * 255
  Clamp: idx = min(max(int(raw_idx), 0), 255)
• Update Statistics
  Increment histogram[idx], increment the total event count, then compute elapsed time since start and the event rate (events/sec).
• Emit Updates
  • Send the raw waveform to the GUI for a "last-pulse" preview
  • Send the bin index to increment the bar or line histogram
  • Send the updated count and rate to the status display
• Loop
  Continue steps 2–8 until the user stops the acquisition.

Software Structure

• SimplePicoScope2000
  Wraps all PicoSDK calls: opening/closing the unit, setting channels and triggers, querying timebases, and retrieving blocks of data.
• AcqThread
  A Qt thread that continuously acquires oscilloscope waveforms and emits them to the main GUI.
• McaThread
  A Qt thread that runs the pulse-spectroscopy loop described above, building a 256-bin histogram and streaming updates back to the GUI.
• MainWindow
  A PyQt5 application window with two tabs:
  
  • Oscilloscope: real-time trace display, noise measurement, and start/stop controls
  • MCA: spectrum acquisition controls, last-pulse preview, and dynamic histogram with rate and elapsed-time readouts

[madexp]

The source code of the script cannot be posted here.
I've published it on GitHub: https://github.com/l-papadopol/PyMCA

[DL8BCN]

That looks interesting!
 

[madexp]

Added baseline restoring plus pulse shape acquisition/recognition. Now it works quite well.
The pictures shows 50keV + 200keV + 300keV peaks from LYSO.
The peak on the righjt is 511keV peak wich is out of range ( I had to use different voltage range or reduce gain)
Sie dürfen in diesem Board keine Dateianhänge sehen.

[Zugpferd]

Crazy!