NUCLEO-F446RE: Dual SiPM Signal Acquisition and Coincidence Detection

[RAJA90]

Hello everyone,

I am a research scholar in High Energy Physics and currently working on a scintillation-based muon detector using SiPMs.

My setup:
- Board: STM32 NUCLEO-F446RE
- Sensors: 2 SiPMs
- Signal conditioning: Op-amp amplification + peak detection
- Output: Analog pulses to STM32 ADC

Objective:
I want to acquire signals from two SiPM channels using ADC and implement coincidence detection (i.e., detect events when both signals occur within a very short time window).

My questions:
1. What is the best approach for precise coincidence detection on STM32?
- ADC polling vs interrupt vs DMA vs timer-based sampling?

2. Since SiPM pulses are very fast (ns scale), how can I reliably detect coincidence using ADC (which is slower)?

3. Would it be better to:
- Use analog comparators + GPIO interrupts?
- Or external coincidence logic (hardware)?
- Or timer input capture method?

4. What timing resolution can realistically be achieved using STM32F446RE?

I am looking for a robust and research-grade solution suitable for muon detection experiments.

Any guidance, example code, or recommended architecture would be highly appreciated.

Thank you!    @Max VIZZINI @mƎALLEm  @GMA @

Peter BENSCH

 

 

 

[TDK]

> detect events when both signals occur within a very short time window

> Since SiPM pulses are very fast (ns scale), how can I reliably detect coincidence using ADC (which is slower)?

You should better define what you are trying to capture Rather than "ns scale", 10 ns? 20? 1? And what voltage, and what offset between the two signals are you trying to capture.

The STM32 ADC won't help you directly here as it is not "ns scale".

[Andrew Neil]

---

@RAJA90 wrote:

using SiPMs.

---

A what?

 

---

@RAJA90 wrote:

I am looking for a robust and research-grade solution suitable for muon detection experiments.

---

As @TDK said, you need to better define your requirements - what does, "research grade" even mean?

 

"very fast (ns scale)" doesn't sound hopeful on a low-cast microcontroller - you're probably into the realms of custom hardware - FPGA ?