Need a power supply design for STM32 based PI detector

[cristiano.sar]

I have found the problem. These breadboards have the power rails divided at the middle and that was the problem. Added some jumpers and now it works.

The Preamp output waveform:

[cristiano.sar]

Now I have another doubt, in the attached 2 images, the best preamp null should be when the DVM measures the most near possible around 0v?

Preamp null at 2.3v



Preamp null at 32mv



My doubt is just because in the first image, the wave form looks better than in the second case, in which the preamp null is near 0v.

Thanks.

[cristiano.sar]

Here is a small vídeo of it detecting for the first time, no my plan is to improve the code and use a better ADC than the 12 bits infernal one

[gdane]

Hey cristiano.sar, interested in your experience using an STM32 instead of the Nano. I'm about to embark on the same adventure once I finish the base Arduino Nano PI build. Any advice or gotchas you can share?

BTW, my plan is to use a STM32H7 with 24-bit ADC in an attempt to avoid having to incorporate an external ADC.

[cristiano.sar]

Hey cristiano.sar, interested in your experience using an STM32 instead of the Nano. I'm about to embark on the same adventure once I finish the base Arduino Nano PI build. Any advice or gotchas you can share?

BTW, my plan is to use a STM32H7 with 24-bit ADC in an attempt to avoid having to incorporate an external ADC.

This is the last version I have of the stm32 detector prototype: https://www.youtube.com/shorts/wLmsmPAFt6U
Im returning to work in this project in the next days. What I usually do is to base the circuit in some already existing good analog designs (like the Eric Foster ones or the ones in the books of Carl Moreland and George Overton) and try to do some things at the digital level, like controlling the TX pulse, delays and sample windows timings. In this specific project I am also doing the detection logic and ground balance at the digital level.
At the MCU level, I am programming in "bare metal" so the things could run as faster as possible. I work with programming, but only with high level languages like Javascript and Python, so pure C and programming at memory register level is not so easy for me, but I am using IA to help in this task.

What I will do next is to use a N-channel mosfet, what was a challenge for me in te past (the way the MCU is powered using the VB+ as GND, very common in N-channel mosfet designs) but with the help of some people from the forum I managed to overcome it.

I will post the progress here.

[GeoMax]

Hi,

I watched your video. The receiver's output signal was visible on the oscilloscope screen. The signal from the test coil isn't as it should be. The test coil needs more attenuation.