2026-05-12 2:29 AM - last edited on 2026-05-12 4:06 AM by mƎALLEm
Hi, we have purchased stm32h750vbt6 microcontroller for analog signal validation with 10k potentiometers to integrate with game controller joystick especially joy.cpl using windows 10 OS.
We have dealt with DMA Circular mode configuration pointing cortex-M7 DRAM to 0x30000000 to retrieve the analog signals associated with respective 10k potentiometers. Totally we have four analog pins such as PA0,PC0,PA3 and PC5 from ADC1 & ADC2 with proper distinguished rank allocation alongside desired maximum sampling cycles but although we have altered and tested with different values of this above configuration it ended up with ADC channel leakage which is declared 16 bit ADC.
Moreover we have tried adding external 100nf capacitors to avoid these linkage but failed.
We kindly request you to resolve the ADC channel leakage in our current stm32 Project. Our entire stm32 project attached for your reference and fault identification.
2026-05-12 4:06 AM
Hello,
I recommend to refer to the AN5354 "Getting started with 16-bit ADC on STM32H7 MCUs" as it explains how the ADC performance varies under various conditions and provides guidelines to exploit the full potential of the STM32 16-bit ADC.
2026-05-19 11:56 PM
I have already studied through the document you mentioned. We are exactly using 10k potentiometer and sometimes alternative IR sensor with photodiode for analog signal interfacing. While we are using all the axes of joy.cpl joystick game controller of window OS channel leakage exists. Our problem had not been solved and this issue is affecting our entire process with ADC GPIOS of STM32H750VBT6. I need your support specifically for interfacing joy.cpl with ADC GPIOS of STM32H750VBT6. Using DMA circular continuous mode configuration in STM32cudeide.
2026-05-20 1:15 AM
Ok you read it.. but did you apply the guidelines described in that application note to your hardware or not?
2026-05-20 8:55 AM
If "leakage" is means interference in between two or more channels, when changing one pot affects readings on another, than capacitors not really helpful here. What necessary is low output impedance buffer in between 10k pot and adc input.
Another things to try is oversampling, when adc does bunch of readings on each channel before switching on the next. Setting, for example, 2-3 megasamples per seconds rate, you can do 256 readings for less than 100 usec, so responsiveness is still very good, but interference lowered by 256 times!
We’re moving the ST Community to a new platform to give you a better and more reliable community experience.