STM32F4 ADC Excessive noise level, Noise level is over 5LSB Опции

[pitt]

Noise level is over 5 LSB. Why? My options?

Using 480 cycles, INJ registers, ADC3, changing prescaler... Don't see improvement.

Let me clarify what I mean 5 LSB: it isn't +/- 5 clocks but 2^5.

Сообщение отредактировал pitt - Sep 23 2015, 15:11

[scifi]

На схему и на разводку смотреть надо. У меня такого безобразия (32 отсчёта) даже близко не было.

[pitt]

На схему и на разводку смотреть надо. У меня такого безобразия (32 отсчёта) даже близко не было.

Thanks for reply.
Unfortunately it is out of my control... Schematics looks clean, nothing really suspicious. Traces, routing and other EE stuff is over my level of expertise.

[EmbedElektrik]

Thanks for reply.
Unfortunately it is out of my control... Schematics looks clean, nothing really suspicious. Traces, routing and other EE stuff is over my level of expertise.

Could you share your code? It seems that there are some mistakes in initialzation of GPIO, RCC or ADC. Check it again.

[pitt]

Could you share your code? It seems that there are some mistakes in initialzation of GPIO, RCC or ADC. Check it again.

All I can share is values of the ADC registers just let me know which do you need. They are fine with me.
Problems with RCC and or GPIO is nonsense: one wouldn't be able to get any readings, but I do - my problem is noise. I suspect that something with the board(made in China) but don't have EE expertise to prove it right or wrong.
The power and Vref are fine: checked with my scope.
I'm talking about raw(unfiltered) readings.

Regardless, the other chat over here is clearly off topic. I'm native Russian speaker but don't and never will have blind typing skills. Replies in Russian are not a problem at all except may be terminology.

Thank you all for response.