STM32F103 noise at ADC pin on MCU side

[NBula.1]

Hi,

I am trying to interface a signal with one of the ADC pins of the STM32F103 MCU. When the signal was scoped using an oscilloscope, it was seen that a noise component was added to the original signal from the MCU side. I have attached images of the following,

1. Original signal (without connecting to the ADC pin)
2. Original signal when it is connected to the ADC pin
3. ADC pin without any signals being connected

Please see that the noise component is periodic and it is present even without any signal connected. 

Please help me debug this, for an example, whether it is a power issue or routing issue?

Nisal

 

[TDK]

Looks like your signal is very high impedance. There will be some drop during ADC sampling as charges up the capacitance inside the ADC. This is normal, expected, and unavoidable for high impedance sources. Extending the sampling time until it reaches equilibrium is one way of mitigating this.

You can use an op-amp in follower mode to transform a high impedance signal to a low-impedance signal.

[AScha.3]

..or read ds . there input impedance on adc pin given - didnt read , eh ?

--otherwise you get , what you see: charging the adc input by your hi-impedance source !

[NBula.1]

Hi TDK,

Thanks for the reply. I already tried with an opamp stage(LM2904). But the result was the same. I will try with another opamp. 

Nisal

[waclawek.jan]

>  I already tried with an opamp stage(LM2904).

Any schematics?

> But the result was the same.

What does it mean, exactly?

JW

[NBula.1]

Hi waclawek.jan,

Please find the attached schematic 

The same result meaning, the pulsating noise can still be seen on the ADC pin.

Thank you.

[Peter BENSCH]

You should try a RRIO opamp such as the TS507 rather than the standard bipolar LM2904 whose output swing only reaches VDD-1.5V.