hello,

you can use stable "reference voltage source" - for example TL431 - see atached schematics.


Of course the OA power voltage not has to be 12V and OA can be rail-to-rail for lover voltage. On the OA out you should have 3.3 V.
Best Regards

 

I agree that the 'scope photos do not show random noise, instead most of the noise is clipped at the same amplitude instead of having random amplitudes.

The LM358 is never used for audio because it produces so much noise that it is not listed on its datasheet. It also produces crossover distortion and has a poor audio high frequency slew rate. Like all amplifiers, its power supply must have a filter caoacitor to ground.

 

Part of what you're seeing is due to the cheap digital scope you are using. On the 20mV scale it looks like the LSB is 0 or 2mV even if the signal is less. If you are looking to measure circuit noise levels below a few mV then you can't do it with the equipment you have.

The 3 inch scope ground lead can add many mV of noise and radio signal. Use the short (1 cm) spring ground attachments that came with your probe.

A bad circuit layout can also add noise.

What noise levels and frequency are you looking for?

 

One more thing. Resistors are inherently noisy as a function of temperature.
Resistor Noise: Deafening and Hard to Reduce | Analog Devices

 

One more thing. Resistors are inherently noisy as a function of temperature.
Resistor Noise: Deafening and Hard to Reduce | Analog Devices

1.8μV in a 20kHz bandwidth for a 10k resistor at room temperature. I don’t think we’d see that on the TS’s scope.

 

1.8μV in a 20kHz bandwidth for a 10k resistor at room temperature. I don’t think we’d see that on the TS’s scope.

Possibly not, but it is there all the same and the lower you go the more significant it becomes.
Toss in an amplifier with 30dB of power gain and it becomes significant.