Noise on ADC A1 input with audio board connected (Teensy 3.2)

Status
Not open for further replies.

FredT

Member
Hi, looking for wisdom on this issue. When I use the audio board with teensy 3.2 analog input A1 has twice the amount of noise. I though it was somehow the noise from the clocks generated by the teensy 3.2. However, just loading a simple code to read A1 without any Audio board configuration, In other words, the audio board is powered on but not configured still give the same amount of noise on A1.

This is happening if the teensy is stacked with the audio board too. (Pin 15 to audio board is cut)

I'm also reading A0,A2,A3,A6,A7 and they are not suffering from this problem.

Thanks
 
A1 is wired to the volume pot pads of the audio board, but guessing you have not actually populated that in anyway have you, so in theory there is just a cap and a connection to Agound on the pin beyond what the other pins have?
 
When stacking them for troubleshooting, I cut the pin A1 from reaching the audio board entirely.

I also saw that even without the audio board connected, A1 was slightly more noisy than analog input but not as bad when the audio board is there. So at least the trend is the same but audio board increase the noise by factor of 2 on A1 compared to the other analog input.

I did more troubleshooting and found something interesting. In my application I'm reading A0,A1,A2,A3,A6,A7 one after another. I found that Changing the order of read has an impact. I guessing that it has something to do with the ADC multiplexer.

The best result was to put the read in order of their internal name in the datasheet. When I do this, noise is pretty even across all ADC input. In this config the noise floor is about 200/4096. For internal 3.3 internal ref and 12bits sampling.

Also I do sampling without averaging.
 
If you're reading analog pins without signals connected, you can expect to get a lot of random noise.

Even when signals are connected, you only get good quality measurements when the signal source impedance is low, like under 5K to 10K. High impedance signals get lot of noise from capacitive coupling of lots of nearby digital stuff.
 
If you're reading analog pins without signals connected, you can expect to get a lot of random noise.

Even when signals are connected, you only get good quality measurements when the signal source impedance is low, like under 5K to 10K. High impedance signals get lot of noise from capacitive coupling of lots of nearby digital stuff.

Indeed, I knew I was more than borderline with the impedance. It was working ok until I added the audio board. I added an opamp buffer (LM324) before the ADC input and things looks very good now.

Thanks for your input.
 
Status
Not open for further replies.
Back
Top