Some math on scope channels behaves strangely

[Anthocyanina]

Was testing the integral and differential math operations, from custom math channels that were shared on the forum before, and I find that they require very large scales, and the derivative is also very noisy, and the bits are too large, and the integral requires an offset, a very large one. Any idea why this behaviour? thank you.

[attila]

Hi @Anthocyanina

Yes. It depends on the noise. What you are seeing is the derivative +/- sample to sample noise it captures. In M1 I didn't multiply it with the rate so you can see a +/-~2mV noise.

You probably want if(isNaN(prev)) prev = C1; and having the initializer code var prev = NaN;  for the first sample to be zero.

[attila]

Hi @Anthocyanina

Use BNC or twisted wires (like 1+/1- and W1/GND) to reduce the noise.

I don't know what is your derivative function but the following M1 for me looks like this:

[Anthocyanina]

Hi @attila thank you. this is what I'm using to get the derivative:

To get the same as you did, I had to acquire the samples at 800Khz, then stop acquisition, and zoom in the timebase to 200µs/div. It looks the same, the only difference being the math scale. But during live acquisition, it is like how i showed in the first picture. Is that the expected behaviour? Thank you.