Discontinuity of AWG signal

[Ricardo Ferreira]

Hi everyone,

I was doing some plots of low voltage signals when I noticed the following anomaly here seen at t=0ms:

Here you see scope channel 2 measuring the output of AWG1 directly (nothing else attached). It seems that when the AC signal crosses 0V, there's a discontinuity in the signal which is quite noticeable with respect to the signal around it. This does not depend on the applied offset voltage so it should be an issue with the waveform generator (probably an artifact of the signal generating hardware when rounding negative vs positive numbers). It happens in both triangle waves and sine waves (did not test others).

When passing this waveform through a derivative circuit it becomes quite noticeable. Mostly a heads up, not sure if it will eventually be possible to correct via some sort of firmware upgrade.

EDIT: of course I should have mentioned that I'm using the AD3...

Best regards,

Edited October 13, 2023 by Ricardo Ferreira

[attila]

Hi @Ricardo Ferreira

For better perspective, this is what it looks like when you average 50 captures versus 1 capture.
The zero crossing of the DAC can be a bit noisy due to a minor digital-analog crosstalk.
For low signal amplitudes an external attenuation can be used, to improve vertical resolution and reduce noise.

 

Edit: You are right, the zero crossing looks like an additional DAC LSbit when only one of the AWGs is running, and there is a bit of crosstalk.