Hi all,

I am trying to determine the error in measurement of the phase using the AD5933. While the calibration method by Analog Devices proposes to eliminate this by using a resistor and then "calibrating" the system. How this works is not exactly explained. I understand that the unknown phase = phase calculated - system phase. And indeed this system phase is what we get when we calibrate the device. But this just seems to be interpolation.

I've looked at a paper that I've attached here which proposed to eliminate this error mathematically. However, I'm wondering where the formulae in the paper come from? I have looked into the datasheet, the user guide for the evaluation board and also the AN-1252 document. None of them seem to have it. Is there some more documentation supporting the AD5933 that perhaps is missing from the AD website?

I see in the documents that DDS starting frequency can be programmed but that is about it. Can someone please explain to me where the above formulae come from? They are all supposedly taken from AD documents.

The unknown phase = phase calculated - system phase is not an interpolation - it is a fundamental property of linear circuits.

Section 3.2 of the article seems to be the authors' puzzling over the phase shift introduced by the low-pass filter they show on Fig.1 - the "box" between the ADC and PGA2. The 90 degree offset is something the AD documentation is silent about.

Hopefully somebody from AD can tell whether the formulae is something the authors came up with on their own or is taken from some AD documentation.