I'm designing a device for power measurement in a 13.56 MHz plasma source. I'm using an ADL5391 to multiply two signals proportional to voltage and current resp.
I used the Z input for adjusting the offset to zero with no input signals present.
Now if I apply a 13.56MHz signal to one of the inputs (other input is still zero) then a small DC offset appears at the output which seems to be almost proportional to the square of the (single) input signal.
My assumption is that there is a small amount of crosstalk from the X to the Y input and vice versa inside the chip that causes this square term to arise at the output.
The crosstalk is not external since I tried shorting the unused inputy directly at the pins. This makes no difference.
My question: Is my assumption correct and is this offset related to crosstalk? Would it be possible to correct this offset by applying a small amount of external crosstalk opposite in phase as a counter measure?