In the datasheet of AD8339 the demodulator clock phase of I/Q channel is not clearly stated, please see below figures, if the 4LO clock input is cos(2pi*4fd*t), I would like to confirm which of the following figures is correct, thanks !
Yes, this is correct. A 10 kHz RF signal with a 10 MHz complex demodulation frequency would mean that the 10 kHz RF would be shifted to a negative frequency ... specifically -10MHz +/- 10 kHz. Since sin() is an odd function, sin(2*pi*f*t) with a negative frequency is -sin(2*pi*f*t). If you try swapping RF and LO frequencies (ie LO < RF), you will find that the Q output leads by 90 degrees.
I'll check into this for you.
The second one is correct. The 0 deg LO reference corresponds with cos() and 90 deg with sin().
We did a simple measurement, sent a very low frequency signal as input (10kHz, the lowest supported input frequency of AD8339, treat it as a DC signal), and watch the output of I/Q channels, it seems on the 90 degree path of AD8339 the mixer mixed input signal with "-sin(wt)" instead of "+sin(wt)". Could you confirm if this is correct or not ? And if so, is there any reason to put a negative sign on the 90 degree path ? This is not intuitive to normal users. Thanks.
Retrieving data ...