phase measurement

If I have two 20 MHz signals, and I want to simply and cheaply (S5 for instance in large quantity) measure the relative phase difference down to a 0.1 degree accuracy between the two signals, how would you guys do it? Some timer circuit, some dsp chip?

i dont know my idea gonna work. I would use a mixer, LPF and ADC. The output if LPF will give you cos(Phi).

A phase-frequency detector as used in synthesizers will have an output proportional to the phase difference. The expensive part will be the ADC to measure the output and the microprocessor or PIC to do the translation from measured voltage to degrees of phase difference.

I thought about the mixer already, but I doubt you could cal out the dc offset over temperature and compensate for AM to PM conversion.

Digital phase detector, interesting...

Take a look at the AD8302. Single chip that gives you amplitude and phase difference between 2 signals. Only good to +/- 1 degree though.

Dave