hello,

i have two general questions concerning the quadrature demodulators.

why do all of the quadrature demodulators you offer have a divide-by-2 or in general divide-by-N local oscillator (LO) injection? what is the purpose/advantage of this circuit? why can't i input my LO-signal directly and it just gets shifted by a 90°-phase-shifter?

what does the phase accuracy mean exactly? if i wanted to simulate the demodulator, how would i do that. lets say my LO-signal is the following:

s_LO= sin(2*pi*f*t)

after the divide-by-2 and the phaseshift i get two mixing signals. lets say the phase accuracy is delta_phi=0.5°. would my mixing signals be:

s_inphase = sin(2*pi* (f/2) *t +- delta_phi) = sin(2*pi* (f/2) *t +- 0.5°)

s_quadphase = cos(2*pi* (f/2) *t +- delta_phi) = cos(2*pi* (f/2) *t +- 0.5°)

or

s_inphase = sin(2*pi* (f/2) *t +- (delta_phi/2) ) = sin(2*pi* (f/2) *t +- 0.25°)

s_quadphase = cos(2*pi* (f/2) *t +- (delta_phi/2) ) = cos(2*pi* (f/2) *t +- 0.25°)

?

and how is the the delta_phi distributed? can i assume it is equally distributed between -0,5°<delta_phi<+0,5° ? are there any statistics?

thank you

jejo86

Yes, the phase accuracy at a particular frequency will be constant. The spec tables provide typical and/or worst case phase error data at spot frequencies and the plot sections of the datasheets show the variation vs frequency.

We do not have an IQ Demodulator that works in the 50-150 MHz range which is 1XLO based. The best suggestion I can give is to use the ADL5802 dual mixer combined with an external phase splitter.