Your main concern will be injection locking, followed by the additive noise of the buffer amps that you use to prevent it. The CMOS output buffers in the oscillators themselves may provide enough isolation, but don't count on it. Injection locking is easy to spot with a scope. You can adjust the tuning voltage of one oscillator to match the frequency of the other, and see if their phases snap into coincidence.
If necessary, LMH6702 opamps can be used at the frequencies covered by the CVHD-950 parts. They will provide plenty of isolation with little or no effect on noise measurements in the -165 dBc/Hz vicinity, given strong signals to work with (+10 dBm or more). If I were driving a mixer with 3.3V HCMOS, I'd try a series 50-ohm termination first. Basically, pretend that you have a perfect voltage source whose output impedance will be that of whatever resistor you place in series with it. It won't be, of course, but the effect will be close enough. The connections between the oscillators and mixers will not be long enough to make exact impedance matching necessary. There should be no need to filter the signal; mixers are switches, and they like square waves. If you want, you can use an L-network to handle both lowpass filtering and matching, of course. The only filter necessary is the LPF that follows the mixer. -- john, KE5FX Miles Design LLC > -----Original Message----- > From: [email protected] [mailto:time-nuts- > [email protected]] On Behalf Of Anders Time > Sent: Monday, March 10, 2014 8:33 AM > To: [email protected] > Subject: [time-nuts] How to measure phase noise of HCMOS oscillators? > > A while a go i bought some low phase noise Crystek CVHD-950 and now I > want > to meas the phase noise of the oscillators. > Is it best to try to impedance match the output signal and filter and > amplify the hcmos before i feed it in to the phase detector? or Just hope > that the signal is high enough to drive the mixer? What about all the > harmonics? Does it alter the phase noise result? > /Anders > _______________________________________________ > time-nuts mailing list -- [email protected] > To unsubscribe, go to https://www.febo.com/cgi- > bin/mailman/listinfo/time-nuts > and follow the instructions there. _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
