Hi These days you can get (for a price) NPO caps up above .1 uf. That makes super high value coils less important in a low frequency beat note system. If you are trying to measure phase noise at 100 KHz then 0.1 uf to ground at multiple places likely is not a real good idea.
Bob On Mar 2, 2010, at 11:28 PM, Brian Kirby wrote: > I got to correspond with a person at NIST today about DMTD - about the HP > 10514B Mixers. They pointed me to a application note at > http://www.thegleam.com/ke5fx/stellex/5952-8217.pdf (hey - thats one of our > time-nuts!) that shows driving the 10514 mixer at +13 dbm for the LO port and > 0 dbm for the RF port. > > This application note is a good read. HP recommends terminating the mixer > into a 3900 pF cap and 50 ohms in series for the RF termination. ANd they > use a LC filter for their VNA set for 90 Khz. I also looked at the > schematic of the 8407A were I got the mixers from. They terminate the IF > into a 1100 pf cap in series with a 51 ohm resistor, and then it passes thru > a 2.7 uH choke going into an amplifier - the IF I beleive is 298 Khz for this > unit. > > I also did a reconfiguration of my temporary DMTD experiment that has also > improved the noise floor. Drawing attached. Also changed the drive levels. > > For reference the previous configuration showed 2x10-11 for 0.1 sec; 2x10-12 > for 1 sec, 5x10-13 for 10 sec, 6x10-14 for 100 sec, 7x10-15 for 1000 sec. for > the noise floor. > > The current configuration (of the drawing attached) shows 8x10-12 for 0.1 > sec; 8x10-13 for 1 sec, 8x10-14 for 10 sec, 8x10-15 for 100 sec and 8x10-16 > for 1000 sec. I used a little more attention in sheilding and spacing to > eliminate crosstalk. Note the noise shows some linearity this time....plus > more drive.... > > Brian - KD4FM > > > > > <DMTD_Plans_1.1.pdf>_______________________________________________ > 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.
