Hi The 70 or 100K resistors are in there because they are a dirt cheap way to isolate the RF on the diode from the DC control signal. If by some strange error, the diode actually *has* leakage near its ratings, they could be an issue. Real world diodes don’t ever seem to have measurable leakage, so it works.
Forward bias (at low current) does nothing bad to the diode. The capacitance goes way up in forward bias. That’s going to swing the frequency more than a bit. The Q of the diode is not so great in forward bias, but not enough to kill most OCXO’s. The net effect may be to tune the oscillator into a region where it was not designed to operate. That could give you poor phase noise or odd spurs. Worst case would be a jump to an unusual mode. The odds of that are pretty small. Most HF OCXO diodes are quite happy below 20V. If you get into VHF or UHF devices, you might find a diode with a rating that is lower. In all cases, it’s like the leakage numbers - yes it’s a spec, does it really matter? Bob > On Dec 28, 2017, at 10:35 AM, Ulf Kylenfall via time-nuts > <[email protected]> wrote: > > Regarding the improved integrator A9 in the 5065A > Rubidium FStandard. (Series 2632) > I have copied some of the improvements (?)that HP made to the 5065. The A7 > AC-Amplifierand the A9 Integrator. Regarding the A9 Integrator: > I noticed that there is nothing that prevents > the [-EFC] signal from going positive in the > improved OPA111 design. > If the XO in the 5065A > is an old 00105-6013, the series resistances > to the varactor in the XO Assembly > amounts to ~70 kOhms which should prevent the > varactor CR1 from forward current destruction. > In the 10811-version I can see 100 kOhms. > How does these oscillators behave should the > -EFC signal swing the full +/- 15V that the > OPA111 is capable of? Does it stop oscillating? > > > Will 15V reverse polarity destroy the varactors? > > I have added two zener diodes to the output > after R3 (21K5) in my design which will > prevent the [-EFC] from exceeding +/- 5V > but the question is if the diodes needs > to be populated or not. > > Comments Invited. > > Ulf Kylenfall > SM6GXV > _______________________________________________ > 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.
