Hi The FE Rb's are dirty enough that you may need a second stage of PLL filtering in the system. Possibly a cheap 10 MHz low power OCXO and a sub 1Hz loop. A lot depends on just how good you want the close in noise and spurs to be.
Put another way - is this PSK-32 at 10 GHz or conventional AM or …? Do we want to tune a bit to find the guy or not? The flip side to the fancy stuff would be SSB at 10 GHz. If I'm within a couple of KHz, I can hear you / you can hear me. We can tune to match things up. That drops the accuracy on the mobile end to 100's of ppb. A <$30 eBay OCXO will handle 1/10th of that nicely. A cell phone TCXO might also do the trick over a narrow range. I'd much rather have the OCXO's stability once I tune on frequency. Bob On Mar 10, 2013, at 12:36 PM, [email protected] wrote: > Having built a few portable frequency sources my recommendation based on > your application and what is reasonably available today would be a FE 5680A > in combination with a 100 MHz clean up XO. We have very good results with a > slightly modified what I call the Dorsten PLL-VCXO by W-H-Rech. I can send > you the article and some plots, the article is in German but you would be > able to understand or I would help you. > Using an AGM battery and a $ 4 up converter I would plug it in to the car > when driving and at home run it off AC and when in the field you will have > plenty of time to operate. 9.8 A 12V Li-ion are available for $ 45. > When home keep it running and use the RS 232 to update Rb frequency. > Bert > > > In a message dated 3/10/2013 10:23:21 A.M. Eastern Daylight Time, > [email protected] writes: > > Asking here on behalf of a friend.. > > With respect to portable amateur microwave operation.. you want good > close in phase noise (so you can use narrow band filters) AND good > frequency accuracy (so you can find the signal)> > > the typical operation is "drive somewhere, operate a bit, drive > somewhere operate a bit" repeated (contacts from different grid > squares/peaks/what haveyou" > > My instinct is that this is an application for a nice quiet OCXO on a > battery. Adjust the frequency before you set out against a good > reference and just go from there. > > Surplus Rb references are apparently also popular, but I think they keep > those on battery too (that is, you need to be ready to go 10 minutes > after arriving, and I don't know that a Rb is "settled in" that quickly). > > So the question from my friend was with reference to GPS disciplined > oscillators. Would that do any better? I'm used to GPSDOs in a fixed > location where you have time to do long term averaging. > > And what about truly mobile operation (there are folks in the SF bay > area apparently doing 10GHz mobile ops.. slotted WG radiator on the roof > of the car, etc.) > > What sort of 1pps timing accuracy do you get from a GPS "on the move". > I assume it would have the usual 10ns sort of uncertainty (in that the > mfr specs don't say "only with the antenna fixed in one place for N > hours"). 10ns is only 1E-8 of a second. Presumably one can average a > bit over many pps ticks. > > > I've got a bunch of Wenzel Streamline units, and they typically do > 1E-10/day aging and 1E-9 over temp. Assuming the temperature doesn't > vary a "lot", seems like the OCXO is "better" than the GPS, at least in > a 1-2 day time frame. (and, of course, isn't that just what a GPSDO is, > in holdover mode, anyway) > > The Rb is good to 1E-11 over the short run (assuming it's been > "calibrated" recently) but I notice that the PRS10 data sheet says 7 > minutes to 1E-9, so in the "non continuously powered" mode of operation, > it's not all that wonderful. > > > The Rb is definitely higher powered.. The PRS10 is 2+ amps at 28V to > start, and 0.6 to run. 15-16 Watts is a lot to keep on a battery. > (Assume you run off a pair of 7Ah 12V batteries.. that gives you 10-12 > hours). > > The Wenzel is a couple watts (after a 5W warmup). The GPS is a LOT > lower power. The Garmin GPS 18x is 0.45W, of course the 1pps on that > receiver is only specified to 1 microsecond.. A moto Oncore UT is a bit > less than a watt and claims <100ns (with SA.. showing the age of the > datasheet I have). > > > _______________________________________________ > 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. _______________________________________________ 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.
