[time-nuts] 40MHz source
grant at ghengineering.co.uk
Mon Mar 25 13:34:54 EDT 2013
I think you're going down the right road with the Chris Bartram transceiver.
My suggestion for a 40MHz source would be to take a 10MHz source and
feed it into two successive doublers, with a bit of inter-stage and
post-stage buffering and filtering.
Frequency doublers can be very simple indeed - as simple as a
full-wave rectifier. It's the sort of thing that somebody in the
local radio club might be interested in building, even if they are
afraid of the 'scary' microwave stuff. Lot's of possibilities for
As for a 10MHz source - I'd be tempted to go for either a good quality
OCXO, or maybe a GPSDO such as the Trimble Thunderbolt (which has a
fairly good 10MHz OCXO inside it). Many Rubidium sources are designed
for medium-long term stability, at the expense of close-in phase
noise, and that might be a problem in this application, as the 10MHz
oscillators in many rubidium sources don't have as low phase noise as
the 10MHz oscillators in the good GPSDOs.
A good quality 10MHz frequency standard will also have other uses in
the lab, of course.
To answer Jim's question - 'inside the box' is an LTC6946-3 integrated
Frac-N PLL/VCO with an o/p at 5GHz, then feeding into a sub-harmonic
mixer to get to 10GHz. The PLL uses a 40MHz reference, which I
believe is also the PLL comparison frequency. I think that the
supplier might be reluctant to change the software to use a lower
comparison frequency, but I haven't asked.
The loop bandwidth of the PLL is somewhere in the 10kHz-20kHz range,
so any close-in phase noise from the reference will get multiplied up
to the final frequency, hence the need for a reference with good phase
Subject: [time-nuts] Are there any rubidiums
programmahttps://mail.google.com/mail/?shva=1#inboxble to 40 MHz?
I'm possibly looking for a 40 MHz source and I know some of the
rubidiums are programmable. But can any of the affordable ones be
programmed to work at 40.0 MHz?
I was looking for a source to drive this 144 MHz -> 10 GHz transceiver.
The TCXO oscillator is off the board and a separate item, but costs
?40 and then one ideally wants to lock that to a more precise source.
The oscillator will lock to an external 10 MHz source, but then one
needs to buy both a 10 MHz rubidium as well as this 40 MHz TCXO. Hence
I was wondering if there was a cheaper more compact solution, which
just used a rubidium, and dispensed with a TCXO.
More information about the time-nuts