Lester Veenstra wrote:
From: [email protected] [mailto:[email protected]] On Behalf Of David
Sent: Tuesday, September 21, 2010 10:55 PM
To: [email protected]
Subject: [qs1r] Looking for good, cheap, external reference
Hi,
I'm interested in doing some ionospheric research, but I don't want to break
the bank with an $8,000 USD GPS disciplined Rubidium clock.
Anyone have some ideas on cheaper alternatives? The QS1R, and the Flex-3000,
both use a TCXO and they have pretty good long-term stability, but there is
a 1-2 Hz wander at 10 MHz imposed on the frequency over a duration of about
45 minutes to one hour. It looks like an exponential rise to one limit,
followed by an exponential decay to the opposite limit.
That sounds an awful lot like an airconditioner/heater/airflow changing
during the day sort of variation.
For long term stability, it obviously requires a temperature stabilized
environment, but I have observed a slow diurnal variation in the TCXO mean
frequency of about 1-2 Hz at 10 MHz just sitting on my desk, exposed in the
morning to some brief sunlight.
For comparison, my old Icom 756ProIII looks astonishingly stable, but with a
fixed frequency offset of around 0.8 Hz (the best I could do by tweaking the
pot on the PLL). At most I see a very slow diurnal wander of around 0.5 Hz
at 10 MHz. Whatever Icom did, they did it very well.
Measured against what?
---
but to practicalities..
Do you actually need to generate a signal stable to fractions of a Hz,
or is being able to calibrate a recorded/digitized signal good enough?
If the latter, you could just record a 1 pps along with your data, and
calibrate it out.
For what it's worth, a GPS disciplined ovenized quarts oscillator is
"cobble together able" for a lot less than $8k (and since you're
considering QS1Rs and Flexradios, you're clearly in the "willing to
cobble" category).
A SRS Rb is uner $2k, by the way, brand new. http://www.thinksrs.com/
PRS10 is list $1495
A FS725 is about $2500
_______________________________________________
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.
