[time-nuts] poor-man's oven
kb8tq at n1k.org
Sun Jun 4 09:36:34 EDT 2017
The gotcha is that you have multiple systems working against each other. The crystal in the TCXO has
one temperature characteristic. The compensation in the TCXO has a temperature characteristic. They
cancel each other out to a limited degree. The residual slope may (or may not) be as shallow as you
might think. Your PTC is at an arbitrary point on the residual curve. A somewhat more subtle issue is
the gradient between your PTC, the crystal, and the compensation as it cycles.
If the TCXO really isn’t a full TCXO, then some of this goes away. A +/- 2 ppm 0-50C “TCXO" may not have
any compensation in it at all. Some 0 to 70C parts are done as 2 ppm 0 to 50 and only compensated at the
hot end. They actually may be worse with the PTC than at room...
Yes this all assumes an AT cut in the TCXO. That’s a pretty good bet ….
> On Jun 4, 2017, at 8:13 AM, jimlux <jimlux at earthlink.net> wrote:
> I recall some years ago folks were talking about putting a PTC thermistor on the TCXO of a FlexRadio SDR1000 to stabilize the frequency as a sort of poor-man's OCXO.
> It's also referenced at
> where he says "order of magnitude improvement" with no numbers (from 1% to 0.1% or from 1 ppb to 0.1 ppb?)
> I wonder how well that actually works.
> Say you bought an inexpensive (perhaps non TC) XO and an equally inexpensive thermistor, glued on on the other, hooked em both up to 3.3 or 5V.
> Yeah, there's issues with room air blowing on it, and tolerances in both the XO and thermistor, so your absolute frequency accuracy may not be so hot. But what sort of medium to long term performance can one expect.
> I did some searches, because I'm sure we've discussed this before, but I couldn't find it. There was some stuff from Oct 2007, but that was in the context of a more complex circuit, and the thermistor was the sensor. (discussions of TE devices too)
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts