[time-nuts] time-nuts Digest, Vol 121, Issue 65

[Bob Camp]

Hi

The typical small Rb’s are temperature compensated. They have a real tempco of a bit less than a  ppb. It gets corrected to about 10X better than that using data from an internal temp sensor. Correction is often three point, so it may or may not track the actual performance of the unit at all temperatures. 

There are a couple of gotcha’s with this approach. The first is that the sensor needs to track what’s going on with the unit. If you do things that change the thermals (heat flow) of the unit, that may no longer be true. The next issue is the step size of the correction. It’s digital, if you vary back and forth barely over a step boundary, it will quite happily modulate your Rb. The net result will be a unit with worst ADEV than one with the correction disabled. 

This is very much a “your MPG may vary” sort of thing. If you happen to have a golden unit that is very flat before correction, the correction will not impact you much. If you have one with a third order curve to it’s pre-correction characteristic, the three point / two line segment correction isn’t gong be as effective as it might be. 

Also o the list of things to be aware of:

Rb’s tune with a magnetic field. Changing the local field can change the output frequency.

Rb’s have a sensitivity to barometric pressure. Eliminating this is a bit hard. Correcting for it may be the better approach.

Again both of these effects vary unit to unit. Your part may not be as sensitive as my part. 

Lots of fun.

Bob

On Aug 23, 2014, at 9:41 PM, Hal Murray <hmurray at megapathdsl.net> wrote:

> 
> kb8tq at n1k.org said:
>> If you have a temperature stable environment (or create one) you can get
>> some very good results with an (good) Rb locked to a (good) GPS via a proper
>> long time constant setup. It’s not easy, but it can be done. 
> 
> What's the temperature sensitivity of the typical telco-surplus Rb unit?
> 
> 
> -- 
> These are my opinions.  I hate spam.
> 
> 
> 
> _______________________________________________
> 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.