[time-nuts] Strange oscillations

Robert Aurand bob at aurand.com
Sat Sep 4 13:46:00 EDT 2004

Alberto, do you have the spec's for this OCXO?   What are the approximate dimensions of the case?    
And just because you don't see evidence of soldering on the pins doesn't mean the oscillator did not 
get tested for this condition.  It is common practice in industry to screen OCXO's for abberant 
behavior by inserting them into test cards using pin sockets. 
I ask all this because a few parts in 10E-11 is a fairly tough spec for a single oven and if it is 2" 
x 2" x 1" or smaller case size with pins, then the crystal is most certainly seeing some room ambient 
temperature change.   A previous poster suggested air flow as a cause for this effect and it could be 
right.  Some isolation from room ambient temperature changes will help but I don't recommend making 
the enclosure too tight or you will likely get some overheating due to the uncontrolled thermal 
dissipation in the oscillator.   One of my early mentors showed me how oscillators make good people 
detectors, as you have experienced.   We used to have fun watching the frequency fluctuations as 
people entered and exited the lab. 
You won't need to measure current to a part in 10e-11 to see the effect.   Most likely the frequency 
is following the current and the current change might be a few hundred microamps.   The current is 
changing because the thermistor is sensing some room temperature change and, as you have calculated, 
the temp change is very small.   IF this is a single oven the thermal gain might be no more than 
about 1000 (or even much less).   Minute air currents can easily cause this effect. I eventually 
developed my own personal technique for "sneaking up" on oscillators on my workbench!! 
Rick's comments are right and I recommend that you check his papers on OCXO design.   They are 
available on the web and I'm sure he will point you to them. 
So, do I understand correctly that you have this oscillator locked to something when you measure the 
AFC??  If so, what?   In industry the usual technique is to ground the tuning voltage and measure 
frequency against (usually) a Cesium.   This is done in a purpose built "aging rack" with carefully 
controlled temperature. 
----- Original Message ----- 
From: Alberto di Bene <dibene at usa.net> 
Date: Sat, 04 Sep 2004 18:55:00 +0200 
To: Discussion of precise time and frequency measurement <time-nuts at febo.com> 
Subject: Re: [time-nuts] Strange oscillations 
Re:    Richard (Rick) Karlquist (N6RK) wrote: 
Re: I also helped to design the E1938A oscillator, which has digital 
Re: control.  It's oven dynamics are vastly superior to the 10811 
Re: because of relieving the capacitor size constraint, among  
Re: other things.  We had to slam the environmental chambers from 
Re: -40? to +80?C as fast as they could go to measure the oven response, 
Re: which was miniscule. 
Re: [..] 

More information about the time-nuts mailing list