[time-nuts] Centering ocxo

Bill Dailey docdailey at gmail.com
Wed Oct 3 13:00:19 EDT 2012


Thanks Said.  This is a learning experience but it is fun watching it settle down.  I will stop fiddling with the settings and just let it do it's thing.  I am watching the output with my qs1r direct digital to spectrumlab.  This way I don't have to worry about any audio glitches. I have been recording frequency every minute the entire time to see what it is doing.

Sent from mobile

On Oct 3, 2012, at 10:39 AM, Said Jackson <saidjack at aol.com> wrote:

> Doc,
> 
> You are on the right track, efc scale affects SD as you can see.
> 
> The phaseco parameter is used to push down the average TI to 0ns. Higher values push faster.
> 
> If your ocxo is still drifting (aging and or retrace) it will take about 48 hours for the aging measurement and correction to kick in, and bring the offset down to 0ns.
> 
> Bye
> Said
> 
> Sent From iPhone
> 
> On Oct 3, 2012, at 3:46, Bill Dailey <docdailey at gmail.com> wrote:
> 
>> ok..  So that may very well be true of this unit.  Electrical tuning is
>> 3E-7  0 - 5v (+/-).  It also lists a digital tuning range of +- 3Hz at
>> 10MHz.  Correct me if I am wrong but that appears to mean 3Hz electrical
>> and 6Hz digital tuning range.   I am not doing digital tuning but thought I
>> would throw that out there.
>> 
>> I have been trying to "optimize" parameters on this Fury board but it seems
>> my "optimization" has just been increasing the deviation.  Running 1.8 ns
>> sd overnight with an average TI of about 26ns (was with my "optimized"
>> settings)...the original settings were giving me a much lower deviation...I
>> didnt log it but looking at the graph of frequency in excel I would say
>> probably between 0.1-0.6 ns.  I just put it back on the original settings
>> and am letting it settle now.  Was adjusting Dampening, EFCSCALE and DAC
>> gain.  My observations reveal dampening makes it a bit slower to respond
>> and perhaps settles it some, The efcscale seems to act as pure gain on top
>> of the baseline dac gain which is essentially determined by the tuning
>> range as you referred to.  What I saw with a low efcscale is that the TI
>> was higher but the SD lower... with efc scale higher the TI was lower but
>> the SD suffered.
>> 
>> Since my goal here is low noise and very good short term stability I prefer
>> the lower efcscale  (low gain with low SD).
>> 
>> Let me know if I have any gross conceptual errors here or if I am looking
>> at this properly.
>> 
>> Doc
>> KX0O
>> 
>> 
>> On Tue, Oct 2, 2012 at 11:54 PM, Hal Murray <hmurray at megapathdsl.net> wrote:
>> 
>>> 
>>> docdailey at gmail.com said:
>>>> I am ok for awhile but how do you center the efc of an ocxo?  I
>>> understand
>>>> there is something (screw) to adjust the ocxo so it is approximately on
>>> freq
>>>> with 2.5v efc.
>>> 
>>>> Specific oscillator datum-1111c. I have he datasheet but doesn't say
>>> "coarse
>>>> frequency adjust this screw" or some such.
>>> 
>>> There may not be a coarse adjustment.  If the tuning range is big enough to
>>> cover the aging over your design life, you don't need one.
>>> 
>>> 
>>> There is a tradeoff between adjustment range and the number of bits you
>>> need
>>> in a DAC to get a required accuracy.
>>> 
>>> Suppose I have an adjustment range of 1 Hz (peak to peak) on a 10 MHz
>>> oscillator.  That's 1 part is 10^7.  If I have a 10 bit DAC, I can adjust
>>> to
>>> 1 part is 10^10.  A 20 bit DAC can get to 1 part is 10^13.
>>> 
>>> But if the tuning range is 10 Hz, the same 20 bit DAC setup only gets you
>>> to
>>> 1 part is 10^12.
>>> 
>>> 
>>> 
>>> --
>>> 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.
>>> 
>> 
>> 
>> 
>> -- 
>> Doc
>> 
>> Bill Dailey
>> KXØO
>> _______________________________________________
>> 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.
> 
> _______________________________________________
> 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 mailing list