[time-nuts] using Fury GPSDO with Rb

Scott Mace smace at intt.net
Fri Apr 11 11:42:35 EDT 2008


Turns out this unit has been set so that the EFC was only useful from within a .5v range
centered around 2.5v, also the output was TTL.  Trying with a different Rb...

	Scott

SAIDJACK at aol.com wrote:
> Hi Scott,
>  
> a good setting is when the unit recovers in minimal time and with  only minor 
> overshoot from a manual coarsedac change.
>  
> Try dacgain 1000 - 2000.
>  
> TI of 1.2ns / SD 2.2ns is not bad. The M12 errors can be expected to be  
> larger than this over 8 hours.
>  
> bye,
> Said
>  
>  
> In a message dated 4/9/2008 07:06:45 Pacific Daylight Time, smace at intt.net  
> writes:
> 
> I've  been running with dacgain 4000, efcscale 1.2, efcdamping 25, 
> phaseco  15.  It has been locked for about 8 hours (since I started the  
> test).  average TI 1.2ns sd 2.2ns.
> 
> I tried bumping the  coarsedac by 2 increments to see what happens.
> It tries to recover, but  will oscillate wildly after about 45min or so.
> I wonder what will happen  after an extended holdover period.
> 
> It looks like the EFC on this this  FRS-A is more than twice as sensitive 
> as the LPRO.
> 
> Scott
> 
> 
> 
> SAIDJACK at aol.com wrote:
>> Hello  Scott,
>>  
>> forgot to mention: also try setting the  serv:efcdamping lower. It's an IIR 
>  
>> low pass filter, and will  introduce phase-shift that can cause instability 
> if 
>> it  is set  too high.
>>  
>> bye,
>> Said
>>   
>>  
>> In a message dated 4/8/2008 22:59:22 Pacific Daylight  Time,  
> SAIDJACK at aol.com 
>> writes:
>>
>> Hello   Scott,
>>
>> this means the gain is set too high so that there is  not enough  phase  
>> margin.
>>
>> You can try to  set the serv:dacgain lower, and  serv:efcs / serv:phaseco  
>>  lower.
>>
>> Contrary to OCXO's, Rb's are  quite stable so  they don't need strong  
>> control, 
>> they just need a   gentle nudge, so you can try starting with low  numbers.
>>
>>  Let me  know if that works,
>> bye,
>> Said
>>
>>  
>>
>>
>> In a message dated  4/8/2008 21:31:50 Pacific  Daylight Time, 
> smace at intt.net   
>> writes:
>>
>>  I've  found that once the units gets into an  overshoot/oscillation  cycle
>> it never  recovers and I have to manually  adjust the  dac to break the 
> cycle.
>> Is the  notion that you use   conservative enough settings so that this
>> never    happens?
>>
>> Scott
>>
>> SAIDJACK at aol.com  wrote:
>>> Hi   Scott,
>>>  
>>> let me  reply on this list since other users  are  also using  Rb's.
>>>  
>>> Here is a quick   pointer:
>>>   
>>> 1) determine EFC slope of Rb  (pos =  frequency goes higher with  higher  
>> EFC  
>>> voltage). Set the  slope with serv:efc:slop   [positive/negative]
>>>  
>>> 2)  Determine maximum  frequency  deviation at EFC = 0V (serv:coarsedac  
> 0)   
>>> using a frequency  counter
>>>   
>>>   
>>> 3) Determine maximum frequency   deviation at EFC = 5V  (serv:coarsedac   
> 255)
>>>   
>>> 4) Set  Dacgain  according to table in user manual depending on the   
>>  frequency  
>>> difference between 2) and  3)
>>>   
>>>  5) Set Integral part of loop to  zero: serv:phaseco 0.0, this  helps for  
>  
>>  now
>>>  
>>> 6) Set Proportional part  of loop to a  number that  makes the unit lock 
> as  
>>> quickly  as  desired without oscillation  and overshoot (for example:    
>> serv:efcs 1.0 
>>> to serv:efcs  6.0
>>>   
>>> 7)  Set the integral part of the loop so that the   phase is driven to 0ns 
>  
>> UTC  
>>> offset as quickly  as desired without   overshoot/oscillation, for 
> example:   
>>> serv:phaseco   20.
>>>  
>>>  Repeat 6) and 7) after some  hours.
>>>   
>>>  Hope that helps,
>>> bye,
>>>  Said
>>>   
>>>>  -----Original Message-----
>>>>  From:  Scott Mace
>>>>  Sent:  Saturday, April 05, 2008   17:51
>>>> To: Said  Jackson
>>>> Subject:  Re:   Fury   GPSDO
>>>>
>>>>
>>>> Said,  I finally  got  my fury up  and  running, I'm using a efratom  
>>>> FRS-A and  also a  LPRO-101.
>>>> Do  you  have any recommending starting  points for  tuning the  servo 
>>>> for a  rubidium    unit?
>>>>
>>>>  Scott
>>>>   
>>>  
>>>  
>>>
>>>
>>>  **************Planning your summer  road trip? Check  out AOL Travel  
>> Guides. 
>>
>>>    
>>  
> (http://travel.aol.com/travel-guide/united-states?ncid=aoltrv00030000000016)
>>>    _______________________________________________
>>> 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.
>>
>>
>>
>>  
>>
>> **************Planning  your summer road trip? Check  out AOL Travel 
> Guides.   
>>  
>>  
> (http://travel.aol.com/travel-guide/united-states?ncid=aoltrv00030000000016)
>>  _______________________________________________
>> 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.
>>
>>
>>
>>
>>  
>> **************Planning your summer road trip? Check out AOL Travel  Guides. 
>    
>>    
> (http://travel.aol.com/travel-guide/united-states?ncid=aoltrv00030000000016)
>>  _______________________________________________
>> 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.
> 
> 
> 
> 
> 
> **************Planning your summer road trip? Check out AOL Travel Guides.    
>   (http://travel.aol.com/travel-guide/united-states?ncid=aoltrv00030000000016)
> _______________________________________________
> 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