[time-nuts] Logging the grid frequency....

[Tom Van Baak]

> No, just designing a very wicked PLL and needing statistics of the 
> frequency derivative... No generator is involved, it´s mains frequency 
> (but you gave me another idea... thanks!).
> 
> Daniel

Hi Daniel,

I collect accurate mains 60 Hz cycle-to-cycle data here. I can send you raw data for the past few minutes or hour. Each cycle is timed to the microsecond. Contact me off-list for details. From the raw data you can compute the derivatives that you want, or make histograms, etc.

Assuming a mean frequency of 60 Hz and a mean period of 16.667 ms what you will find is that the rms variation in period is about 0.005 ms, cycle-to-cycle, the relative phase or frequency instability is 0.005 / 16.667 = about 3e-4, or 300 ppm.

Since you want statistics to help design a PLL you will be interested in frequency variation over some range of averaging or integration times. An Allan deviation plot shows this quite nicely. Attached is a ADEV,MDEV plot (http://leapsecond.com/pages/mains/mains-60hz-cycle-stability.gif) of the short-term stability (that is, short-term variation in frequency) that you may find useful. It covers averaging times from 1 cycle to 100 seconds.

If you have any questions, let us know.

Thanks,
/tvb
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mains-60hz-cycle-stability.gif
Type: image/gif
Size: 44203 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20130222/bd4b382d/attachment-0001.gif>