[time-nuts] Allan deviation

Magnus Danielson magnus at rubidium.dyndns.org
Mon Apr 28 16:41:48 EDT 2014


Hi,

On 04/28/2014 07:56 PM, Артём Щербаков wrote:
> Hello! I built a graph Allan Deviation. Correctly built them? How to know
> the accuracy of time on them?
> http://www.pictureshack.ru/view_71342_image_GPS_20140425.png
> http://www.pictureshack.ru/view_96798_image_GPS_20140426.png
> http://www.pictureshack.ru/view_34327_image_GPS_20140427.png

A few notes:

Your Y scale is in PPM, which might be handy for what you are doing, but 
usually you use 10^-6 type of marking.

What kind of measurement is this? How was the data collected?

It looks like you are white phase modulation limited of quite severe 
level. This makes me wonder how the data was collected and how it looks 
before collected. If it has uncompensated wrappings etc it can look like 
this.

I recommend you to pull down the TimeLab software, where you can acquire 
data from an ASCII file. Make sure that you specify the distance between 
the samples as well as the frequency (and thus period of unwrapping) and 
it will unwrapp it for you. You can then view the data in unwrapped 
phase (p), wrapped phase (w) and frequency (f) to see that your data 
looks sane, then swap over to Allan deviation (a), modified Allan 
deviation (m), Hadamard deviation (h) and Time deviation (t).

Cheers,
Magnus



More information about the time-nuts mailing list