[GPS_Standard] Self Test experiment with strange result

[Dimitri.p]

In attempting to calculate the average frequency offset , i hit the 
wall described below:

A)
If 6800 is 26624 , which equals  10000000 exactly
and the next value up 6801 translates to 26625 which equals 10000000.0625
(using the formula 10000000 + (X - 26624) / 16  to convert the sample 
frequency to MHz)

that's only a 6.25e-2  step increase.

I dont' understand how montrol is able to calculate a min,max and 
average using what appear to be such "coarse steps".

B)
There is stil more mystery regarding the two sample lines below
D|U|0201E|.|.|6805|000C|0000|0CA8|00
D|U|0201E|.|.|6800|000D|0000|0CA9|00
The sample frequency is 6805 and 6800
but the respective Frequency difference reported is 0000


What part of the magic I am missing ?


Dimitri




At 12:02 PM 11/4/2015, Bert, VE2ZAZ wrote:
>Dimitri,
>
>Use the raw count hexadecimal 68xx. This is the counter reading for 
>each 16 second window.
>
>Cheers,
>
>Bert, VE2ZAZ
>
> > Date: Wed, 4 Nov 2015 11:45:06 -0800
> > To: ve2zaz at sympatico.ca; gps_standard at mailman.qth.net
> > From: dimitri at dotp.com
> > Subject: RE: [GPS_Standard] Self Test experiment with strange result
> >
> > Calculating the Average offset manually sounds like a good idea.
> > I was hoping to do that from the logged serial data,
> > I started logging the serial data, and my eye cought the two lines i
> > pasted below:
> >
> > D|U|0201E|.|.|6805|000C|0000|0CA8|00
> > D|U|0201E|.|.|6800|000D|0000|0CA9|00
> >
> > The sample frequency is 6805 and 6800
> > but the respective Frequency difference reported is 0000
> >
> > Now as far as average offsets go, I'm pretty sure I don't know which
> > values to use to calculate the average offset.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > At 07:28 AM 11/4/2015, Bert, VE2ZAZ wrote:
> > >Hi Dimitri.
> > >
> > >My first thoughts are that this might be caused be the way the
> > >built-in 16-bit counter works. It is plausible that the counter
> > >could mis-behave when everything is absolutely synchronous. This is
> > >a condition that should not occur in real life, and averaging will
> > >do the work to clean this up...
> > >
> > >You should try something else. Turn off the FLL. Do manual computing
> > >of the average offset and see whether it correlates with the
> > >described behavior. This will isolate it to the counter behavior.
> > >
> > >Cheers,
> > >
> > >Bert, VE2ZAZ
> > >
> > >
> > > > Date: Wed, 4 Nov 2015 00:22:17 -0800
> > > > To: gps_standard at mailman.qth.net
> > > > From: dimitri at dotp.com
> > > > Subject: Re: [GPS_Standard] Self Test experiment with strange result
> > > >
> > > > I took the 10 Mhz output from the board, run it through a divider
> > > > chain (74HC390s) to end up with 1Hz and used that to feed the 
> 1PPS input .
> > > >
> > > > Montrol reports
> > > > Avg frequency offset -3.29e-10
> > > > Std Dev 1.46e-08
> > > > Min -3.12e-08
> > > > Max 2.50e-08
> > > >
> > > > These values are in the same neighborhood of what I get using the
> > > > 1PPS from the GPS.
> > > >
> > > > On frequency counters when you measure the counter's own timebase you
> > > > end up with 10 and as many zeroes as can fit on the display.
> > > >
> > > > I was expecting values in the e-12 (or better) neighborhood.
> > > >
> > > > The idea that there is a frequency "offset" when the 1PPS is derived
> > > > from the same 10MHz being measured doesn't make much sense at 
> the moment.
> > > >
> > > > Can anyone shed any light on this ?
> > > >
> > > >
> > > > Dimitri
> > > >
> > > >
> > > > Dimitri
> > > >
> > > >
> > > >
> > > > ______________________________________________________________
> > > > GPS_Standard mailing list
> > > > Home: http://mailman.qth.net/mailman/listinfo/gps_standard
> > > > Help: http://mailman.qth.net/mmfaq.htm
> > > > Post: mailto:GPS_Standard at mailman.qth.net
> > > >
> > > > This list hosted by: http://www.qsl.net
> > > > Please help support this email list: http://www.qsl.net/donate.html
> >