[GPS_Standard] Self Test experiment with strange result

[Dimitri.p]

Now that I seem to have "Average Frequency 
Offset" seperated from "Average frequency offset 
ERROR"  in my head I'll stop banging it against the wall :)


Thanks !

Dimitri


At 01:06 PM 11/5/2015, Peter Mulhare wrote:
>Hi Dimitri
>
>What Montrol is giving you are Average Frequency 
>Offset, Standard Deviation, and Minimum Offset 
>and Maximum Offset.  These are done in the 
>Current Cycle, the Previous Cycle, and the Accumulated Totals.
>
>If you reset Montrol completely ie quit the 
>program when the GPSDO is running and is stable 
>and locked, then reboot Montrol, and watch the 
>Average offset in the yellow Column for a while.
>
>Immediately Montrol starts there are no averages 
>and the Current Cycle Offset you will see are 
>the actual offset figures, the Accumulated value 
>will also exactly follow for a while but begin 
>to change less as the number of totals build up 
>over time.  Once there are a reasonable amount 
>of completed cycles you will start to see the 
>average offset develop of all the cycles from Montrol switch on.
>
>You will notice that over a much longer time 
>that the figures in the Blue Column (Accumulated 
>Totals) will get closer and closer to the 
>Statistical Offset of the GPSDO.  Because the 
>GPSDO is a Frequency Locked Loop, not a Phase 
>Locked Loop, the actual GPSDO frequency will 
>change slightly over time, this is because of 
>jitter in the 1pps from the GPS, due to various 
>factors, and also the tendency for the crystal 
>to wander slightly in Frequency, all these 
>errors accumulate over time and Montrol is 
>showing you, what the statistical error of the 
>GPSDO is (from a Statistical Bell Curve) at any 
>given time.  It is not giving you at any time the Actual Frequency!
>
>Montrol is showing you 3 slightly different ways 
>to express the Statistical version of all the 
>offsets to date since Montrol was switched on.
>
>This is a simplistic explanation but I think 
>that it will explain it correctly for you.
>
>With regard to your last post that is correct it 
>is not able to read any better than 6.25 x 
>10-9.  But remember that is the max offset above 
>& min offset below, which I would expect to be 
>greater than that.  Mine is typical around 3.12 
>x 10-8.  It is not an average.  And remember 
>that 6.25 x 10-9 is 6.25 parts in 1GHz!
>
>With regard to noise on the EFC line, I took the 
>precaution of using shielded audio cable and 
>only earthed the shield at the electronics end 
>and not at the Oscillator end.  I also shielded 
>the 1pps line from the GPS, and earthed that at 
>the GPS end only.  I’m using a 12Volt CIC 
>ST2145A double oven Oscillator, and an Adafruit 
>GPS on a small breakout board.  The GPS has a 
>rated jitter of ­­+10 nS rms.  My system is 
>working as I would expect it to.  I have 
>replaced the oscillator 3 times, twice with 
>Isotemp OCXO131-100, one of which went faulty, 
>and finally with the double oven CIC 
>ST2145A.  To be honest there has not been much 
>difference statistically between all 3 
>OCXO’s.  I built the unit nearly 4 years ago.
>
>Pete Mulhare
>ZL2iK
>
>
>
>-----Original Message-----
>From: GPS_Standard 
>[mailto:gps_standard-bounces at mailman.qth.net] On Behalf Of Dimitri.p
>Sent: Friday, November 6, 2015 08:48
>To: gps_standard at mailman.qth.net
>Subject: Re: [GPS_Standard] Self Test experiment with strange result
>
>Thanks for the reply.
>Although it justifies the existence of "0x6800" , if we take
>160000001 and divide it by 16  it results 
>in  10000000.0625 Mhz. Now unless I am still 
>missing something any change less that .0625 cannot be "seen" by the FLL.
>So my realization is in a "best" case scenario when the FLL displays
>6800 the actual oscillator frequency is 
>somewhere either exactly  10000000.00000000 
>or  between  plus or minus 0.06249  MHz
>
>And, correct me if I'm wrong,  if indeed  the 
>Min Max display in Montrol is the ppm error the 
>way you described, it will never be able to 
>measure anything better than plus or minus 6.25e-9
>
>
>Dimitri
>