[GPS_Standard] My first attempt

[John Knott]

I have just finished my implementation of the standard design.
Built on .1" PTH matrix board with ground plane in stick on copper  
strip. OCXO is Micro Crystal (ebay) 0-5V tune +- 40 Hz.
GPS adafruit ultimate gps breakout board.
Each element uproc and opamp,OXCO , Buffer (HC240), 232driver and 1pps  
divider uproc have separate 7805 regulators
pre regulated with LM317 set to 8 volt.
To get a 1pps output from 10 Mhz  synchronized to pps in a 12F675  
version pD15 is included see Tom Van Baak leapsecond.com/pic.

First task was configure jumpers on 12F675 to make pps from external  
10MHz (Racal nanosecond counter 10MHz out or Rubidium  standard)  
synchronized to
pps in. Display on HP54510a Digital 250MHz 1G sample scope, pps in v  
pps from ext 10MHz. With infinite persistence pk-pk jitter is less  
than 200nSec. Working on philosophy that external 10MHz may
not be accurate but should not jitter this is the GPS pps sample -  
sample jitter. Display drifted 400ns in 10 mins ie .66ns / sec so ext  
10MHz was pretty accurate.
Buffered pps from gps (12" BNC Cable) had severe 1 cycle ring on  
rising edge, gps out is 2.8 v cmos, no impedance quoted, 750R  
terminator to 0V with 750R feedback around
HC240 buffer cures this.

Next task 2nd 12F675 to make pps from OCXO 10MHz, can then compare  
jitter out to in and out to external derived pps Racal nanosecond  
counter 10MHz out or Rubidium  standard.

System seems to settle with default parameters except samples to 2 to  
count of  target+- 3 or 4. OCXO frequency measured with jitter of one  
count on Racal nanosecond counter (1)000 000 06/7  E6 leading 1 has  
overflowed, difference 60/70 mili Hz jitter 10.

A most interesting project thank you for  putting in public domain.

John Knott 2E0KNJ


PS could  not compile gps_stdmain.asm with Mplab8.9 objected to CCP2MX  
= B3,  changed to CCP2MX = OFF compiled and my hex code same as  
supplied.