At last, I'm running my first Rb disciplining algorithm. Using the time-to-digital posted time ago, I get the difference from the M12M PPS and the Rb 10MHz divided down PPS. The difference (that is phase) is read every second, is sawtooth corrected, is accumulated for 100 seconds. The difference of this difference (that is frequency) is also accumulated but for 1000 seconds. At the end of 1000 seconds the accumulated value is applied to the DAC as an integral action (DACIntegralnew= DACIntegralprev+Freq1000*Ki) and the accumulated phase is applied as a proportional action (DAC= DACIntegralnew+Phase100*Kp). Now the resulting Rb PPS is in a 25nS window compared with the Z3815A (using an oscilloscope with the infinite persistence, last 3 hours after 22 hours running). _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
