[time-nuts] Holdover, RTC for Pi as NTP GPS source

Trent Piepho tpiepho at gmail.com
Wed Nov 1 22:44:35 EDT 2017

Some of the DS chips have a square wave input that can be fed from a GPS
PPS and the RTC will discipline itself.  It can also produce a PPS when the
GPS stops.

The question is, is NTP's holdover on the system clock better than NTP
disciplining the system clock with an RTC PPS.  I'd guess not.

Forget about accurate time from the RTC's registers.  The slowness and
variably in I2C is quite poor and most of them don't even have sub-second
resolution when read or written.

On Nov 1, 2017 4:22 PM, "MLewis" <mlewis000 at rogers.com> wrote:

Is this a workable or worthwhile strategy?:
- RTC providing date & time to second to system on boot
- RTC frequency output driving a counter/divider to produce PPS
- GPS module providing UTC PPS
- GPS module's secondary PPS disciplining the RTC-counter-divider PPS by
resetting the RTC's counter/divider (I'm assuming there's one that will
rest fast enough to sync; I've never looked into these...)

If GPS PPS is lost, then the RTC counter/divider is producing a recently
disciplined PPS.

Valid or invalid?

And the DS3231 has:
- a 32K output, and
- an Active-Low-Interrupt/SQW output that can be set to PPS.
It's unclear to me how to sync the DS3231 PPS to the GPS PPS, or if that
can be done.



time-nuts mailing list -- time-nuts at febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/m
and follow the instructions there.

More information about the time-nuts mailing list