Re: Timing via serial port

[Kai Harrekilde-Petersen]

I've measured the delay between the 1PPS signal and the TX uart output on the NEO-M9N board with a logic analyzer. I would need to find the or re-measure the data for you tonight.

Front edge of the UART frame is the most precise. Rear edge depends too much on the length of the message.

As I recall, the latency from the PPS to the front edge increases(!) with increasing update rate. I did not check for the variation between the 1PPS and the start of the UART message.

Regards,

Kai

---

From: "O'Connor, Daniel"
Sent: Wed Sep 30 14:08:21 GMT+02:00 2020
To: Hal Murray
Cc: gpsd-dev@nongnu.org
Subject: Re: Timing via serial port

 On 30 Sep 2020, at 18:28, Hal Murray <hmurray@megapathdsl.net> wrote:
 Has anybody tested various GPS devices to see how good the timing via the 
 serial port is?
 
 Some of the NMEA sentences have sub-second time stamps.  Are any of them 
 useful?
 
 Has anybody looked at whether the front of the sentence (or burst) has better 
 timing than the end?

I would say it depends how tight your requirements are.
I setup USB serial to a GPS engine and fed NTP with it and it seemed quite good - certainly NTP was happy to use it and all the numbers look good.

I imagine it would also depend a lot on your hardware and OS as to what sort of jitter you would get.

--
Daniel O'Connor
"The nice thing about standards is that there
are so many of them to choose from."
 -- Andrew Tanenbaum