|
From: | Ed W |
Subject: | Re: [gpsd-dev] PPS over USB |
Date: | Mon, 07 May 2012 19:15:33 +0100 |
User-agent: | Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:12.0) Gecko/20120428 Thunderbird/12.0.1 |
Granted NMEA time is not traditionally considered incredibly accurate, but *if* the calculation is as I describe above, then we could potentially increase measurement accuracy by 0.5-1ms?There are two issues: accuracy and stability. Your suggestion might improve accuracy. It won't help stability. In the context of bufferbloat, accuracy is not a big deal. We could go through a calibration dance if that helps.
Can you explain what you mean in more detail? I get the difference between accuracy, stability, jitter, but not what you mean here?
As I see it, specifically the Venus chipset "has PPS" already, just over the serial data, not via a separate line. This doesn't seem to be true of other chipsets of course. In fact the Venus appears to give better than PPS because it appears we get multiple observations per second (each extra character in the ZDA sentence). So with the addition of a PPL loop it should be possible to use the Venus to get significantly better than 1ms accuracy
The new device seems to give you the 1 ms that the PPS output appears on, but give or take an offset that is the same as the 1ms timeslot that the $ in the ZDA sentence appears in. What have you gained by PPS?
In fact the $ should be much more accurate, because you can estimate the sub millisec arrival time within that 1ms, whereas with the PPS all you know is that it occured in a given 1ms slot...
This seems like quite an exciting thing to explore. With a chipset giving stable ZDA sentences we should be able to get perhaps 100us accuracy out of a 1ms accurate observation - cool!
I'm going to collect a couple of hours results and plot them. My observations over the last few months (without having thought it special) are that the Venus6+CP210x never visibly deviates more than around 1ms on the NMEA time. I never see any of these few hundred ms drifts that others talk about?That's reasonable. Thanks for the data point. It just means that the firmware on the Venus chip isn't as brain damaged (from a timing perspective) as the firmware on most other chips.
If you check google there are a number of observations on the stability of the NMEA output. I think sparkfun and "tz" on this list have observed the output with a scope and confirmed that you can use the NMEA output as a stable PPS on several of the Venus chipsets
Ed W
[Prev in Thread] | Current Thread | [Next in Thread] |