On 2012-02-09, Ron Frazier (NTP) <[email protected]> wrote: > On 2/9/2012 4:18 AM, David J Taylor wrote: >>> Hi David, >>> >>> I'm certainly considering that board. It looks very attractive. I'm >>> going to have to read over the documentation page you've put >>> together. It looks very comprehensive. >>> >>> Chris Albertson said to get a unit which had: >>> >>> 1) - a stationary mode, or possibly only a stationary mode >>> 2) - self survey >>> 3) - PPS >>> 4) - serial output >>> 5) - NTPD driver support >>> >>> Now I know the Sure board has 3) and 4). And I presume you can do 5) >>> through the generic NMEA driver. However, do you know if it can do >>> 1) and 2)? Actually, does it have native PPS, or are you adding it >>> by using the LED output? >>> >>> Sincerely, >>> >>> Ron >> >> (1) and (2) are not required for the microsecond level accuracy you >> are going to see in an NTP application, Ron. In common with other GPS >> timing devices, the Sure board doesn't have those features, or if it >> does, I've not used them. It's my understanding that they only become >> significant below the microsecond level. Other uncertainties will >> contribute rather more than (1) or (2) will reduce. (1) might help if >> you don't have a reliable view of the sky - in practice with today's >> sensitive receivers being on the top floor of a building with just the >> roof in the way may well be enough, particularly in a domestic >> environment. See how well other GPS devices perform in your location. >> >> Any further questions which aren't answered on my Web page, please ask. >> >> Cheers, >> David > > Hi David, > > I just finished reading your Sure board web page. That's a really good > write up. I now understand how you're getting the PPS signal. Could > you possibly share the rev level of the board, since, if they change the > design, you patch instructions may not work any more? Some boards don't > have a rev level. > > As suggested by Brent Gordon's post, and confirmed in the manual for the > Trimble Copernicus II, I have switched my GPS to output only the GPZDA > sentence and changed the NMEA driver to respond to that. I don't think > the GPZDA sentence has any position information, so it's length should > be very consistent. Whatever the reason, the results are amazing. > Although I've only been running this way for 8 hours, my peak offsets > are now in the +/- 6 ms range. Of course, we'll have to see if the > system continues this way, or if my GPS goes phycho again. This is > twice the level of performance and accuracy than I had before.
PPS, at least on Linux, can give over 1000 times better accuracy. > > Now, I will probably try some PPS stuff, just for intellectual reasons > and just because it would be really cool to see that chart drop under 2 > or 1 ms. However, for my simple purpose of just keeping all my PC > clocks right to the second, if I can keep this thing under 10 ms of > error, I'm pretty happy with it. And this is still USB only. > > Assuming it's possible to program the Sure board for GPZDA, it would be > interesting to see if doing that affects the performance you're seeing. > Of course, if you're using PPS, the specific content of the NMEA > sentence may not matter as much. GEt the manual from Mediatex MTK NMEA Packet User Manual, which gives a far far more extensive set of nmea programming instructions for the chipset that Sure uses. > > Sincerely, > > Ron > _______________________________________________ questions mailing list [email protected] http://lists.ntp.org/listinfo/questions
