WXforum.net

Weather Station Hardware => Blitzortung => Topic started by: n2qew on September 09, 2014, 12:23:47 PM

Title: 1 pps accuracy question
Post by: n2qew on September 09, 2014, 12:23:47 PM
Just wondering what most people see for the 1 pps accuracy. Asking because I'm using a different GPS than usual because the one I received with the kit was defective. Not worth the shipping to replace it, so using something else.
Title: Re: 1 pps accuracy question
Post by: Cutty Sark Sailor on September 09, 2014, 12:45:41 PM
Hey... N2Qew... what's your station??? Are you 1156 Wilmington?
You mean something like this poor sample?
Title: Re: 1 pps accuracy question
Post by: miraculon on September 09, 2014, 01:21:06 PM
Here is what mine is (as of now):

PPS Accuracy Mean: 109.1ns, Current: 47ns

Greg H.

Title: Re: 1 pps accuracy question
Post by: n2qew on September 09, 2014, 01:25:16 PM
Yes I am in Wilmington, NC. And, yes, I'm talking about the mean accuracy at the bottom of the screen. That is an indication of how accurate the GPS receiver is at marking the time as referenced to UTC. It is pretty much the basis of this entire system. I suspect that the way it's done is that the running counter is stored at the moment that the board decides that a strike has occurred. And the delta from the last UTC timestamp as marked by the time pulse.

The receiver I'm using to test is a FEI GPS / Rubidium. Older, but intended purpose was precision timing. I just wanted a sanity check on what looks "normal". I'm in the ~40ns range for accuracy. I've got some Trimble ace ii receivers that should work OK.
Title: Re: 1 pps accuracy question
Post by: W3DRM on September 10, 2014, 01:00:48 PM
Mine is as follows:

PPS Accuracy Mean: 75.8ns, Current: 47ns

However, I also see those values bouncing all over the place with Current values going from 66ns to -156ns periodically.
Title: Re: 1 pps accuracy question
Post by: Einar on September 10, 2014, 03:25:37 PM
I think you're talking about an influence on TOA location of 20-50m here.
When the system as a whole becomes accurate enough that 156ns is significant that will be jawdropping!
There are other places to hunt for perfection.
Time to relax and have a homebrew?
Title: Re: 1 pps accuracy question
Post by: b.e.wilson on September 10, 2014, 09:20:06 PM
I thought I read in the manual that the clock used to check the 1 pps signal has a frequency corresponding to 12 ns per sample, so the 1 pps signal has at a minimum that much slop in the measurement, no matter how accurate the GPS signal.