- Hi... When you did a re-install, did you clean out (after backing up) all the old data? Only ask, as what you see in the delay statistics, may be from theMessage 1 of 16 , Mar 8 1:59 AMView SourceHi...
When you did a re-install, did you clean out (after backing up) all the
Only ask, as what you see in the delay statistics, may be from the prior
install, if you just installed over the old one.
It's always worth clearing that data in the app, then letting the system
gather new data for a day or so, before deciding what's going on. You
may find a new peak in the data up in the 100's of mS as a result.
Check your NTP time source(s). It's not unknown for ISP's to do silly
things, like decide to take a NTP server off the public access list, or
to move the service to another system that varies so much as to make it
useless for apps'a like Faros.
Faros is also "a little unkind" to NTP servers too I've been told, due
to the poll rate it uses, from prolonged discussions on a NTP newsgroup
with people who are "in the know" re these things.
I've ended up (for several reasons) creating my own local (on my LAN)
NTP server, locked to GPS time, just like the beacons themselves. Not
a trivial task as it turned out, but I started doing that based on this
There's a lot to read, and even then, it's not exactly a "how to" as to
do it. But I can help there if needed. It does work, and works very
well indeed, even running on Windows! More than good enough for Faros.
Even without the GPS locking, a local NTP server on your LAN (even
running in the background on an existing box) will service Faros just
fine, and only sync periodicaly with external NTP sources, if you still
have a good one or three to use.
However... If your ISP or WAN provider is messing about, and the ping
times to the servers vary significantly throught the day (as mine were
doing) then you *need* the local GPS locked type of NTP source.
Most ISP's sadly will be happy if their NTP servers are within 1S of
each other, let alone within mS (or better) of UTC ticks. Contact them
to complain about jitter and huge offsets, and if you find someone who
doesnt think you're talking about FTP, they still wont have a clue what
you're on about.
Hope you find out what and why, and it's not NTP issues. Resolvable if
that is the case, but it's a bit of a chore. Help available if needed.
- Jerry, an interesting article I just noticed that might pertain to this issue. A project worth following, I think. If NASA is concerned why not FAROS!?Message 2 of 16 , Mar 11 5:55 PMView SourceJerry, an interesting article I just noticed that might pertain to this
issue. A project worth following, I think. If NASA is concerned why not
On 3/5/2010 5:28 PM, Bruce Tanner wrote:
> Wow! I don't know if this is of any value or significance to the issue,
> Jerry, but I have heard that an effect of the recent Earthquake in Chile
> there scientists have been able to record both a very minor change in
> the tilt of the Earth on her axis and a very slight speed up (I think)
> of the period of rotation. Could that be showing up in the Faros data?
> The Chilean earthquake was about two weeks ago now, Feb. 27.
> There are other reports if you Google 'earthquake rotation axis'.
> Bruce, K2BET
> On 3/5/2010 8:02 AM, Jerry Flanders wrote:
> > Faros was running FB until a couple weeks ago, with almost all delays
> > on the short path line as expected. Now delays appear to be in the
> > range of 200 - 300 ms and all delay points are at top of details screen.
> > I can't find any setting that appears wrong to me. I have uninstalled
> > and reinstalled Faros - no help.
> > Current settings:
> > Signal delay red vertical line is about 17 ms with a "normal" (no pun
> > intended) distribution of green lines centering at about 15 ms,
> > extending from about 0 to about 46 ms.
> > Delay correction is set to 17 ms.
> > UTC Clock info std error 0.99 ms, max error 18.6 ms.
> > Log file shows even high SNR stations are recording delays of 200 -
> > 300 ms.
> > Any ideas?
> > Jerry W4UK
> [Non-text portions of this message have been removed]
[Non-text portions of this message have been removed]