[ntp:questions] question about logs of PPS

Christian Pinedo Zamalloa christian.pinedo.zamalloaNOSPAM at NOKKgmail.com
Thu May 17 20:12:47 UTC 2007

On 2007-05-16, Jason Rabel <jason at extremeoverclocking.com> wrote:
> Christian,
> Did you disable your FIFO in the serial port settings? In the
> /boot/device.hints file? This *really* helped my PPS stability on my little
> NTP box.
> http://ntp.isc.org/bin/view/Support/KnownHardwareIssues
> For example, I have my serial port(s) set with the following flag:
> hint.sio.0.flags="0x12"
> ------------------------
> As to answer your other questions:
> 1. Best thing to do would be to run winoncore and see what your SNR levels
> are. Then you can determine if the bottle/tape is causing any problems. You
> can usually find good all-weather GPS antenna's on eBay for like $30.
> 2. If your coordinates are completely incorrect it could lead to some weird
> issues. Again, I like to run Winoncore and let it do its self-survey then so
> you can monitor the data.
> 3. Being near other equipment shouldn't be an issue, but you never know.
> Again, running winoncore and monitoring the data for a while should let you
> know where the trouble lies, and also configure your receiver completely.

First of all thanks for your comments, Jason and Richard, i think i have
find the fail: the /etc/ntp.oncore0 had a bad value of Longitude, it
should be negative and not possitive because the value of the longitude
is towards the west. I understand that this fail could produce the
unstable behaviour of the clock.

Anyway i have tested other things:

1. With winoncore i have tested the receiver and the aerial and they were ok.
2. The GPS finds several satellites (9 or more) with good levels of SNR
(now i discard the lack of visibility and the possibility of electronic
3. I have disable FIFO settings of serial port and ACPI support of
FreeBSD server that runs the NTP daemon.

Now the NTP server is not available (ntp.i2t.ehu.es) because the GPS is
running winoncore in a windows host to calculate position. I expect the
next week to run again the server and to see if these changes improve
the precision of the clock.

I will have you informed :). Thanks again,

Christian Pinedo Zamalloa

More information about the questions mailing list