[Soekris] NET4501 and ntp server.

Sat Sep 2 15:30:47 UTC 2006

Did either of you two figure this problem out?

I have very much the same configuration as Gabriel. Soekris 4501, FreeBSD
5.4-RELEASE #0:, and the Ilotus (Motorola) M12M GPS.
I plan to use the GPIO and TMR1IN inputs, but first thought I would try the
unit using the serial DCD input for the PPS to see how it performs as a
standard(?) stratum 1 unit.

In the ntp log I get the following frequent messages.

 2 Sep 23:04:54 ntpd[413]: kernel time sync disabled 2907
 2 Sep 23:05:11 ntpd[413]: kernel time sync enabled 2107
 2 Sep 23:05:26 ntpd[413]: kernel time sync disabled 2907
 2 Sep 23:05:42 ntpd[413]: kernel time sync enabled 2107
 2 Sep 23:06:44 ntpd[413]: kernel time sync disabled 2907
 2 Sep 23:07:02 ntpd[413]: kernel time sync enabled 2107

Also ntptime shows errors equal to the intervals count.
eg.   intervals 479, jitter exceeded 119, stability exceeded 0, errors 479.

Some relavent messages out of dmesg are as follows. I also would like to fix
the last two messages, but at this stage I think they are irrelavent.

Copyright (c) 1992-2005 The FreeBSD Project.
Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
        The Regents of the University of California. All rights reserved.
FreeBSD 5.4-RELEASE #0: Sat May 27 00:13:00 WST 2006
    root at FreeServer:/usr/obj/usr/src/sys/SOEKRIS3
Timecounter "i8254" frequency 1193182 Hz quality 0
CPU: AMD Enhanced Am486DX4/Am5x86 Write-Back (486-class CPU)
  Origin = "AuthenticAMD"  Id = 0x494  Stepping = 4
  Features=0x1<FPU>
real memory  = 67108864 (64 MB)
avail memory = 60358656 (57 MB)
npx0: <math processor> on motherboard
npx0: INT 16 interface
cpu0 on motherboard
sysctl machdep.i8254_freq=1189161 returns 0
Timecounter "ELAN" frequency 8333333 Hz quality 1000
Timecounters tick every 6.666 msec
Elan-mmcr driver: MMCR at 0xc5888000. PPS support.
ad0: FAILURE - SETFEATURES ENABLE RCACHE status=51<READY,DSC,ERROR>
error=4<ABORTED>
ad0: FAILURE - SETFEATURES ENABLE WCACHE status=51<READY,DSC,ERROR>
error=4<ABORTED>

Any ideas from anyone would be welcome.
Thanks.

----- Original Message -----
From: "Frank Mayhar" <frank at exit.com>
To: "Gabriel A. Reyes-Davila" <gard at cgic.ucol.mx>
Cc: <soekris-tech at lists.soekris.com>
Sent: Sunday, January 29, 2006 2:59 PM
Subject: Re: [Soekris] NET4501 and ntp server.

> Digging through old email I found this...  I don't know if you've
> figured things out, but I'm replying because I've run into one of the
> problems you encountered.
>
> On Thu, 2005-12-08 at 16:37 +0000, Gabriel A. Reyes-Davila wrote:
> > I am implementing a ntp server using an out of the box NET4501, freebsd
> > 5.4,
>
> FreeBSD 6-stable, in my case.
>
> > and a GsynQ/T gps (Motorola's M12+ Timing) connected to serial port 1.
> > Problems:
> > 1.- the PPS signal is not being echoed to pin 6 of JP3 in the 4501.
>
> This sounds like a misconfiguration somewhere.  My setup is echoing the
> PPS just fine.
>
> > 2.- Output of ntptime is:
> >      ntp_gettime() returns code 0 (OK)
> >          time c742ce86.44e6c07c  Thu, Dec  8 2005 15:23:50.269,
> > (.269146468),
> >          maximum error 1229 us, estimated error 10 us, TAI offset 0
> >      ntp_adjtime() returns code 0 (OK)
> >          modes 0x0 (),
> >          offset -0.296 us, frequency -20.811 ppm, interval 4 s,
> >         maximum error 1229 us, estimated error 10 us,
> >         status 0x2107 (PLL,PPSFREQ,PPSTIME,PPSSIGNAL,NANO),
> >         time constant 4, precision 0.001 us, tolerance 496 ppm,
> >         pps frequency -20.811 ppm, stability 0.000 ppm, jitter 1.569 us,
> >         intervals 1033, jitter exceeded 139, stability exceeded 0,
> > errors 1033.
>
> Unfortunately, I'm seeing this.  Basically, the FLL never calibrates
> because the cumulative time count (over the calibration interval) is too
> high, in my case by more than two times.  The cutoff is 2000000 over
> four seconds, but I'm seeing numbers in the range 4894797 to 4895397 (to
> pick two examples).  My suspicion is that the TMR1IN signal has
> problems; I'm saving the TMR1 count from the last second and displaying
> it; it increases over four seconds then wraps.  This leads me to believe
> that TMR1 isn't being reset by the PPS signal.  Why that might be I
> don't know, although I'll check the solder connection (again) tomorrow.
> The GPIO pin, though, is certainly seeing it.
>
> Any hints you (or anyone else on the list) could give me would be
> appreciated.  Thanks.
> --
> Frank Mayhar frank at exit.com     http://www.exit.com/
> Exit Consulting                 http://www.gpsclock.com/
>                                 http://www.exit.com/blog/frank/
> _______________________________________________
> Soekris-tech mailing list
> Soekris-tech at lists.soekris.com
> http://lists.soekris.com/mailman/listinfo/soekris-tech
>