"David Schwartz" <[EMAIL PROTECTED]> writes:
[...]
> No. But unless you have some specialized hardware, you're just kidding
> yourself if you think you can get more than that. Even with a PPS input
> accurate to 200nS and the PPS kernel patches, accuracy is still only about 3
> microseconds
Bjorn Gabrielsson <[EMAIL PROTECTED]> writes:
> Ulrich Windl <[EMAIL PROTECTED]> writes:
> >
> > You may play with some PPS API offset adjustments or NTP driver fudge flags.
>
> Could you elaborate on this. I have a system with refclock_NMEA,
> 2.4.21-nano-kernel. I want to adjust this system 13
In article <[EMAIL PROTECTED]> Bjorn Gabrielsson
<[EMAIL PROTECTED]> writes:
>
>I know what TIOCMIWAIT gives. Is there a user-space function that can
>return time with better than than the 1us resolution that
>gettimeofday() gives?
To answer your question in the literal sense, "there is" the funct
"X.B." <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> The computer i'm using has no cmos battery, so each time it is cold
> rebooted, the internal clock is set to something around 01-01-2000.
> I've gone through the source code, and found that antenna's NTP packets
> are thrown
I recall an open bug on this item, but I can't find it offhand.
If you have a filesystem on the machine you can periodically touch a file
(perhaps /etc/ntp.drift will do) and at boot time (before ntpd -gN) set the
system time to the timestamp on that file.
H
__
Tony wrote:
> Having a fairly large NTP deployment, I would like to perform a
> hardware
> measurement of one of the NTP servers (Stratum 3) against the PPS
> signal
> from the GPS on the Stratum 1
Actually, a user-space program works quite well. I use the parallel
port to generate a 1 pps. T
NTP Public Services Project rehosts the NTP Bugzilla
Redwood City, CA - May 9, 2005 - The NTP Public Services Project
(http://ntp.isc.org) is pleased to announce the rehosting of the
NTP Bugzilla to its facilities at ISC.
The NTP Public Services Project Bugzilla is located at
http://bugs.ntp.isc.
"David Schwartz" <[EMAIL PROTECTED]> writes:
> "Bjorn Gabrielsson" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]
>
> > "David Schwartz" <[EMAIL PROTECTED]> writes:
>
> >> No. But unless you have some specialized hardware, you're just
> >> kidding
> >> yourself if you think
"Bjorn Gabrielsson" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> David, I am not kidding myself. Lets skip the absolute accuracy. Say
> you want to measure the time between two positive edges of a
> signal. Let this signal be a 1PPS, 10PPS or 100PPS generated by a
> high quality
"Bjorn Gabrielsson" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> "David Schwartz" <[EMAIL PROTECTED]> writes:
>> No. But unless you have some specialized hardware, you're just
>> kidding
>> yourself if you think you can get more than that. Even with a PPS input
>> accurate
"David Schwartz" <[EMAIL PROTECTED]> writes:
> No. But unless you have some specialized hardware, you're just kidding
> yourself if you think you can get more than that. Even with a PPS input
> accurate to 200nS and the PPS kernel patches, accuracy is still only about 3
> microseconds.
>
>
"David Schwartz" <[EMAIL PROTECTED]> writes:
> "Bjorn Gabrielsson" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]
>
> > I know what TIOCMIWAIT gives. Is there a user-space function that can
> > return time with better than than the 1us resolution that
> > gettimeofday() gives?
>
At 9:29 PM + 2005-05-09, Bjorn Gabrielsson wrote:
I know what TIOCMIWAIT gives. Is there a user-space function that can
return time with better than than the 1us resolution that
gettimeofday() gives?
No. You're limited to the level of resolution of the system
clock, and it only runs so fa
"Bjorn Gabrielsson" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> I know what TIOCMIWAIT gives. Is there a user-space function that can
> return time with better than than the 1us resolution that
> gettimeofday() gives?
No. But unless you have some specialized hardware, you'
"David Schwartz" <[EMAIL PROTECTED]> writes:
> "Bjorn Gabrielsson" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]
>
> > "David Schwartz" <[EMAIL PROTECTED]> writes:
>
> >> You don't need the PPSkit because Linux TIOCMIWAIT. However you will
> >> get
> >> better accuracy with
"Bjorn Gabrielsson" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> "David Schwartz" <[EMAIL PROTECTED]> writes:
>> You don't need the PPSkit because Linux TIOCMIWAIT. However you will
>> get
>> better accuracy with the PPSkit.
>> http://www.webchat.org/~jk/shm_linux_clock.c
"David Schwartz" <[EMAIL PROTECTED]> writes:
> You don't need the PPSkit because Linux TIOCMIWAIT. However you will get
> better accuracy with the PPSkit.
> http://www.webchat.org/~jk/shm_linux_clock.c
>
Does it exist a "gettimeofday()"-version or replacement that gives you
nano-second res
"Ulrich Windl" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> "MarkO" <[EMAIL PROTECTED]> writes:
>> We're currently running ntp version [EMAIL PROTECTED] on linux kernel version
>> 2.4.18-4. We have a Garmin NMEA GPS connected to the serial port and
>> sending a (default) 100ms
Look at the refclock driver list. There are many add-in cards from
Meinberg, Symmetricom and others which are supported by refclock
drivers. These cards typically include a voltage-controlled crystal
oscillator and dac control circuit to provide much better stability than
your standard xtal as we
Ulrich Windl <[EMAIL PROTECTED]> writes:
>
> You may play with some PPS API offset adjustments or NTP driver fudge flags.
Could you elaborate on this. I have a system with refclock_NMEA,
2.4.21-nano-kernel. I want to adjust this system 13us. Another system
need a little more than 2 ms adjustment.
X.B. wrote:
Hello,
i'm using ntp ( 4.2.0 ) with a GPS Trimble 'Acutime' antenna and the
palisade driver.
Everything works ok, but i've the following problem :
The computer i'm using has no cmos battery, so each time it is cold
rebooted, the internal clock is set to something around 01-01-2000.
On Mon, 09 May 2005 15:45:47 +0200 in comp.protocols.time.ntp, "X.B."
<[EMAIL PROTECTED]> wrote:
>The computer i'm using has no cmos battery, so each time it is cold
>rebooted, the internal clock is set to something around 01-01-2000.
>I couldn't find any reasonable solution to overcome this pro
At 3:45 PM +0200 2005-05-09, X.B. wrote:
The computer i'm using has no cmos battery, so each time it is cold
rebooted, the internal clock is set to something around 01-01-2000.
In this case, ntpd -g -q cannot set up the date correctly.
Nor can ntpd -g ...
There is a 1000 second sanity check, a
Hello,
i'm using ntp ( 4.2.0 ) with a GPS Trimble 'Acutime' antenna and the
palisade driver.
Everything works ok, but i've the following problem :
The computer i'm using has no cmos battery, so each time it is cold
rebooted, the internal clock is set to something around 01-01-2000.
In this case
Ulrich,
Not all messages received from servers result in an update to the system
clock. This has always been the case and is very definitely intentional.
Only those messages that do result in update result in loopstats
messages. See the briefings at the NTP project page. The Autokey
provision
Hi,
for a long time I did not try newer versions of ntpd. However recently I tried
the version from SuSE Linux 9.2 with autokey (requires update from SuSE) from
a startum-1 GPS source. I notices that even when the "peer poll" was 64s, the
loopstats file wasn't updated in that period. Furthermose t
"Tony" <[EMAIL PROTECTED]> writes:
> Having a fairly large NTP deployment, I would like to perform a
> hardware
> measurement of one of the NTP servers (Stratum 3) against the PPS
> signal
> from the GPS on the Stratum 1. However I can't think of a way of
> getting the
> Linux ST3 to generate a pu
Maybe ask the people at Meinberg; clocks are their business:
http://www.meinberg.de
Regards,
Ulrich
"Joshua Coombs" <[EMAIL PROTECTED]> writes:
> Given standard pc clocks are less than ideal, with wide drift, etc, does
> anyone make a good quality oscillator as an add in, say via PCI? Ideally
>
"MarkO" <[EMAIL PROTECTED]> writes:
> We're currently running ntp version [EMAIL PROTECTED] on linux kernel version
> 2.4.18-4. We have a Garmin NMEA GPS connected to the serial port and
> sending a (default) 100ms PPS pulse on pin #1.
>
> I have a couple very basic questions. First, do we need
Rodolfo Giometti <[EMAIL PROTECTED]> writes:
> Sorry for delay...
>
> On Fri, Apr 08, 2005 at 02:39:55PM +0200, Ulrich Windl wrote:
> > What I undertstood is that this is simply still all in kernel space; right?
> > What should be the benefits from your point of view?
>
> The benefits are:
>
>
Abandoning the right to remain silent, Jonathan Buzzard at Sat, 07 May
2005 21:08:53 +0100 said:
> On Sat, 07 May 2005 03:05:10 -0700, David Schwartz wrote:
>
>>
>> "Jonathan Buzzard" <[EMAIL PROTECTED]> wrote in message
>> news:[EMAIL PROTECTED]
>>
>>> That is some very cheap GPS receiver or
31 matches
Mail list logo
