ref clock for network POP locations - NTP

This is a discussion on ref clock for network POP locations - NTP ; I have some network locations where I have no way to access GPS or CDMA. But, I would still like to have a very accurate time source. (I would like sub-microsecond accuracy from NTPD if I can get it.) These ...

+ Reply to Thread
Results 1 to 12 of 12

Thread: ref clock for network POP locations

  1. ref clock for network POP locations

    I have some network locations where I have no way to access GPS or CDMA. But, I
    would still like to have a very accurate time source. (I would like
    sub-microsecond accuracy from NTPD if I can get it.) These locations are in
    telco-hotels, so they are typically very well shielded making any broadcast
    technology unworkable. (We are not using sonet - so getting time from there is
    also not an option.)

    I was thinking that it might be possible to get an oscillator to provide a PPS
    signal, and then use a very well connected (possibly even a dedicated network
    link with no router/switch) network peer as the preferred peer. Is this a
    reasonable solution? Can anyone suggest anything better?

    The PPS would need to be calibrated... Is the normal ntp calibration reasonable
    for this? I would guess it depends on the oscillator...

    If this is a reasonable solution, does anyone have recommendations for
    oscillators? Most of the ones I have seen want a reference source of their own.
    This of course does not work in my case...

    Thanks,
    jeff

    P.S. I will happily add information to the wiki on this after figuring out how
    to do it. (If?)

  2. Re: ref clock for network POP locations

    Jeff W. Boote wrote:

    > I have some network locations where I have no way to access GPS or CDMA.
    > But, I would still like to have a very accurate time source. (I would
    > like sub-microsecond accuracy from NTPD if I can get it.) These
    > locations are in telco-hotels, so they are typically very well shielded
    > making any broadcast technology unworkable. (We are not using sonet - so
    > getting time from there is also not an option.)
    >
    > I was thinking that it might be possible to get an oscillator to provide
    > a PPS signal, and then use a very well connected (possibly even a
    > dedicated network link with no router/switch) network peer as the
    > preferred peer. Is this a reasonable solution? Can anyone suggest
    > anything better?
    >
    > The PPS would need to be calibrated... Is the normal ntp calibration
    > reasonable for this? I would guess it depends on the oscillator...
    >
    > If this is a reasonable solution, does anyone have recommendations for
    > oscillators? Most of the ones I have seen want a reference source of
    > their own. This of course does not work in my case...
    >
    > Thanks,
    > jeff
    >
    > P.S. I will happily add information to the wiki on this after figuring
    > out how to do it. (If?)


    The only ways I know of to get "sub microsecond accuracy" from NTPD are:
    a. a GPS based reference clock (the PPS output is usually rated at 25-50
    nanoseconds) By the time you get to the computer's clock that degrades
    to a microsecond or two.
    b. a CDMA timing receiver
    c. a LORAN receiver
    d. an atomic (Cesium) clock

    Forget about doing it over a network. It might be possible but I
    suspect that a Cesium frequency standard with clock option would be
    easier and cheaper. (I believe that Cesium clocks go for $40,000 and up!)

    The problem with networks is that they introduce random "phase noise";
    e.g. if you set up a server in New York that sends packets at EXACT one
    second intervals, those packets will not arrive in Los Angeles at
    anything like one second intervals. Some of them may not arrive at all
    if you are using UDP. The best you can hope for using internet servers
    is 2-5 millisecond accuracy. The worst is much worse than that; much
    depends on the quality of the servers involved and more depends on the
    quality of the network connection between you and the server.

    If worst comes to worst, you can put the GPS receiver on the roof and
    run fiber from it to where you need the time.

    FWIW, Telcos tend to have some pretty serious timing stuff of their own;
    Cellular phones, T1 and T3 lines all rely on very precise timing.
    This may take the form of frequency standards rather than time standards
    but the two are intimately related; t=1/f and f=1/t

  3. Re: ref clock for network POP locations

    On Tue, 26 Sep 2006 13:56:09 -0600, "Jeff W. Boote"
    wrote for the entire planet to see:

    >I have some network locations where I have no way to access GPS or CDMA. But, I
    >would still like to have a very accurate time source. (I would like
    >sub-microsecond accuracy from NTPD if I can get it.) These locations are in
    >telco-hotels, so they are typically very well shielded making any broadcast
    >technology unworkable.


    >If this is a reasonable solution, does anyone have recommendations for
    >oscillators? Most of the ones I have seen want a reference source of their own.
    >This of course does not work in my case...


    Dr. Mills has related stories of him synchronizing his (portable)
    cesium clock at the USNO and carrying the time back to Delaware. I
    guess you could fire up a good oscillator locally and get a portable
    one to resynchronize and carry the time into your building every few
    months.

    - Eric


  4. Re: ref clock for network POP locations

    I believe rubidium is cheaper than caesium.

    Take a look at this site: http://www.quartzlock.com/ (don't be put off
    by the name!).

    Paul


  5. Re: ref clock for network POP locations

    Jeff W. Boote schrieb:
    > I have some network locations where I have no way to access GPS or CDMA.
    > But, I would still like to have a very accurate time source. (I would
    > like sub-microsecond accuracy from NTPD if I can get it.) These
    > locations are in telco-hotels, so they are typically very well shielded
    > making any broadcast technology unworkable. (We are not using sonet - so
    > getting time from there is also not an option.)
    >
    > I was thinking that it might be possible to get an oscillator to provide
    > a PPS signal, and then use a very well connected (possibly even a
    > dedicated network link with no router/switch) network peer as the
    > preferred peer. Is this a reasonable solution? Can anyone suggest
    > anything better?
    >
    > The PPS would need to be calibrated... Is the normal ntp calibration
    > reasonable for this? I would guess it depends on the oscillator...
    >
    > If this is a reasonable solution, does anyone have recommendations for
    > oscillators? Most of the ones I have seen want a reference source of
    > their own. This of course does not work in my case...
    >
    > Thanks,
    > jeff
    >
    > P.S. I will happily add information to the wiki on this after figuring
    > out how to do it. (If?)


    Hi Jeff,

    as already mentioned, sub-microsecond accuracy is not a possibility with
    NTP these days. Did you check out PTP (IEEE 1588) instead
    (ieee1588.nist.gov)?

    We offer a NTP appliance that has its own free-running OCXO-HQ that is
    disciplined by NTP over the network. Using a very smooth filtering and
    applying only very small corrections allows us to get down to 50 to 100
    microsecond accuracy in an "ideal" network environment, i.e.
    crossover-cable.

    The model is called LANTIME/NDT (Network Disciplined Timeserver) and you
    can find it on our website (www.meinberg.de, see Products section).

    I would say that you have to check out PTP if you really need to sync
    down to the nanoseconds ...

    Best regards,
    Heiko


    --
    Meinberg radio clocks: 25 years of accurate time worldwide

    MEINBERG Radio Clocks
    www.meinberg.de

    Stand alone ntp time servers and radio clocks based on GPS, DCF77 and
    IRIG. Rackmount and desktop versions and PCI slot cards.

  6. Re: ref clock for network POP locations


    Paul.Croome@softwareag.com wrote:
    > I believe rubidium is cheaper than caesium.
    >


    You'll want to look carefully at the specs before buying a rubidium.
    I've used a number of USD2000 rubidiums - these typically age at
    1 part in 10^12 per day so you would very quickly accumulate a
    significant
    phase error with one of these.

    Cheers
    Michael.


  7. Re: ref clock for network POP locations


    Heiko Gerstung wrote:

    > We offer a NTP appliance that has its own free-running OCXO-HQ that is
    > disciplined by NTP over the network. Using a very smooth filtering and
    > applying only very small corrections allows us to get down to 50 to 100
    > microsecond accuracy in an "ideal" network environment, i.e.
    > crossover-cable.


    I look after a national network (physical separations of thousands of
    km, maybe 6 to 10 router hops apart) of NTP servers which used to have
    a good network connection (high speed academic). Typically ntpd would
    report offsets between these of less than 100 microseconds, with a FWHM
    on the distribution of about 100 microseconds as well. So you CAN do as
    well as point to point.

    Cheers
    Michael


  8. Re: ref clock for network POP locations

    Heiko Gerstung wrote:

    > Jeff W. Boote schrieb:
    >
    >> I have some network locations where I have no way to access GPS or
    >> CDMA. But, I would still like to have a very accurate time source. (I
    >> would like sub-microsecond accuracy from NTPD if I can get it.) These
    >> locations are in telco-hotels, so they are typically very well
    >> shielded making any broadcast technology unworkable. (We are not using
    >> sonet - so getting time from there is also not an option.)


    > Hi Jeff,
    >
    > as already mentioned, sub-microsecond accuracy is not a possibility with
    > NTP these days. Did you check out PTP (IEEE 1588) instead
    > (ieee1588.nist.gov)?
    >


    That is not true. You can get microsecond accuracy or better using a
    reference clock. What you can't do is get that kind of accuracy over
    the internet.

    My only knowledge of PTP is based on some earlier messages here but I
    believe that it must have the same problems as NTP over the internet.




  9. Re: ref clock for network POP locations

    Richard B. Gilbert schrieb:
    > Heiko Gerstung wrote:
    >
    >> Jeff W. Boote schrieb:
    >>
    >>> I have some network locations where I have no way to access GPS or
    >>> CDMA. But, I would still like to have a very accurate time source. (I
    >>> would like sub-microsecond accuracy from NTPD if I can get it.) These
    >>> locations are in telco-hotels, so they are typically very well
    >>> shielded making any broadcast technology unworkable. (We are not
    >>> using sonet - so getting time from there is also not an option.)

    >
    >> Hi Jeff,
    >>
    >> as already mentioned, sub-microsecond accuracy is not a possibility
    >> with NTP these days. Did you check out PTP (IEEE 1588) instead
    >> (ieee1588.nist.gov)?
    >>

    >
    > That is not true. You can get microsecond accuracy or better using a
    > reference clock. What you can't do is get that kind of accuracy over
    > the internet.


    I was referring to sub-microsecond accuracy over the network, it seems
    that the OP has no chance to use a hardware ref clock.

    > My only knowledge of PTP is based on some earlier messages here but I
    > believe that it must have the same problems as NTP over the internet.


    PTP uses hardware timestamping at the MII/PHY level, this will not help
    when you use the Internet, but maybe the OP has his own nice WAN with
    low-jitter connections available. You can get in the lower nanoseconds
    with PTP over Ethernet, but only in very small networks or by using
    PTP-aware infrastructure components like switches with integrated
    hardware timestamping..

    >


    The mentioned LANTIME/NDT is basically an oscillator that is disciplined
    by NTP. And yes, its too expensive for using it at home :-)

    Best regards,
    Heiko


    --
    Meinberg radio clocks: 25 years of accurate time worldwide

    MEINBERG Radio Clocks
    www.meinberg.de

    Stand alone ntp time servers and radio clocks based on GPS, DCF77 and
    IRIG. Rackmount and desktop versions and PCI slot cards.

  10. Re: ref clock for network POP locations

    Heiko Gerstung wrote:
    > Richard B. Gilbert schrieb:
    > I was referring to sub-microsecond accuracy over the network, it seems
    > that the OP has no chance to use a hardware ref clock.
    >
    >> My only knowledge of PTP is based on some earlier messages here but I
    >> believe that it must have the same problems as NTP over the internet.

    >
    >
    > PTP uses hardware timestamping at the MII/PHY level, this will not help
    > when you use the Internet, but maybe the OP has his own nice WAN with
    > low-jitter connections available. You can get in the lower nanoseconds
    > with PTP over Ethernet, but only in very small networks or by using
    > PTP-aware infrastructure components like switches with integrated
    > hardware timestamping..


    I do have access to a very nice WAN with extremely low-jitter connections. I can
    likely remove nearly all buffered devices from the path - and get nearly the
    equivalent of a really long cross-over cable.

    >>

    >
    >
    > The mentioned LANTIME/NDT is basically an oscillator that is disciplined
    > by NTP. And yes, its too expensive for using it at home :-)


    Right - eventually it all comes down to price. How much accuracy can you
    afford... But, this at least sounds interesting.

    Thanks for the pointers.

    jeff

  11. Re: ref clock for network POP locations

    Jeff W. Boote wrote:

    > Heiko Gerstung wrote:
    >
    >> Richard B. Gilbert schrieb:
    >> I was referring to sub-microsecond accuracy over the network, it seems
    >> that the OP has no chance to use a hardware ref clock.
    >>
    >>> My only knowledge of PTP is based on some earlier messages here but I
    >>> believe that it must have the same problems as NTP over the internet.

    >>
    >>
    >>
    >> PTP uses hardware timestamping at the MII/PHY level, this will not
    >> help when you use the Internet, but maybe the OP has his own nice WAN
    >> with low-jitter connections available. You can get in the lower
    >> nanoseconds with PTP over Ethernet, but only in very small networks or
    >> by using PTP-aware infrastructure components like switches with
    >> integrated hardware timestamping..

    >
    >
    > I do have access to a very nice WAN with extremely low-jitter
    > connections. I can likely remove nearly all buffered devices from the
    > path - and get nearly the equivalent of a really long cross-over cable.
    >
    >>>

    >>
    >>
    >>
    >> The mentioned LANTIME/NDT is basically an oscillator that is
    >> disciplined by NTP. And yes, its too expensive for using it at home :-)

    >
    >
    > Right - eventually it all comes down to price. How much accuracy can you
    > afford... But, this at least sounds interesting.
    >
    > Thanks for the pointers.
    >
    > jeff


    Jeff,

    I think the killer keywords here are "disciplined by NTP". (BTW, I
    thought it was GPS disciplined) If it is, in fact, disciplined by NTP,
    it is no better than your NTP time source, be it reference clock, or
    internet server. If it is, as I thought, disciplined by GPS, then it's
    just a high grade crystal oscillator if you cannot apply the GPS signal
    to it. I question whether it can maintain "sub microsecond accuracy"
    without the GPS discipline.

    It's a very nice toy and, if I should hit the lottery for a couple of
    million, I'll probably buy one.

    If you really need that sub microsecond accuracy and must live with the
    constraint of "no radio reception" your best bet is a Cesium clock.
    It's expensive, but once calibrated by NIST it should drift no more than
    400 nanoseconds per year. If that 400 nanoseconds is significant,
    you'll want to have it calibrated annually.

  12. Re: ref clock for network POP locations

    Richard B. Gilbert schrieb:
    > Jeff W. Boote wrote:
    >
    >> Heiko Gerstung wrote:
    >>
    >>> Richard B. Gilbert schrieb:
    >>> I was referring to sub-microsecond accuracy over the network, it
    >>> seems that the OP has no chance to use a hardware ref clock.
    >>>
    >>>> My only knowledge of PTP is based on some earlier messages here but
    >>>> I believe that it must have the same problems as NTP over the internet.
    >>>
    >>>
    >>>
    >>> PTP uses hardware timestamping at the MII/PHY level, this will not
    >>> help when you use the Internet, but maybe the OP has his own nice WAN
    >>> with low-jitter connections available. You can get in the lower
    >>> nanoseconds with PTP over Ethernet, but only in very small networks
    >>> or by using PTP-aware infrastructure components like switches with
    >>> integrated hardware timestamping..

    >>
    >>
    >> I do have access to a very nice WAN with extremely low-jitter
    >> connections. I can likely remove nearly all buffered devices from the
    >> path - and get nearly the equivalent of a really long cross-over cable.
    >>
    >>>>
    >>>
    >>>
    >>>
    >>> The mentioned LANTIME/NDT is basically an oscillator that is
    >>> disciplined by NTP. And yes, its too expensive for using it at home :-)

    >>
    >>
    >> Right - eventually it all comes down to price. How much accuracy can
    >> you afford... But, this at least sounds interesting.
    >>
    >> Thanks for the pointers.
    >>
    >> jeff

    >
    > Jeff,
    >
    > I think the killer keywords here are "disciplined by NTP". (BTW, I
    > thought it was GPS disciplined) If it is, in fact, disciplined by NTP,

    We have several different NTP time servers, synchronized to GPS, NTP,
    IRIG B and so on. The NDT is the one that is using another NTP server as
    its primary reference.

    > it is no better than your NTP time source, be it reference clock, or
    > internet server. If it is, as I thought, disciplined by GPS, then it's
    > just a high grade crystal oscillator if you cannot apply the GPS signal
    > to it. I question whether it can maintain "sub microsecond accuracy"
    > without the GPS discipline.

    Sub-Microsecond accuracy is not possible with it, our measurements in a
    separated LAN show an accuracy of +/- 30 microseconds.

    > It's a very nice toy and, if I should hit the lottery for a couple of
    > million, I'll probably buy one.

    I would not call it a toy, there are actually people doing serious
    things with it, but we are getting off-topic :-)

    > If you really need that sub microsecond accuracy and must live with the
    > constraint of "no radio reception" your best bet is a Cesium clock. It's
    > expensive, but once calibrated by NIST it should drift no more than 400
    > nanoseconds per year. If that 400 nanoseconds is significant, you'll
    > want to have it calibrated annually.


    Best regards,
    Heiko

+ Reply to Thread