[ntp:questions] What is the NTP recovery time from 16s step in GPSserver?

unruh unruh at invalid.ca
Thu Nov 1 03:39:51 UTC 2012

On 2012-10-31, Richard B. Gilbert <rgilbert88 at comcast.net> wrote:
> On 10/31/2012 5:04 PM, unruh wrote:
>> On 2012-10-31, Richard B. Gilbert <rgilbert88 at comcast.net> wrote:
>>> On 10/31/2012 4:30 AM, David Woolley wrote:
>>>> Kennedy, Paul wrote:
>>>>> I believe the answer to your question is 12.5 minutes.
>>>>> This is the time it takes to receive the full set of 25 almanac frames,
>>>>> which contains the GPSTime/UTC offset (amongst other things).
>>>>> http://en.wikipedia.org/wiki/GPS_signals#Almanac
>>>> I think he knows the time taken for the GPS receiver, which is a lot
>>>> less than that.  His concern is about how long ntpd takes once the GPS
>>>> receiver is reporting the correct time.  As noted, ntpd is not specified
>>>> for this case, so makes no attempt to recover any faster than any other
>>>> broken local clock case.
>>>> The almanac you are referring to is a low resolution one to aid the
>>>> receiver in finding satellites after a cold start.  Once it has found a
>>>> satellite, it should have a high resolution almanac for that satellite
>>>> in about 30 seconds.  Modern receivers tend to decode multiple
>>>> satellites at once, which is how they get a fast start, so they may be
>>>> fully acquired in 30 seconds.  However, if there is no memory at all, it
>>>> may take them some tome to find their first satellite, and locating
>>>> subsequent ones may be slow until the full coarse almanac is received.
>>> NTPD is a "slow starter"!  Ideally, you will only start it once and
>>> let it run for a few months.
>>> How slow is a "slow start"?.  It can take NTPD up to ten hours to
>>> synchronize within + or - 50 nanoseconds with whatever you are using as
>> It will never get to within 50nsec. The interrupt processing is far more
>> variable than that. You might get to withing a few micro seconds.
>>> a time source.  If you must boot your computer at 8:30 every morning,
>>> NTPD is a poor choice!
>>> There is another "product" which will give you a "reasonable facsimile"
>>> of the correct time in a very short time.  I've never used it. I've
>>> forgotten its name.  Sorry about that.  I'm sure that someone here
>>> can recall the name I've forgotten!
>> Chrony.
>> It also gives better accuracy.
> Accuracy long term or short term?

The standard deviation of the time as compared with a reference time is
about a factor of three less than with ntpd.
That is over a long term-- days and weeks.


More information about the questions mailing list