Author Topic: Are there any GPS NTP sever available with 1 Nano second/day accuracy ?  (Read 1307 times)

0 Members and 1 Guest are viewing this topic.

Offline MrOmnosTopic starter

  • Frequent Contributor
  • **
  • Posts: 268
  • Country: np
  • BE in Electronics and Communication
Hi, I was looking for a rack mounted NTP sever with 1 nano second/ day accuracy for a personal project. I have found few with 15nano second. Do you guys have any suggestions?
 

Offline DC1MC

  • Super Contributor
  • ***
  • Posts: 1917
  • Country: de
The things you're talking about are military/nation state things. I don't mean Nepal military, but US military  :-DD, so I would say that chances are pretty slim to get one.

Cheers,
DC1MC
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 21227
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
Hi, I was looking for a rack mounted NTP sever with 1 nano second/ day accuracy for a personal project. I have found few with 15nano second. Do you guys have any suggestions?

Make sure you don't change the length of the connections to other equipment by more than a foot :)

There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline MrOmnosTopic starter

  • Frequent Contributor
  • **
  • Posts: 268
  • Country: np
  • BE in Electronics and Communication
I have one more question. When a manufacturer says the accuracy is 15 nano second or 1 nano second do they mean the hardware can determine time up to 3 nano second and not really the accuracy of the acutal time they recive?
 

Offline switchabl

  • Frequent Contributor
  • **
  • Posts: 445
  • Country: de
Hi, I was looking for a rack mounted NTP sever with 1 nano second/ day accuracy for a personal project. I have found few with 15nano second. Do you guys have any suggestions?

What exactly are you trying to achieve? The precision of NTP is usually somewhere in the millisecond range.

If you are worried about long-term stability, don't be. Over the long term, GPS timing is as good as USNO's master clock (= as good as it gets).

If you really need 1 ns/day short-term, then you would need a hydrogen maser (and you definitely can't use NTP).
« Last Edit: April 23, 2023, 08:52:18 am by switchabl »
 

Offline MrOmnosTopic starter

  • Frequent Contributor
  • **
  • Posts: 268
  • Country: np
  • BE in Electronics and Communication
I have question for you since you seem to know about the clocks. When the manufacturers say accuracy is 3ns/day, does it mean that they are talking about the receiver accuracy, the receiver can keep time within 3ns?  we know that the GPS signal can never give that accuracy? Are there two different types of accuracy, GPS signal has an accuracy  and the receiver has its own accuracy which could be better than the GPS signal.
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 21227
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
I have one more question. When a manufacturer says the accuracy is 15 nano second or 1 nano second do they mean the hardware can determine time up to 3 nano second and not really the accuracy of the acutal time they recive?

For deep fundamental reasons, there is no such thing as "the actual time".

Start by considering three boxes. A transmits a signal simultaneously to box B sitting next to it and to box C one km away. On receiving that signal, boxes B and C instantly send a signal back to A. If there is a single "actual time", why doesn't A see two signal return at the same instant? Since A doesn't see that, what is the implication for the concept of "the actual time".
There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Online tggzzz

  • Super Contributor
  • ***
  • Posts: 21227
  • Country: gb
  • Numbers, not adjectives
    • Having fun doing more, with less
I have question for you since you seem to know about the clocks. When the manufacturers say accuracy is 3ns/day, does it mean that they are talking about the receiver accuracy, the receiver can keep time within 3ns?  we know that the GPS signal can never give that accuracy? Are there two different types of accuracy, GPS signal has an accuracy  and the receiver has its own accuracy which could be better than the GPS signal.

Make sure you understand this:

There are lies, damned lies, statistics - and ADC/DAC specs.
Glider pilot's aphorism: "there is no substitute for span". Retort: "There is a substitute: skill+imagination. But you can buy span".
Having fun doing more, with less
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
I have question for you since you seem to know about the clocks. When the manufacturers say accuracy is 3ns/day, does it mean that they are talking about the receiver accuracy, the receiver can keep time within 3ns?  we know that the GPS signal can never give that accuracy? Are there two different types of accuracy, GPS signal has an accuracy  and the receiver has its own accuracy which could be better than the GPS signal.
3ns/day means the clock can drift by 3ns a day. But this would be a pretty good clock (better than Rubidium IIRC)

Over a long term (> week) time from GPS is accurate. Short term (<48 hours) it will drift about +/-50ns. So if you want to have stable (I'll get to accurate later on) time from a GPS receiver you will need to have a local oscillator (OCXO, Rubidium or Cesium) that is corrected slowly using the GPS time. With a Cesium clock you can achieve good stability by doing a tiny amount of trimming (= averaging the GPS time); so little that the Cesium is basically freerunning. OCXO and Rubidium will drift much quicker and require more adjustments so the time is not that stable.

Where it comes to accuracy: you'll need to calibrate the cables for the GPS receiver to take their time delays into account but also think about the internal delays up to the point where you need to use the time (for timestamping for example).

The GPS time is sourced from the USNO (US Navy Observatory)  by using a whole bunch of Cesium clocks. And typically this time source is also compared with many other institutes who have their own clocks. There are lists available that tell you how much GPS time was off so you can correct timestamps afterwards as well.

However, the use of NTP suggests you are looking for a clock source that stays on track over a long period. For that a GPS receiver will do just fine. Likely the NTP server will have an OCXO internally to keep track of the GPS time and keep going during a GPS outage / breakage. That is also where the 15ns/day number is coming from: the internal oscillator (likely Rubidium) can drift that much when it is not corrected from the GPS time.
« Last Edit: April 23, 2023, 12:56:33 pm by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline mendip_discovery

  • Super Contributor
  • ***
  • Posts: 1024
  • Country: gb
Take a look at the BIPM Circular-T to get an idea of how the national labs compare on time.

IIRC several national labs offer a time service for the likes of stock market places.

Motorcyclist, Nerd, and I work in a Calibration Lab :-)
--
So everyone is clear, Calibration = Taking Measurement against a known source, Verification = Checking Calibration against Specification, Adjustment = Adjusting the unit to be within specifications.
 

Online Halcyon

  • Global Moderator
  • *****
  • Posts: 6126
  • Country: au
« Last Edit: April 23, 2023, 01:57:55 pm by Halcyon »
 

Offline 5U4GB

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: au
Re: Are there any GPS NTP sever available with 1 Nano second/day accuracy ?
« Reply #11 on: April 29, 2023, 10:58:55 am »
https://timemachinescorp.com/product/gps-ntpptp-network-time-server-tm2000/

There is also the https://timemachinescorp.com/product/gps-ntpptp-network-time-server-10mz-output-tm2500/ if you want a 1 PPS or 10 MHz output signal.

Their web pages don't mention 1ns accuracy but the manual claims +/- 10ns, and +/- 20ns on the 1PPS output.  OTOH if all you want is a generic GPS NTP server then the TZT one does the same thing and is far cheaper, they're all over eBay and Aliexpress.  They claim 100ns accuracy, which is close enough to the TimeMachines one.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf