How to use GPS to synchronize measurement at 2 different locations.

[MattBradley]

Sorry, been out of touch for a while...

Here's the link for the product.  If you have any questions, get back to me...

http://www.symmetricom.com/products/gps-solutions/bus-level-timing/bc637PCI-V2/

The 637 has the GPS option (the 635 does not).

[alex.]

Hi Mat,

Thanks a lot. May I know how you do the measurement.

What I plan to do is connect the 1PPS as a external triger signal for DAQ card. But still want to make it longer than 1 second for the time interval.

Also I heard there is another method, when the DAQ start collecting data, it will send a signal to GPS, and GPS reply a time info of it. But cannot find detail info on it.

Any suggestion, thanks.

I googled, and found some brand can reach 30ns for 1PPS, but never test it yet.

[Adam_U]

Sorry but just to confirm, have you considered the 6682 (I'm not sure if you're on PXI). If you are using PXI, the 6682 is probably the best way to go since it can generate not just a PPS but various clock frequencies all locked to GPS/IRIG/1588/etc. This way you don't incur clock drift or have to deal with resampling issues.

[alex.]

I am exlporing 6288 but comparing other product at same time.

The NI guy said the error is only 15ns for 1PPS, but on the manual said 100ns, a bit confused of it.

Except using 1PPS as external trigger for DAQ, is there any other good way to use it, thanks.

[MattBradley]

Not sure if I can answer your question on how to make the measurement without knowing more details.  There are LabVIEW drivers available for the 637, if that's what you mean.

[Adam_U]

This can depend on a series of factors. There are some methods to reduce the jitter/uncertainty in a generated signal if you're interested. Is your specification of "<10ns" a standard deviation or max/min offset? I'd be very suspect of a GPS solution that claims <10ns max/min.

The 6682 can be used to generate the sample clock directly (and trigger it) or timestamp acquisitions. All of these signals are synchronized to GPS or your selected time reference. It can also route discrete digital signals from the front or bridge to other time synchronization like IEEE-1588.

I believe the best performance for your solution can probably be achieved with the 6682. 

[alex.]

The application is to locate a fault inside the cable. The fault is a source of pulse signal.

When the pulse propogating inside the cable,  2 data acqusition device were put on both end of the cable to capture the pulse. By recording the arriving time difference, we can tell the pulse location.

But we need GPS to synchronize the time base at both ends. The pulse speed is half of light, so we need 10ns error for 1PPS output.

The soluction has not fixed yet.

[Adam_U]

Take a look at this link on sync'ing DSA with GPS. While you're using DAQ, the technology is comparable. Note that we will soon have the ability to use 2 NI devices to get very similar performance (<10ns std. deviation) so you don't need an external device.

[alex.]

May I know the link.

May I know the approximate concept or info of the new member?

[Adam_U]

Sorry about that: http://zone.ni.com/devzone/cda/tut/p/id/6818