How can I test IRIG-B'a accuracy using Labview

[JYC]

We have a GPS source as reference, how can we test second GPS clock with IRIG-B output ?  Can we by using Labview ?

[Ryan_V[DE]]

JYC,

What other hardware do you have? Traditionally, to test the accuracy of anything, you need a device with accuracy much greater than that which you are testing. The NI-6608 will allow you to decode an IRIG or PPS signal, but its measurement capabilites will not likely provide the versatility to decode a complex waveform, like that of IRIG-B. If you were performing simple PPS measurements, you could get more accurate data (the period of the pulse), limited only by the 6608's onboard 80Mhz oscillator. We also offer high speed digital boards (656x series) which would allow you to capture your entire IRIG waveform, then post-analyze it. Please let us know what exactly you are trying to do, and I will be more than happy to help.

Regards,

[jspaarg]

Generally speaking most systems that generate IRIG B also generate a PPS.  The IRIG B zero crossing is then concurrent with the PPS plus any analog delays.  Good systems have ways of compensating for the typical 15-30 microseconds of analog delays.

Rather than testing the IRIG B, can you test the PPS (assuming there is one).

I spent a number of years doing time code generation and analysis and never found any tool better than an oscilloscope and delayed triggers for IRIG.

[Ryan_V[DE]]

jspaarg,

Is the IRIG B delayed from the PPS by 15-30us? I guess that is caused by the digital logic to create the IRIG B signal (I'm assuming TTL). Thanks for the great info!

[jspaarg]

Most timecode generators work by taking an external input (GPS, IRIG, PPS, or any of a number of other timecodes such as MILA, NASA36, IRIG A, etc) and outputting an on-time signal.

The most common systems now use a GPS receiver to steer an oscillator to some fixed frequency (10MHz is a favorite) and then use that steered oscillator to generate the outputs.  The one PPS from a GPS receiver is not very smooth and jumps around easily a few hundred nano-seconds every PPS (at least they used to).

Time codes are usually done by generating an on-time digital signal in which the rising edge of the frame bits is co-incident with the PPS timing.  The digital signal is then sent through some analog magic (op amps and whatever) which delays the output on time.

In digital signals, the rising edge is considered on time, in analog it is the zero-crossing.

Good systems will calibrate out the analog delay in the factory by moving back the on-time timing that is fed to the analog so that all on-time marks are co-incident.

[JYC]

Really appreciate the responses.   Here is what I think.    We have one GPS source with 1 PPS as well as IRIG-B and a to-be-tested GPS receiver.  and I could hook both of them to the testing platform.  Then  I could do the analysis by comparing two. The test platform could specify which IRIG-B and also log 24 hours to calcuate the accuracy since the accuracy is statistic number.  

[jspaarg]

It sounds like you are using one GPS source to test another.

If the first GPS source has a one PPS output, then compare that to the GPS under test PPS output.

You don't need to use the IRIG as a comparison.

Keep in mind that unless the PPS is steered (unlikely) the PPS will jump by a few hundred nano seconds every second.

Check the spec for your receiver.

That means that you can have a 2X error between your two receivers and still be in spec.

[Ryan_V[DE]]

Thanks again for the info jspaarg!

JYC,

What are the capabilities of this test platform you speak of? Does it already exist, or are you considering building it? What is it about the signals you intend to test? Do you want to test for compliance of the IRIG waveform, or do you want to test the timing accuracy? As jspaarg explained, it would be far easier to test the PPS signals if you want timing infomation, as there is nothing to decode, and no output circuitry to potentially skew your accuracy (if not properly calibrated).

Regards,

[Ryan_V[DE]]

I second that jspaarg. I was replying as you were posting...

[JYC]

The problem is that our end user needs to check IRIG-B too.  I know there is standalone instrument which can check two PPS.  But we need to test IRIG-B too, even its driving/loading ability