Time stamping and recording analog data with gps time

[mschleppe]

I am very new to the LabView programming environment so I do not yet fully understand how to use some of the features. I am trying to take analog data and time stamp it with gps time using the PXI 6608 device and LabView. So far I have been able to synchronize the clock in the PXI 6608 device with the pps from a gps reciever, but I have not been able to been able to tag analog events with gps time. So I guess my question is, how do I take a continuous steam of analog data, measure the values of the data, and then stamp it with a gps time and record it to a file?

I know this is a big question, but I already know how to do some of this such as bring in the data, I just don't seem to be able to put a gps time with it.

Thankyou

[Ray.R]

This is an interesting challenge indeed.

The gps provides pulses at 27MHz, right, with an amazing precision (1E-10??)??

My first question is why do you want to use a gps? This means having the ability of tracking pulses that are occuring faster than the time to process the data.

It has been a while since I played with a gps system. Which may correct my second line above. We had used a gpm device to provide an accurate 10MHz clock signal to synchronize RF test equipment.

However, in your case, are you obtaining pulses (clock signal only)? or is there an intermediate device which translates the pulses into some sort of counter? If you have some sort of counter, then that simplifies your life greatly... Let me explain.

You are acquiring analog data. And using the gps to synchronize the PXI6608. However, you also want to use the clock signal as a timestamp to tag the analog data (or events). If I understand this correctly, then the challenge is to be able to count each and every clock signal and convert it to a timestamp. I doubt that LV will be able to process everything including the timestamp to permit you to do this... (I am trying to think of a solution)...

Do you actually need the accuracy? You can take an initial timestamp from the computer clock and deltas from then on and store this with the data. How many significant digits do you need for the timestamps and what if the frequency of the events?

There was a post (2 years ago??) that mentionned the time per iteration of a simple while loop. The value 1ms or 4ms comes to mind. If that is the case, the best resolution of time would be in the order of msec. Therefore the pc clock would be provide a relatively accurate timestamp for the data.

I seem to understand what you want to do. Have a precise snapshot of "pps" for each event. This would provide a very precise delta of time between the events. However, the limitation may be with the speed of LV to process the event and the timestamp..

I hope I am not the bearer of bad new... 😉

I will re-visit this thread. Hopefully someone can come up with some sort of solution.

-JLV-

[mschleppe]

I must not have worked my question completely right, but your on the right track. I do know how to synchronize the PXI6608 clock and the gps. What I do not know how to do is to print the time stamp with the data value to a relatively high accuracy. Using the system clock is fine, as it is synchronized with the gps pps and has a calculable offset to the gps time, but I just need a way to put the time of the analog event and the value of the analog reading into a usable file.

[unclebump]

Here is a functional global vi that I use to save cluster data into a file at close of my vi. When I start the vi again, I read this file so that I have the last data values.

[Ben]

If you do a search (of entire site) on "IRIG-B" you will find this link.

http://zone.ni.com/devzone/conceptd.nsf/webmain/2FAD3B4E67BA766186256B9C00679F75

The 6608 is smart enough to handle the GPS tioming feed.

Ben

[Ray.R]

Thanks Ben,

This is excellent information.

JLV

[Ben]

Do you remeber that opening round of "Shock and Awe" where a half dozen Tomahawks showed in downtown Bagdad at the same time?

They started out from different ships and subs in different oceans.

IRIG-B was used to get them all to show up at the same time (I believe).

Ben

[Ray.R]

Yes Ben,

And they have access to more accurate gps data..

However, civilian access is remarkably accurate as well.
I wish I remembered who made the gps device to provide timestamps. It was at 10MHz. The 27MHz was another timestamp for DVB-RCS.

Now Back to this thread...

So does the PXI 6608 provide "numeric" timestamp information? If so, then it is a simple matter of combining that information with the sampled data. So create an array of data & timestamps and write them to a speadsheet or text file.

I will look for examples of storing data & timestamps, and post them. Unfortunately "duty calls". I have to run back to the lab.

-JLV-

[MF]

Hello,

I saw your post and have a similar set-up. I have a PXI chassis with a PXI-6608 card in it and a Symmetricom GPS card that I am connecting the DC level signal from. I cannot however read the GPS time.  Can you tell me which pins specifically you connect the GPS DC Level signal to on the 6608 card and whatever other connections are needed. I am using DAQmx drivers in LabVIEW.
Thank-you.

Michael Froehlich

[Ryan_V[DE]]

Michael,

Assuming that "GPS DC Level" is the signal on which you want to count events, you would wire this to the sources of one of your counters and tie the gate of that counter high. On your card, for counter 0, the source is pin 2 and the gate is pin 3.

Hope this helps,
Ryan V.
National Instruments

Message Edited by Ryan V on 07-15-2005 05:34 PM