Using NI 9467 GPS module

[Harss]

Hello,

My application requires me to timestamp the samples at the rate of microseconds (i.e. 1 MS/sec).

I am using NI cRIO 9030 and the gps module NI 9467 for this.

I am unable to find any example which would allow me to timestamp at microseconds.

I am using the analog modle NI 9223 for data acquisition.

Could you please guide me through the necessary steps to carry out this interfacing?

[AmitShachaf]

Hi,  We are sending signals from NI-1473R and sampling the I/Os in 40MHz.  Obliviously the I/Os are not changing that fast. What we are doing is to compress the signal. We are attaching 32bit timestamp for every number we send over the DMA. The timestampe is of delta T from previous measurement.  I can give you example code if it helps. I just want to point out that we are not sending the data in full rate.  We are creating timestamp and sending the data only when one of the I/Os is changing. I think that sending data in 1MS/sec continuously is very fast. I can send you example if it helps.  Amit 

[Harss]

Hello Amit,

Could you please share your example.

Thanks.

[Harss]

After acquiring signal at 1 MS/sec using NI 9223. The timekeeper logic  timestamps signal at 1 pps.(using NI 9467)

Is there a way in which both the VI's can be integrated to timestamp at microseconds.

Please help.

Thanks.

[Harss]

Hi,

I am facing another issue with the basic acquisition  I am using cRIO 9030 controller and NI 9223 . It seems the User controlled I/O sampling does sample at 1 MS/sec . However, it does not acquire data continously and gets timed out frequently. I have referred to a similar post

http://forums.ni.com/t5/Real-Time-Measurement-and/NI-9223-at-1Mhz-on-a-cRIO-9024 /td-p/2892950/page/...

However, my application does not require me to plot a graph so can this reduce additional overhead of the host and allow me to acquire the signal at 1 MS/sec continously and log the data on the crio.If I use the 'Producer' 'Consumer' architecture.

Could you also please suggest if there is any other architecture to accomodate this.

Increasing the FIFO size also causes the labview to not work as the size to which the FIFO can be increased is not known and increasing it causes labview to not coompile the code

Can anyone pls. help !!

Thanks.

[Harss]

Hi,

Please find attached the basic signal acquisition VI.

I have tried to run this VI and the NI 9467 Data logging VI from the labview examples . on the same FPGA block diagram and the front panel.

The timestampig VI timestamps at 1 sec and the FPGA VI acquires the signal at 1 MS/sec. I am unable to link the two VI's to enable them to acquire as well as time stamp the signals at 1 microsecond interval.

Please help.

[Harss]

Hi,

I need some guidance on the timestamping, if the procedure I am using is correct because it seems the FPGA timekeeper is capable of timestamping at 1 sec . I am unable to understand how to timestamp the signal at 1 microsecond to keep up with the acquisition and later send the data to the front panel for display.

Could anyone please help me in timestamping at microseconds interval.

Thanks in advance.

Regards.

[JCC_(SK)]

Hi

In my project I used NI TimeSync FPGA Timekeeper

[Harss]

Hi,

Thanks a lot for your feedback . I will check the example and get back to you.

Thanks once again.

Regards.

[Harss]

Hi,

I have used NI 9223 in 'Úser controlled I/O Sampling' from labview examples section to enable it to acquire signal at 1 MS/sec. This VI needs a user input in terms of the number of samples to acquire which implies this needs to be a top VI because it needs user defined parameters. I am not sure how to exactly link this VI to the FPGA timekeeper VI because the time keeper VI would work on a PPS signal and would timestamp at an interval of 1 sec.

I need to link my timekeeper VI with my acquisition VI , keeping the acquisition VI as the topmost VI and enable the timestamping at microseconds interval. I have thought of one possible solution to this which could be timestamping at one second and counting the number of samples between this one second interval. The time difference between the two consecutive one second interval GPS timestamp could be subtracted and divided by the number of samples acquired between these one second interval and this could be used to timestamp each missing sample between this one second interval.

I am unable to find any means or resource about how to implement this. Could you please help me to understand if you have used any such kind of approach to timestamp at micoseconds interval because I am unable to implement this  or if anyone could suggest any other better alternative to this it would be really helpful.

Thanks in advance.

Regards.