digital data regeneration with retrigerring

[nirtzi1]

Hi,

first of all:

we are using LabVIEW 8.0 with PCI6534.

my system include external clock of 4.2MHZ and my trigger is 17 HZ.

we try to transmit digital data with retriggering using while loop.

the trigger is config outside of the loop.

we expect to get the data valid when the trigger rising and not valid when the trigger falling.

but the result is not good...

most of the result are ok but every few bursts of data there is no valid data for all of the burst period (14ms).

the file of the loop is attached.

we are waiting for any suggestion to solve the problem or to make it better.

thanks

[Abhinav T]

Hi nirtzi1,

I took your vi and did some testing on it. The best way to retrigger a digital output task is with a counter, since your board does not have any counters, the method you are using of software retriggering is the only options. The reason why you see some lapses in missed trigger transitions is because the DAQ card takes some time to reset when you stop and re start the task. The vi attached to this post is a slightly better was to accomplish what you were trying to do. It checks to see if the task is done outputting your waveform and then resets the task.

I hope this helps!

Abhinav T
Applications Engineer
National Instruments

[nirtzi1]

Hi,

Thank u for trying to help but it still not working.

We tried to transmit only one burst of data and we still can observe the same problem every few slots.

We attached file which can give a view of the problem. (Yellow line is data, purple  line is the trigger and we expect that the data will apear

in it's period of time only - 14ms).

if you have any answer or idea what we can do please let me know.

p.s

we use labview 8.0 with PCI6534, the trigger is 17HZ, external clock of 4.2 MHZ.

Best regards

Nir

[nirtzi1]

Hi again abhinav,

this time i attached a good burst too...

(in addition to my previous mail :

Hi,

Thank u for trying to help but it still not working.

We tried to transmit only one burst of data and we still can observe the same problem every few slots.

We attached file which can give a view of the problem. (Yellow line is data, purple  line is the trigger and we expect that the data will apear

in it's period of time only - 14ms).

if you have any answer or idea what we can do please let me know.

p.s

we use labview 8.0 with PCI6534, the trigger is 17HZ, external clock of 4.2 MHZ.

Best regards

Nir)

[Abhinav T]

Hi Nir,

Your issue here is that the DAQ card is missing the trigger while it is resetting (with the DAQmx Stop and Start). This is the only way to make your task retriggerable without using counters.

In order to improve the accuracy of your output, the only way to circumvent this issue is with a counter - here is a Knowledge Base article that shows you how to use counters to attain better output.

Abhinav T.
Applications Engineer
National Instruments

[nirtzi1]

Hi,

we use PCI 6534 with two groups: group A and group B simultaneously,

we want to improve our accuracy by counter.

1. will the counter solve the problem if our trigger is 17Hz?

2. which card do you recommend us to use which include counters and two groups like PCI 6534?

Best regards

Nir

[Abhinav T]

Hi Nir,

I wanted to know some more information to solve this issue:

What are the specs of your computer?
How long is your data burst?
Is it more than the period of the trigger (1/17s) because the program will have to wait until its done transmitting the first trigger's signal it may miss the second trigger.

If you want a counter I/O board, the PCI-6602 is an 8 counter board. You will also need a RTSI cable to route the counter output to the sample clock on your.

Abhinav

[nirtzi1]

Hi,

My computer specs. are,

IBM PC, 1.4GHz , 512 MB & Windows XP

---------------------------------------------------------

My trigger signal is a digital pulse with duty cycle of 1/4 and rate of ~ 17 Hz.

(My trigger is on ("1") for ~14 msec, and then off ("0") for ~ 42 msec.)

While the trigger is on, I would like to transmit data, (~14 msec) in rate of 4.2 MHz using external clock (4.2 MHz).

--------------------------------------------------------------------------------------------------------------------------------------------------

This application works quite well, except that ~ 20% of the transmitted bursts are missing.

(You were saying, that this effect is caused due to "Software Re Triggering")

Does the PCI - 6602 should be added to our PCI - 6534?

How the PCI - 6602 interface to the PCI 6354? (Can we get a simple schematics)

If this approach is the prefered solution to our application, can we get a *.vi file for it?

Best Regards,

Nir

[Abhinav T]

Nir,

Yes you will need to add the PCI-6602 to your PCI slots. You will also need to connect a RTSI cable between your 6602 and 6534, this synchronizes the two cards together within hardware. There are several examples of how you can use the 6602 to get better performance out of your retriggerable task. The Knowledge Base article I placed in an earlier post, also linked here. There is an NI-Daqmx Shipping Example that shows how to retrigger an Analog Input task using counters (It is found in the NI Example Finder (Help » Find Examples) under Hardware Input and Output » DAQmx » Synchronization » Multi-function). Another resource you may want to look through is this Developer Zone article on RTSI.

Abhinav T.

[nirtzi1]

Hi,

We have the PCI 6602 with PCI 6534 and we try to regenerate date with triggering.

Whave a pulse clock (56ms with 0.25 duty cycle) out of the PCI 6602 which is at rate of 4.2MHz.

We are trying to sync the data with the clock but it dosn't work.

we try now to make the sample clock as a "finite ample" and the number of samples is the same as the clk.

It still doesn't work. we get an error :

*************************************************************************

"Possible reason(s):

Attempted to write a sample beyond the final sample generated. The generation has stopped, therefore the sample specified by the combination of position and offset will never be available.

Specify a position and offset which selects a sample up to, but not beyond, the final sample generated. The final sample generated can be determined by querying the total samples generated after a generation has stopped.

Attempted to Write Sample: 128008

Property: RelativeTo
Corresponding Value: Current Write Position

Property: Offset
Corresponding Value:

Task Name: _unnamedTask<139>"

******************************************************************************************************

WHAT IS IT??? HOW CAN WE SOLVE IT???

transmission loop is attached.

best regards

Nir Berkovich