6561 TClk Synchroization Problem

[gt3000]

Hi bmann2000,

Looking further into the errors, I approached this from the side of NI-TClk.  If the error indeed is that NI-TClk needs a the master session Start Trigger to be set explicitly, there is a Property Node that can be used.  One of the existing references that is being exported can be input into the property node.  This is seen in the image below:

The Property node is found in the palette as shown below:

If this Property Node generates an error for you, we can also export the Start Trigger of one card onto the PXI Trigger lines on the PXI-1045 backplane.  This code below should be added only for a single reference at some point before the "niHSDIO Initiate VI" from the previous post.  Previously, I have expected that this NI-TClk VIs would assign the 10MHz Reference trigger from the backplane, and always choose a Start Trigger Master as well, which I why I have not mentioned this before.

As for additional documenation, I am searching for some whitepapers that can be followed.  For the moment, the examples in Help->NI Example Finder  can be used as a reference.  I am still looking for additional sources to speed up the responses to your questions.

[gt3000]

One more thing,

Just for clarification, can you verify that example Multi-Device Dynamic Acquisition (TClk).vi works correctly without errors?

The recent issues have appeared because of the introduction of having to trigger only when a specific pattern is matched, both for the start of a new record, and for subsequent records.

Regards, 

[bmann2000]

Hi George,

Yes, Multi-Device Dynamic Acquisition (TClk).vi does works correctly without errors, code attached with results set as default.

[bmann2000]

Adding the StartTrigMasterSession Property Node results in error -250033 "You cannot perform this action while the session is in the running state".

[bmann2000]

If you add niHSDIO Export Signal, the previous -250033 error persists.  If you then remove Property Node StartTrigMasterSession you end up back at error -250037.

VIs with results set as default attached.

[gt3000]

Hello bmann2000,

In an effort to speed up this process, I am taking a step back and re-reading your requirements.  I am attempting to write my own VI that performs the task you have had in mind in this post.  

I willl be posting it up as soon as I have made substantial progress.  You can then copy some of concepts for your own project.

Regards, 

[gt3000]

Hello bmann2000,

 The approach to solving your issue has had to be changed.  Initially, we were debugging the code that was posted on the forum, which involved triggering when a certain pattern was matched.  It was concluded that the Advanced Trigger function was not suppored by NI-TClk, and the Start Trigger was recommended instead.

Re-reading the criteria and trying to develop my own application, I can now say that it is not possible to overcome the errors in the previous posts that were generated through NI-TClk.  When synchronizing multiple cards, NI-TClk could use a Start Trigger as a master trigger for all cards.  However, the specification of the HSDIO cards is that they will not produce an exportable Start Trigger when a Pattern Match has been detected.

To perform that task described in your original post, the channels must continuously stream to disk for the application duration of 1 hour.  The individual records could be synchronized with NI-TClk and then compared in post-processing to find where the patterns did not match. Triggers cannot be used in the way we thought in the example we have been debugging together.

Regards, 

[bmann2000]

Okay, are we talking about a TClk sync to start the acquisition, sampling all 72-channels continuously streaming to TDMS, then stripping out the zeros post-processing ?  2M x 6 = 12M samples per sec, it's well within the bandwidth capabilities of the PXI.  I'll create a VI to try it out.

[gt3000]

Hi bmann2000,

Please find the example that I mentioned I would make attached.  It actually follows the pattern you described previously, where the data is stored in a TDMS file.  I have not tested the VI on hardware and a few further modifications will be needed from yourself.  However, I was hoping to post this before you made your own.

1. The location of the TDMS file must be changed on the Block Diagram.
2. Writing to a TDMS file should be inspected to ensure that different cards are writing to different channels.
3. The stop logic with the 'greater than' in the Producer Loop should be checked.
4. The code should be tested first with two cards, and then with an increasing number of boards.

Please let me know if you have any questions.  I will continue working on this in the mean time to refine it.  The example should open in LabVIEW 2009 or greater.

Regards, 

[gt3000]

Hi bmann2000,

This is just an update to the VI from the previous post.  It just takes care of points 1) and 2) I mentioned earlier.  It should reduce the amount of coding you have to do for writing to file correctly so that focus can be on the actual synchronized HSDIO side. 

Please let me know how the application tests with the actual hardware.   

Regards,