Shared clokck among DSAs and a MIO

[Brad_K]

Hi golubovski,

I'll leave your DSA questions to Jason, but to address question #2, your DAQmx is not crippled. By default, DAQmx property nodes in LabVIEW filter out properties that are not supported by the boards you have installed. To show all of the properties, right click on the property node and select "Select Filter..." from the context menu, then check "Show All Attributes".

Brad

[golubovski]

Thank you Brad, you are right.

This raises another issue for me regarding the clock I may use from PXI-4472 to feed PXI-6713 to avoid division? The chassis is PXI-1006 (3 in 1). This should really be documented considering the devices are so much different. So Jason, I'd be very gratefull if you can tell me which clock from 4472 can be feed directly to 6713 (which options from the Timing Node) without the need of dividing it (like I had to do in the traditional NI-DAQ).

Thanks in advance,

[RT4CY]

Hello again golubovski,

How are you today?  I see you have a couple more questions regarding the synchronization of the DSA and AO devices.  It looks like Brad answered your second question, so I'll answer your other questions here.

1.a.

How (where on the diagram) is it defined that DSA in slot #2 is sampling (clock) master for the rest of the DSAs (slots #3 to #10) - I can see MIO (slot #11) receives clock and is triggered by the AI master? Is it implicit by its "reserved" ST location in slot #2?

1.b.

 Who generates the "PXI_Clk 10" (I beleive a 10 MHz clock)? If it is DSA master (slot #2) which trigger line should I direct LEFT-to-RIGH, or is it taken care of automatically by the NI-DAQmx? I do not see explicite definition in the AI task!?

So Jason, I'd be very gratefull if you can tell me which clock from 4472 can be feed directly to 6713 (which options from the Timing Node) without the need of dividing it (like I had to do in the traditional NI-DAQ).

The 10 MHz clock used to synchronize the devices is built into the PXI backplane.  When you set the sampling rate, DAQmx will divide down the PXI_CLK10 reference clock for you.  For more info, check out the Multi-Device Synchronization section of the following devzone article.

M Series Synchronization with LabVIEW and NI-DAQmx

The PXI-4472 in slot 2 is generating the start trigger, so it can still be considered the master DSA device.  However, the clock being used by the devices is coming from the PXI chassis backplane.  I took a look at your program, and everything seems to be good.  Let me know if you have additional questions.

[golubovski]

Rod,

Glad to hear from you again! My problem is with the "RefClk.Src" attribute which is not available (is filtered) for the PXI-4472 and PXI-6713...?!? I have copied that node from the example you gave me previously but normally I cannot access it and the "PXI-Clk10" feed for it...?!? So if RefClk.Src is unavailable for my devices, which clock do I use with the nodes?

Anyway, is it safe to formulate my findings the following way?:

1.

By gropuping all 9 DSAs in a single task they all inherit the (I hope) 10MHz sampling or ADC clock from the backplane.

2.

Implicitely, all PXI DSAs from  slots #3 to #10 trigger themselves from the DSA in slot #2?

3.

What is the explicit use of the 3 nodes (2 in the AI task and 1 in the AO task)? If the 10MHz clock is available to all PXI slots and the MIO waits upon trigger from the DSA in slot #2 why is the wiring between the 2nd (AI) and 3rd (AO) nodeneeded? I have dificulties distinguishing all the options of the Timing Node - if the 1st makes the DSAs to listen for the 10MHx clock, the 2nd delivers it to the MIO allthough the MIO can listen fo it itself...?!?

4.

In order to allow the StartTrigger to propagate from slot #2 (bus 1 in PXI-1006) to slot#11 (bus 3 in PXI-1006) do I leave all 8 trigger lines to be "Dynamic" or do I need to specify one "Away from slot #2"?

Sorry for bothering, but I am really confused.

Sincerely,

[RT4CY]

Hi golubovski,

Thank you for the reply, and for pointing out that the RefClk.Src attribute is not available for the PXI-4472.  You have discovered something that I overlooked, that the PXI-4472 does not support using the PXI 10 MHz reference clock.  This is possible with the PXI-449x and PXI-446x devices, but not the PXI-4472. 

So in this case, you will do the following:

1. Create an AI task for the Master DSA device (in slot 2), that will output a clock and trigger to be used by the slave devices

2. Group together the slave DSA devices in a second AI task and have them receive the clock and trigger (it will also receive a Sync pulse)

3. Create an AO task for the PXI-6713 device that will receive the clock and trigger

We have a great example which you can modify that ships with LabVIEW.  Go to Help»Find Examples to launch the NI Example Finder.  Browse by task to Hardware Input and Output»DAQmx»Synchronization»Multi-Device folder.  The very first example is what you need (Multi-Device Sync-AI and AO-Shared Timebase & Trig-DSA.vi).

Case 1 uses the Reference clock method that I've been discussing.  You will use Case 2, which exports the clock and sync pulse from the master DSA.  You can remove the second AO task, since you only have one PXI-6713. 

Sorry about the confusion, and I hope this helps!

[golubovski]

Rod,

I have attached my final version.

1.

Is this OK with PXI-4472s and a PXI6713?

2.

This should really be delivered as formal tutorial to the customers? Why does the sampling synchronization need the SyncPulse in addition to the SampleClock? Is the SampleClock an ADC timebase or the AI ScanClock?

3.

DSAs can sample up to 102.4 kS/s - can the MIO follow this way?

4.

Do I leave the PXI trigger lines all to be "Dynamic" for the NI-DAQmx to use, or need I configure routing between the 3 buses within the PXI-1006 backplane?

Thanks again,

[RT4CY]

Hi golubovski

1. Is this OK with PXI-4472s and a PXI6713?

From what I can tell, everything looks good in your program.  These are the proper signals to be using with the PXI-4472.  This is where I was mixed up at first, because other DSA devices (such as the PXI-449x and 446x) can indeed use the 10 MHz reference clock from the PXI backplane.

2. This should really be delivered as formal tutorial to the customers? Why does the sampling synchronization need the SyncPulse in addition to the SampleClock? Is the SampleClock an ADC timebase or the AI ScanClock?
4. Do I leave the PXI trigger lines all to be "Dynamic" for the NI-DAQmx to use, or need I configure routing between the 3 buses within the PXI-1006 backplane?

The Sync Pulse simultaneously resets the internal clock dividers and converters on each DSA device in the system.  You can read more in the NI-DAQmx Help.  Go to Start»Programs»National Instruments»NI-DAQ»NI-DAQmx Help.  Search for "SyncPulse", and you will find the answer to question number 4 as well.  Essentially, the SyncPulse is routed to Trig5.  The timebase from the master is routed to PXIStar.  DAQmx is smart enough to do this for you.  I'm sorry for the confusion from one of my previous posts.  As you can see from this KB, the PXI-4472 can indeed route signals to the Star line.  This would not be necessary for the PXI-449x and 446x devices, since they can PLL to the 10MHz reference clock.

3. DSAs can sample up to 102.4 kS/s - can the MIO follow this way?

The PXI-6713 can generate as fast as 1 MHz, so this should not be a problem.  As a general rule, usually you will use the slowest device as the master.

Let me know if you have any more questions/concerns.  I'm here to help.

[golubovski]

Rod,

Tank you for the support and your time. I will have the opportunity to test this. You have been more than helpful. If I understood correctly, I am actually sending clock via the StarTrigger, the SyncPulse via Trig#5 and StartTrigger via one of the routed trigger lines. Another note from me - I think PXI-6713 cannot be a ST master so I cannot put it in slot #2.

Till next time,

[golubovski]

Rod,

I just gained access to the DAQ setup - so far it was doing tests and I could not interfere. There is a problem with the timing nodes for the AO task (PXI-6713). When I run the last version of the VI, NI-DAQmx reports error -200452. I removed the SyncPulse.Src than error - 200132 is reported implicating the NEEDED clock division from the 4472 for the 6713. As I sad at the beggining of this thread NI-DAQmx does not divide theat synchronization clock, it has to be done explicitely (like with the Traditional NI-DAQ).

Now my problem is I do not understand (distinguish) the different TiminNode proerties allthough I am perfectly aware what I must do. Can you help me wire the divisor properties? Can SampleClk.Timebase.Src be used to be divided down as clock for the signl generation of the 6713? How to accomplish the division - which properties to connect with what?

Sorry for the inconveniance and thanks in advance,