Synchronizing signal acquisition on PXI-4472 and PXI-6115 cards

[Asimo]

Hello,

I synchronize signal acquisition on several PXI-4472 devices simultaneously. I start the acquisition with an external trigger analog source. Now I'd like to add a PXI-6115 device to this synchronized acquisition, this one at a higher frequency, let's say 5 times larger. I have installed NI-DAQmx 7.4 drivers and it seems to work properly.

Any idea how I could do it? Ideally, the trigger source signal would be directed to a channel on a 4472 card, which is located in slot 2. Does the highest rate frequency card still has to be the master with NI-DAQmx 7.4?

[e-Cash-Money]

Hello Asimo,

If you want to have the 6115 use the same trigger as the 4472 devices, you should be able to route the trigger signal to the PXI backplane on one of the trigger lines, then use this the source of the trigger for the 6115. Because you want the 6115 sampling at a different frequency, I would reccomend using the onboard clock for sampling. This means that you don't have to worry about master/slave settings with the extra board.

Take care,
E.Lee

[reddog]

If you don't want the clocks to drift over time, the only thing I can think of is to use the sample timebase from the 4472 in slot 2 as the sample timebase of the 6115 and to use the start trigger from the 4472 as the start trigger of the 6115. Since the 4472 either oversamples at 64 or 128 times the sample rate, the 6115 will at most be able to sample 32 or 64 times as fast as the 4472 using this scheme (the 32 and 64 are due to the fact that the 6115 has to divide down its sample timebase by at least 2 to create its sample clock). Since the 6115 is using the timebase from the 4472 (max timebase of the 4472 is ~13 MHz), you won't get as good of sample clock resolution you would get if using the onboard 20 MHz timebase, but you should be in the ballpark. Also, this scheme will work fine if you stick to analog triggers on the 4472. If you start using a digital trigger, you'll have to deal with the filter delay on the 4472. There are ways to compensate for this, but I won't go into the details unless you plan on using digital triggers in the future. I've attached an example and picture that will hopefully illustrate what I've described. If you have any questions, let me know. Good luck!

[Asimo]

Hello guys,

Thank you very much, I really appreciate your help. Sadly I don't know a lot about all this so, E. Lee, I'm not sure to understand everything you explained.

Reddog, I modified your VI for my system but no matter what I do (modifying the sample rates, setting the number of samples to read to -1...) I get -200284 error. After the reading I did on the subject I suspect the problem is not related to the sample rate, so it must have something to do with the clock?

I join a bitmap of my block diagram. If anyone could help me with this problem, I would be very grateful.

[reddog]

Asimo,

In the 6115 task, you need to change the SampClk.Timebase.Src timing property from /PXI1Slot2/ai/SampleClock to /PXI1Slot2/ai/SampleClockTimebase. Give this a try and let me know how things work out.

[Asimo]

Hi reddog,

I tried to do what you suggest but now I can't acquire signal at all. I get a 89136 error (see bitmap), telling that the route cannot be satisfied because the hardware does not support it. I guess it has something with the configuration of my cards. I'll look this way. If you have any idea, please tell me.

Thank you very much.

[reddog]

I was finally able to track down some hardware and was able to get the attached example to work. For some reason, implicitly pulling the sample clock timebase from the 4472 isn't working. Instead, I had to explicitly export the sample clock timebase on the 4472 across the PXI-Star trigger line to the appropriate PXI slot. Then the 6115 could import the timebase by using the PXI-Star trigger line as it's sample clock timebase. You'll notice this is done in the revised example through the DAQmx Export Signal property node.

Finally, some explanation on the topology of the PXI-Star trigger lines may be in order. The star trigger lines are equal length point to point connections connecting slot 2 to slots 3 - 15. This enables the device in slot 2 to selectively fan out a trigger or clock to any of the devices in slots 3 - 15. It is across these trigger lines that the 4472 is sending the sample clock timebase to the 6115. However, the name of the trigger line to use in the terminal control may not be intuitive. The trigger line that connects slots 2 and 3 is PXI_Star 0. The line that connects slots 2 and 4 is PXI_Star 1, and so on up to PXI_Star 12 on an 18 slot chassis. In the current example, I am connecting a 4472 in slot 2 with a 6115 in slot 4 by using PXI_Star 1. You may need to change this depending on the placement of your boards. Also, if you plan to synchronize more the one 6115 with the 4472, you will need to export the sample clock timebase on each dedicated star trigger line. This can be done through the DAQmx Export Signal Property Node by using a comma separated list of terminal names instead of just a single terminal name.

Hopefully this will be the one that gets things working for you. Out of curiousity, what type of application are you trying to solve by synchronizing these two boards?

[Asimo]

Hi Reddog,

You made my week. I integrated what you suggest into my vi and it's working! Thank you so, so much. I used to acquire signals on two separate PXIs simultaneously. Of course there was no synchronization. Not very effective and smart.

To answer your question, I work in a university research center specialized in high performance machining. We often have to monitor dynamic signals of several different natures during the cutting process: cutting forces and torque, acoustic emissions (on a 6115 card, by the way), vibrations, power required by the spindle holding the tool... One typical application is to predict tool failure, but the possibilities are endless.

Again, I really appreciate your help with this.

[AndrejK]

Hi

Could be the same solution used for sync of PXI-4472 and PXI-4220 cards?  If not, what is the correct solution?

(I tryed the solution with PXI-4220 Simulated device and I get the Error No. 89125)

Message Edited by AndrejK on 05-25-2006 09:23 AM

[Daniel_L]

Hi AndrejK,

The PXI 4220 is most similar to E Series devices in terms of timing and synchronization. Like the E Series boards, the 4220 uses a DAQ-STC (systems timing controller) chip for time-related functions. The NI PXI-4220 uses the PXI trigger bus to easily synchronize several measurement functions between multiple NI PXI-4220 devices to a common trigger or timing event. The PXI trigger bus is connected through the rear signal connector to the PXI chassis backplane. The DAQ-STC provides a flexible interface for connecting timing signals to other devices or external circuitry. The NI PXI-4220 can also use the PXI trigger bus to interconnect timing signals between PXI devices and the programmable function input (PFI) pin on the front SMB connector to connect the device to external circuitry. These connections are designed to enable the device to both control and be controlled by other devices and circuits. The DAQ-STC has internal timing signals you can control by an external source. These timing signals can also be controlled by signals internally generated by the DAQ-STC, and these selections are software configurable.

Therefore, based on this KB "Can I Synchronize a DSA Device with Other NI Hardware?" [link removed; article no longer available] , I would consider option 2, "NI-PXI 4472 with Analog I/O on PXI E Series". I'm fairly certain that this would be the best option. 

Regards,

Daniel L.

Applications Engineer

National Instruments

Message Edited by Daniel L on 05-26-2006 11:19 AM