Synchronizing FPGA modules 5674 and 7976R

[Dorino]

Hello there,

I had originally posted this in the Labview board, but it occurred to me that PXI might be a better place.

My setup is the following:

PXIe-1075 Chassis:

     --PXIe 7976R FPGA module + AT1212 pulse generator FAM - Operating at 156.25MHz (IO Module\IoModClipClock0)

     --PXIe 5764 FlexRIO Digitizer module (KU060) - Operating at 125MHz (IO Socket\Data Clock)

     --PXIe 6363 PXI Multifunction IO module

I am attempting to generate pulse patterns with the 7976R, which triggers the 5764 for acquisition of the waveform. Both work fine on their own: I am able to trigger either with DIO from the 6363. For the 5764, I am using the 'Getting Started Example 5764' from labview examples.

The problem happens when I try and trigger the 5764 with the 7976R. Operation is as follows: The 6363 sends a digital output to the 7976R on PXI_Trig0 -> 7976R sends a trigger across PXI_Trig1 to the 5764 to start acquisition.

The problem is that I do not see a waveform every time I send a trigger. Is the reason because the FPGA modules are running at different clock speeds and need to be synchronized? If so, which method of synchronization is the most suited for this setup? I have found the following link Synchronization Explained , that explains a few synchronization methods using DAQmx, but other pages have mentioned using NI-TClk.

After further debugging, I've noticed that only sending a trigger to the 5764 is triggering the 7976R, which shouldn't happen because the PXI trigger lines are different. This only seems to happen when I perform routing of PXI_Trig1 to Destination1 on the 5764.

I am very new at NI FPGA modules, so any advice in the right direction would be greatly appreciated.

For those curious, I have attached the FPGA vi for both targets, as well as the host vis and the vi to control the DIO on 6363. Please let me know if I am missing something.

Thank you,

Dorino

[MingSheng]

Hi Dorino,

I met exactly the same problem when I used 5764 as you did then, have you solved it? It would be greatly appreciated if you can share some experiences.

Thank you,

Sheng Ming 

[Dorino]

Hi Sheng,

It's been awhile since i've worked on this project. What precisely are you having troubles with? Is one card not triggering the other? After the post on this thread, I found some other issues I had to fix, so i'm curious what exactly might be your problem.

Sincerely,

Dorino

[MingSheng]

Hi Dorino, thanks for your reply. I'm working with PXIe 5764 digitizer, 5745 AWG and 6674 clock card now. All these cards are connected to a chassis. Then I use 5745 to generate a square wave as trigger and an arbitrary waveform for measurement test. The trigger signal is routed through the chassis. When I use 5764 to acquire the arbitrary waveform, the trigger mode can be 3 cases: immediate, digital and analog. Since I use square as trigger signal, I set the trigger mode as digital edge trigger, then 5764 returns an error. But when I change the trigger mode to digital level trigger, 5764 acquires the waveform, but repeating acquision gets different waveform. I'm not sure about the difference between these two trigger modes. Is there a setup that I can acquire a definite output as I send in with 5745?

[Dorino]

Hi Sheng,

I think I understand your problem. The difference between the digital edge and level trigger is where the receiving hardware looks for a change. The edge would be from whenever the level of the trigger line drops/rises and the level would trigger after a certain time that the trigger line is at a certain voltage. I apologize if I stated the obvious. Functionally, I dont know why its triggering on one and not the other. Perhaps a timing issue with the edges?

I dont have a 5745 so I'm not sure if it will work, but have you tried sending a trigger on the PXI lines (the PXI chassis backplane triggers, usually 0-7) from the 5745, waiting x clock cycles, then producing the waveform? The 5764 would then receive the trigger and start acquiring.

Let me know if this doesnt make sense.

Sincerely,

Dorino