High-Speed Digitizers

cancel
Showing results for 
Search instead for 
Did you mean: 

PXIe-5122 + PXI-6552 TClk Synchronization

Hello,

 

I'm trying to implement a continuous, and synchronized, acquisition on a system comprising a PXIe-5122 and a PXI-6552.

I'm using TClk to get these 2 devices synchronized as tightly as possible.

The idea is to acquire an analog channel (CH0 on my PXIe-5122) as well as a 16 bits digital bus (Dynamic Channel 0:15 on the 6552) simultaneously and to log data to my HDD-8164 RAID array disk for post processing and analysis.

I could find the simple (simplistic?) "mixedSignalScopeUsingTclk" example on NI's website and tried to adapt this VI to perform continous acquisition.

Unfortunately, the way continuous acquisition is performed on the 5122 (never actually triggers - use pre-triggers samples only) doesn't seem compatible with the way homogeneous triggers work in TClk...

I'm reviewing the TClk documentation and trying different combinations of settings but have not been very successful so far.

Did anyone ever tried to implement such a mixed-signal synchronized acquisition ? Any tips that would allow me get this exemple up and running?

Thanks in advance for your help !

Best regards

Gerald Albertini

0 Kudos
Message 1 of 3
(4,932 Views)

I have used the Tclock to "mix" the 5421/5122 and also 5122/5442 in CVI. The required code is farily short but it took a couple of tries to get it done.

 

A few questions first,

 

What language? Looks like Labview based on the VI reference.

 

You want both cards to trigger simultaneously and do a single collection, correct? (No Multirecords etc.)

 

You want to continuously collect to RAID for both cards (cyclical buffer onboard to RAM buffer to RAID or similar), correct?

What sampling rate are you attempting (100 MSample / s )?

In this case can you verify by running without TClock that the throughput is available.

 

Embedded or remote (MXI) controlled?

 

I can probably ferret out the CVI code and then you can relate the corresponding VI to the code.

 

0 Kudos
Message 2 of 3
(4,863 Views)

I just wanted to include some of the comments that I wrote to Gerald via email, for further reference of anyone reading this post.

 

This is slightly confusing since HSDIO uses a different method of continuous acquisition than the NI-SCOPE instruments. Since the digitizer requires the user to configure a reference trigger that never gets sent, it must be configured as the master session for the reference trigger. This means that the scope must configure a software trigger and no reference trigger should be configured for the HSDIO device. Thus, neither board will actually ever receive a reference trigger. As far as the start trigger goes, either device may be the master for the start triggger session (meaning that you can configure the start trigger source for one board and not configure anything on the other). In general, T-Clk will ensure that the boards begin at the same time, so it actually only matters to configure a start trigger if you would like the boards to start synchronous to some external event. In terms of fetching the data, the method used in the niScope EX Fetch Forever example should be used, meaning that you will have to fetch relative to the Read pointer in a loop. The HSDIO and SCOPE fetches can each be performed independently just as they are done in their respective continuous acquisition examples.

 

Regards,

Daniel S.
National Instruments
Message 3 of 3
(4,835 Views)