Syncronize NI Scope 5122 with DAQ 6124

[Alex PTB]

Hi all,

I try to syncronize the PXIe 5122 with a PXIe 6124. I´m using the PXI Clock 10MHz of the PXI Chassi 1042 as Reference clock for both, but I still get a phase shift between the two digitizers.

I attached a sample vi.

Does somebody know how to solve this problem?

Is it possible to syncronize such a measurement, maybe by tieing one sampleclock to the other or is there a simpler way to do that?

I want to realize a continous measurement using both cards. At the end with different sample rates....

Any help would be great!

Thanks

Alex

[dansch]

Hi Alex,

It is harder to synchronize a digitizer when acquiring continuously because you need a common starting point on both devices in order for them to start sampling at the same time, and the digitizer doesn't use a normal reference trigger by default when performing continuous sampling. However, you can use an Acquisition Arm (Start) Trigger on a digitizer to tell it when to begin sampling pre-trigger samples (which, for a continuous acquisition, all samples are considered pre-trigger as no reference trigger ever gets sent).

I think the best way to do this, in addition to sharing the same reference clock as you have already done, is to export the start event/trigger from the DAQmx card and use that as the source of the Start Trigger on the digitizer. To configure a Start Trigger on the digitizer, you will need to use an niScope property node found under Synchronization » Start Trigger (Acq. Arm) » Source. Since the digitizer will be waiting for the exported signal from the DAQ card, you will want to initiate the digitizer first, then start the DAQ card. To make the operations sequential, you will probably need to route your error wires between the niScope and DAQmx functions so that proper timing and data flow is maintained. Also, keep in mind that because the digitizer is sampling at a much higher rate than the DAQ card, there may be some small differences between the phase due to the delay of exporting the start trigger.

Hope this helps,

[dansch]

Hi Alex,

Actually, I completely forgot to mention the fact that you could probably do this with an easier method, which would be to export a divided down sample clock from the 5122 to use as the sample clock on the S Series card. This will only work if you can sample on the S Series at a rate which is an integer divide down from the 5122's 100 MS/s rate. In this case though, you would start the DAQmx device first, and then as soon as the digitizer initiates its acquisition, the sample clock will be sent to the S Series device and it will also start sampling. Of course, as before, there will be some small difference between the cards due to the propagation delay of exporting the sample clock to the other card (the digitizer will already be sampling while the clock gets sent to the DAQ device).

Regards, 

[Alex PTB]

Hi Daniel,

thank you very much for your help!

Luky me. I just got  an another scope card 5105. So I use this one instead of the 6124. And I don´t have the synchronize problem anymore 🙂

Anyways thanks you.

It´s sometimes realy hard to figure out what the digitizers are actually doing....

But I have an another question about streaming.

If I use the High Speed digitzier streaming example, I´m not able to stream at 100MS/s using the 5122. The max is 50MS/s and 33MS/s for two channels.

First I thought it depends on the storage device, so I modified the vi to write to the computer onboard memory via Queue instead of writing to hard disk.

The onboard memory should be fast enough. But it is the same problem. I still get 200004. Data is overwritten in the card-onboard memory.

Unfortunatly I just have 8MB card memory...

Is it not possible to use the 100MS/s in streaming mode?

Thanks!

Cheers

Alex

[DFGray]

What variety of 5122 do you have (PCI, PXI, ...)?  If memory serves the PCI bus has an aggregate bandwidth of 133MB/s (all devices on this bus share this bandwidth).  If you have a PCI or PXI flavor, this is one possible reason.  The PXIe version should not have this problem (400MB/s transfer bandwidth).

Remember that each sample is two bytes, so your bandwidth is 100MB/s at 50MS/s.

[Alex PTB]

I´m using the PXIe 5122. Labview 2010 on Win7 32Bit.

So it should be possible. I can fetch some samples but after a few milliseconds I get an error (200004).

I think I can´t fetch fast enough. If I increase the number of fetches the error occures even earlier. I´m aware of the bandwith problem. But the Comp. RAM should be fast enough to cover the 100MB/s.

[DFGray]

Error 200004 is an input overload error:

---

Input termination resistor for one or more channels was overloaded. Your data might be invalid.

---

You may have an overvoltage input.

You are correct that the PXIe-5122 should be able to stream both channels at full speed to RAM.  Could you post your code on how you are doing this?  If you are building arrays or plotting data as you do this, it could slow you down enough to fail.  Some of this is dependent on your computer hardware architecture, as well, although it should not be an issue with most relatively modern computers.