NI-5133 Measurement slow

[Wolley]

For further defense of my assertion:

From user DFGray in 2005, even:

1. You can fetch the same waveform multiple times from the digitizer. It will stay in the card's memory until you initiate another acquisition.

Therefore for each data-set, I need a Start Acquisition and Abort Acquisition for each set of data I want, ie one per inner loop.

[Joshua-B]

Hi Wolley,

Please look at the example in LabVIEW called niScope EX Multi Record Fetch More Than Available Memory.vi.  You can get to the example finder in LabVIEW by going to Help >> Find Examples.  Once you are in the example finder navigate to Hardware Input and Output >> Modular Instruments >> NI-SCOPE (High-Speed Digitizers) >> Continuous Acquisition.  Here you will find the example I mentioned.  

I believe you have configured your task to only acquire one record which is probably why it isn't recording more records unless you abort and re-initiate the acquisition.  I recommend modifying the VI I posted to read the number of records that you want to read instead of just one.  The default value for the number of records input on the Configure Horizontal Timing VI is one.

[Wolley]

I don't have the VI in front of me, but in several different configurations as what I've been working on has varied, I've tried increasing the number of records - any time I do so, it returns an error if the number is greater than 1. 

Wolley

[Joshua-B]

Hi Wolley,

It looks like you are correct.  I just got access to a USB-5133 and tried to configure it for a mult-record acquistion.  I received the error you are mentioning.  The 5132/5133 is apparently the only scope/digitizer we sell that doesn't have multi-record acquisition capabilities.  This can be seen in the section of the High-Speed Digitizers Help file titled Features Supported by SMC-Based Devices and USB Devices under the Acquisition table.  

Since you can only acquire one record at time, you will have to abort and then re-initiate the acquisition every time you want to acquire a new record around a reference trigger. 

If you want to speed up your acquisition, you may want to consider continuously acquiring data and then post-processing the data to pull out the samples you want rather than trying to only pull out pieces or data around a reference trigger initially.  The example niScope EX Fetch Forever.vi is a good place to start if you want to pursue this route.

[Wolley]

I unfortunately (again) am away from my access to Labview for now, but in the mean-time - does that example include advice in targetting the data? I did try an instance where instead of 128 samples in the inner loop it was 4 collections of a duration that would include 25% each of the signals coming in - but in trying to process it, I could only readily access the first signal which the acquisition was triggered from because the following signals didn't have a precise enough timing to assume a time-spacing from signal-to-signal. This wouldn't be so problematic were it not for the fact that the signal duration is so small compared to the frequency - a 20ns pulse roughly each millisecond doesn't amount to a great ratio of data to noise, you know?

If not, do you maybe have any advice? In all actuality I mostly just need a way to characterize the voltage of the pulses, but considering that signal-to-noise ratio the data-measurement itself isn't worth much. I considered doing some sort of averaging of the measured voltages to a higher power in order to try and emphasize the pulses over the noise - do you foresee this being a worthwhile pursuit, or maybe have an alternative to try?

Thank you very much,

Wolley

[Joshua-B · Accepted solution]

Hi Wolley,

Without being able to do multirecord acquisitions, I think that a continuous acquisition is probably the best way to go.  The example I mentioned doesn't have anything built in that targets specific sections of the data.  You would need to write you own algorithm for that after the data has been collected, but I do think that the Threshold Detector VI in LabVIEW would probably go a long way toward helping you narrow down the areas of interest within your data.