Continuous multiple channel and multiple record fetching with the 5105

[Jennifer O]

Hi Jswalwell,

I understand you were in contact via phone with one of the Applications Engineers last week.  With your permission I will get your contact info from him.  I would like to discuss some potential solutions with you for your specific configuration that will be valid while we work fix the root of the issue.

Jennifer O.

Product Support Engineer - High Speed Digitizer

[jswalwell]

tried calling

[jswalwell]

So, when do you expect this to be fixed?  If I cannot use scope 3.4 and I cannot set enable records > memory, I have to run initiate.vi in a loop on intervals that do not fill up the onboard memory.  This slows down my continuous acquisition speed and I miss data every time initiate.vi runs: each channel acquires 200 points x 2 bytes per point x 8 channels = 3200 bytes per record.  Then the number of records I can allow to pile up in onboard memory is 16777216/3200 = ~ 5242 (reality seems to be 4672).  There seems to be an optimum number of records to pull across the pxi bus somewhere in the neighborhood of 32 (32x3200  =  102,400 bytes...is there a known optimum size??) Attached is my current rev of the code.

[jswalwell]

Make that 4512 records before running initiate again...hadn't saved some of my front panle defaults.

[RatedPG]

I just upgraded to NI-SCOPE 3.5 and Measurement Studio 8.6 and am seeing the same behavior. It seems that the integrity of the data structure is not being maintained and I will see data “jump” from one channel to another and back again. I cannot determine a pattern.

Has this issue been resolved? Is there a patch or temporary work-around available?

[Jennifer O]

A fix will be available with the next release of the driver.  In the meantime here are some work-arounds to help.

1) Enable an even number of channels (i.e 2, 4...8).  One channel should also be fine. You do not need to fetch the channels that you are not using.

2) Fetch at most 2 records at a time.

Of course working with NI-SCOPE 3.4 is also an option if you wish to temporarily revert back and you do not require the new features released with NI-SCOPE 3.5.

 Hope this helps,

Jennifer O.

[jswalwell]

Hi.  NI-SCOPE 3.5.1 was released earlier this month.  This problem is not listed in the known bug fixes section of the read me.  Has it been fixed?

Thanks,

Jarred

[ESD]

I too am curious to hear if this problem has been resolved in the new driver version.

[dklipec]

Yes, NI-Scope 3.5.1 does fix the multirecord fetch bug with the 5105.  Any combination of number of channels and records should now be properly supported.