lsdaq doesn't list my PXI-6289 (Scientific Linux 5.5)

[torturedbabycow]

Hi, I'm new to both NI and Linux, but I've just been tasked with getting our M-series DAQ (PXI-6289) to work. The computer is running Scientific Linux 5.5, and has both DAQmx and DAQmx Base installed already. It's connected in the PXI-1036 chassis with PXI-8336 controller.

But, lsdaq doesn't find anything:

$ lsdaq
--------------------------------
Detecting National Instruments DAQ Devices
Found the following DAQ Devices:
--------------------------------

Even though everything seems connected otherwise:

# nilsdev
NI PXI-6289: "Dev1" 
NI PXI-6289: "Dev3" [Not Present] 
NI PXI-6289: "PXI1Slot6" [Not Present]

# nipxiconfig -l

List of identifications:

Chassis 1 (PXI-1036)
Slot 1: PXI-MXI-4
Slot 5: PXI-6289 (PXI5::12::INSTR)

Any ideas on what I could try next? Any other info that might be helpful for me to post?

Thanks!

[amezam]

Hello Torturedbabycow,

Can you please tell what versions of the DAQmx drivers do you have?

Best Regards,

[torturedbabycow]

Oh of course: DAQmx 8.0.1 and mxbase 3.4.0.

(p.s. There was an unexpected rearrangement of computing resources, but I should be able to reply right away now. All the previous info still stands.)

[amezam]

Hello Torturedbabycow,

Why do you have both drivers installed? The DAQmx 8.0.1 is for Windows 2000 and XP. In the DAQmx base 3.4.0 readme file, it states it is supported in Scientific Linux 5.5 and it includes support for the PXI-6289. My recommendation would be to uninstall DAQmx 8.0.1 and reinstall DAQmx base 3.4.0. 

Best Regards,

[torturedbabycow]

Thanks for the responses! Just for the record: I'd inherited the original machine from a long, storied history of running various NI devices. We ended up moving the DAQ setup to a computer with fresh installs of everything, essentially bypassing the earlier problems. The questions remaining are more software-related, so should probably go in a new thread if they are not easily fixed..

Thanks again!

[JoeFriedchicken · Accepted solution]

Hi everyone,

I'd like to clarify a few things in case there was any confusion.

DAQmx 8.0.1 was also released for Linux [1], but my recommendation for the PXI-6289 is to use DAQmx 8.0.2 [2] and not DAQmx Base (the latest version is now 3.6 [3]). Both DAQmx 8.0.2 and DAQmx 3.6 support the 6289, but DAQmx has better performance and exposes the complete feature set of the hardware. DAQmx Base performs all measurement types, but does not implement the less popular features (like trigger delay, sample clock delay, subsystem routing, and more). The best time to use DAQmx Base on Linux is for USB DAQ hardware since it is the only NI driver that supports those devices.

Finally, about the list output: the nilsdev command asks DAQmx which hardware it has under its control and the lsdaq command does the same for DAQmx Base. As with any OS, only one driver can associate with each device, and DAQmx will pre-empt DAQmx Base when enumering devices since it has complete support for the hardware. Thus, nilsdev reported the 6259 since DAQmx had taken ownership and lsdaq reported nothing since there were no USB devices attached.

[1] NI DAQmx 8.0.1 for Linux
http://joule.ni.com/nidu/cds/view/p/id/1194/lang/en 

[2] NI DAQmx 8.0.2 for Linux
http://joule.ni.com/nidu/cds/view/p/id/2322/lang/en 

[2] NI DAQmx Base 3.6 for Linux
http://joule.ni.com/nidu/cds/view/p/id/3434/lang/en 

[torturedbabycow]

Thanks for clarifying! I'm glad to know that it was, in fact, just an issue of two competing drivers.

We do have nidaqmx 8.0.2 on the new setup, since there's also some existing code of our own that requires it (rather than mxbase). I do want to point out that one of the first things I came across early in the process was this chart: http://digital.ni.com/public.nsf/$CXIV/ATTACH-AEEE-8NDLGL/$FILE/LinuxSupportMatrix_subset.png. Had I followed this chart, I'd be struggling to force old code designed to work with nidaqmx to instead work with only the subset in mxbase. Granted, getting nidaqmx installed was a struggle (which required someone with vastly more Linux-fu than me to fix, so I can't comment on details), but it still seems like there's some conflicting recommendations and resources out there.

In any case, thanks very much for the details and the explanation.