getting MAX to recognise SCXI modules

[LozJLR]

Hello, Trying to keep it short, we are trying to get Labview 5.1 running on XP machine, until we get new updated kit in future ( current NT machines are failing ). I have tried NiDAQ 7.1 and 8.1 drivers with a PCI-MIO-16-4E card in a Dell optiplex 755, and the current sticking point is configuring the system through MAX to specify the chassis type and modules. The modules are not found automatically, and inputting them manually results in the slots being declared empty, on a test of the chassis (SCXI 1001). The SCXI modules we have are SCXI 1100, SCXI 1162HV, 2 x SCXI 1124, and 2 x SCXI 1160. What should I be doing to get this to work? Different driver? I am not at all familiar with Labview's inner workings!  Thank you.

[rsw]

Hi,

A few questions to narrow down the issue:

1. What you say DAQ 7.1 and 8.1 drivers, do you mean traditional DAQ 7.1 and DAQmx 8.1?
2. Has the system ever worked?
3. Does the PCI card pass the self test within MAX
4. Is the cabling all correct?

All the best, 

[LozJLR · Accepted solution]

Rob,

Thank you for replying. The problem was down to the SCSI cable coming loose from the SCXI chassis end ( those connections seem very prone to coming adrift ). The modules then self populated, and the system now works. The system did work very nicely for ten years, but the old Compaq PC's are failing now, and we've only just got some replacement Dell's, but of course they come with Windows XP, whereas the Compaqs ran WinNT, so it has been interesting trying to get Labview 5.1 running on the new boxes with the old DAQ cards, until we get funding for new cards/ Labview 2010/ Teststand. Problem solved for now!

[rsw]

Hi,

Thanks for letting me know and for marking the solution for others to see, I can imagine getting LV 5.1 working under an XP machine is quite a challenge!

All the best, 

[BME_genuis]

Hey,

I have a similar situation.  I have this old messy code in LV 5.1 which is running on Windows ME using almost the exact same SCXI modules as you.  I tried to get it on a computer running Windows XP to no avail.  Do you have any suggestions?  What is your application?  Are you going to recode in a newer labview version?

-Scott

[LozJLR]

Good Morning Scott,

Our application is to run pneumatic circuits using approx 22 pre-coded special VI's in a front end designed for us in 2000 to run on windows NT Compaq PC's. We don't do the coding thing ourselves, as we are mostly mechanical engineers, running test rigs, and we don't have the time or the will, to learn Labview. We had been promised new PC's and new Labview versions / frontend, but we ended up with some second hand Dells (to replace the failing Compaqs) running XP, and asked to see what we could do, with what we had.

The NI reps told us that LV 5.1 on XP wouldn't work, and if you try to install it, it immediately tells you "wrong OS", but ignore that and install anyway.

The order of operations was as follows;

-Take the XP PC and physically install the DAQ card in one of the PCI slots

-Switch PC on, and let it fuss about the new hardware, tell it to forget about the card, and then get the software ready.

-First load in NIDaq 7.1 ( I think up to 7.4 is OK too ) and ask for traditional DAQ.

-Load in Labview 5.1 ( wrong OS !) . It has NIDaq 6.5 in there somewhere, so that can be skipped.

-Re-start PC and the DAQ card is recognized, and the NI program file structure is created.

-Go back to your ME PC and take a copy of the NI folder in Program Files. Whilst you are there, take a copy of the specific file structure relating to your code or frontend in Labview.This may mean a complete copy of the Labview folders. Transfer these files over to the XP machine, and copy them into the appropriate folders. You can say no overwriting, so that only new stuff is added.

- You should now have Labview 5.1 installed to an XP machine with your specific VI libraries and Labview folders added.

-Hook the PC up to the SCXI chassis with the SCSI cable, and switch that kit on, so that MAX can find the chassis and modules.

- Run MAX using traditional DAQ, add a new chassis, let it find the modules.

-Labview should now run, if you have transferred the files over from ME to XP machine.

We found that we had to fire up our frontend, by finding the "about.vi" file manually. We also found that it complained about missing library files, and so we had to go back to the original hard drive on NT and find those, but we got it going in the end. I'm sure you can too. Let us know, OK?

Good luck.

[BME_genuis]

How do you guys control your timing and did it affect your timing?  I am worried when I move it over it will screw up my timing.  I have a lot of processes going on the same time in parallel loops such as my data acquisition and it is all at the mercy of Windows.

[LozJLR]

Re message No. 7:

I'm not sure what you mean by timing, and if it involves anything intricate within Labview, then I have to duck out, as I don't know much about it at all. Point and shoot user, that's me!

If it helps, since transfering over to XP PC's we have used the kit we have, in anger, for approx ten running weeks,( between 4 workstations) and no sign that it runs differently on XP to what it did on WinNT. Our applications do not rely on split second accuracy, it has to be said. We have rig cycle times between 3 seconds and maybe two minutes. A stopwatch would be good enough!

All the best, Loz

[BME_genuis]

Loz,

I just meant the timing of our analog output might be affected by changing operating systems and yes our timing is more critical (1 sec).  When you say new kit do you mean a newer version of Labview? Are you then planning on recoding the 5.1 stuff?

[LozJLR]

We use analogue inputs for timing light switches, and I would guess that their performance hasn't changed. Things stop and start much as they did. I couldn't see a change of OS affecting that. We were going to get new PC's with XP, and new version of Labview (2010), but this costs money we weren't going to get and also involved another expensive session with someone who would write new VI's for us. ( We just use the equipment, we don't want to have to learn Labview to do so, so we have had consultants write us pre-set VI's in the past. Luckily, this stuff still works, as it cost a fortune! ) We can't afford the time or the expense of one of our number becoming proficient enough in Labview to do everything from scratch, so given that the change from WinNT to XP has been successful, I cannot see any benefit in moving to a later Labview plus Teststand, for instance. If our company changes from XP to windows 7 at some point, I can see things changing then. Has anyone run any version of Labview on a Linux distribution before?