VirtualBench

[Pazzer1]

Hi, I run virtual bench scope to capture valve response traces, these are normally around 10 ms opening time and display a dip when the valve is open, however on the windows 2000 machine sometimes I get what looks like two response times, the first one, then the second one is superimposed on it but delayed.

I also run the same software on an XP machine and do not get the same problem.

Is there an incompatibility with windows 2000 and virtual bench?

I am running a labview application as well as the virtual bench scope, I send a command from labview to a fieldpoint module to produce a pulse to a valve driver card, this then supplies a 60ms pulse to the valve.

The virtual bench scope then captures the current and voltage trace of the valve operation.

The digitiser is and NI5102.

attachments TLV4B shows a good trace, and TLV7A shows a strange trace.

Any advice?

Thanks

Allan

[JamesC]

Hi Allan,

This might be a silly question, but as I can not find any compatibility issues with Virtual Bench and win2k this is what i'm left with.

Looking at your file names, you seem to be controlling 2 valves, TLV4B and TLV7A, have you tried the Win2K machine connected to TLV4B (the valvue you are getting a good trace from) and can you connect up the WinXP machine to TLV7A (Valve with the bad trace). Just to make sure it is a system problem or the valve showing wear.

Another thing you should try is to switch the 5102's between machines, to see if we have a calibration issue with the 2 digitisers.

I hope I have given you some useful suggestions, if you could try them and get back to me that would be great.

Regards
JamesC
NIUK and Ireland

[Pazzer1]

Hi James,

The good valve is taken off a machine running XP, generally we don't have a problem on the xp machines, however on the win2k machine it appears randomly, i.e if we start getting the problem and we restart the PC, the traces generally return to normal.

The 5102 card has been swapped to another machine and that is ok, so the only common factor seems to be the win2k.

The valves also show nominal traces if we use an external scope.

I think we will probably update the machine to XP, but I was interested to know if anybody had seen anything similar and had a possible explanation.

Thanks

Allan

[Kennon_Cotton]

Allan,

I don't remember of hearing of this sort of issue and I can't think of any incompatibilities between VirtualBench and 2000.  I'm thinking it could be one of 2 things, either a difference in the NI-DAQ driver between your 2000 and XP machines or something weird about the grounding on that 2000 machine. 

If the problem is with the driver it could be a version difference NI-DAQ 6.something vs NI-DAQ 7.something or it could be the driver behaving differently on 2000 vs XP.  If you upgrade that machine to XP you'll probably eliminate the driver from being the issue.  But upgrading just the driver should be easier than changing to OS so you might try it first.

If it is some sort of grounding issue, changing the software won't help, but getting the OS and drivers to be the same and knowing the board is good would eliminate everything but the hardware in the computer itself.

Kennon