NI PCI-GPIB vs. HP 3488A SCU

[Methis Elddir]

On Thu, 04 Mar 2004 11:22:04 -0600, AllanH wrote:

> I'll see if I can set up a test like you describe. Unfortunately, this
> hardware and software is part of a deployed system so time to test out
> theories is hard to come by.
>
> Thanks again,
> Allan

Let us know how it turns out. Good luck!

[AllanH]

Hello again. Just a note to let you know that the new version of the driver (2.1) did not solve the problem. All three SCUs continue to reset simultaneously after random amounts of time communicating with them. Please let me know if you have any other suggestions (configuration suggestions for old equipment, for example) as we feel at this point we have no choice but to purchase and try some Agilent controller cards instead.

Thanks for your help,

Allan

[Koninika]

Hello Allan,
PCI-GPIB should work fine with Windows if it worked fine with HP.
Try communicating with your instrument on Measurement and Automation Explorer (that is installed with the GPIB driver). If you expand the drivers and interfaces section on their, highlight your board and then right click and bring up scan for instruments it will bring up your instrument on their. Then highlight the instrument and right click and do scan for instruements. Then you can type in *IDN? to see if it identifies the instrument. You can also type in several commands on their (the commands recognized by your instrument).
Also, try doing a NI-SPY capture by highlighting the instrument and clicking on NI-SPY on the menu bar.
You can see how your instrument is communicating w
ith the GPIB board and if there is an error in the method of communication.
Also, what is your application program? C? Labview? You can probably see what is happening through the ni-spy.
Please email me back so I can assist you further.
Hope some of the suggestions helped.
Good luck and have a great day!
Koninika
National Instruments.

[AllanH]

Not sure what you mean by 'PCI-GPIB should work fine with Windows if it worked fine with HP' exactly. The HP machines did not use PCI-GPIB.

I can communicate with the instrument fine, and in fact it works flawlessly for hours at a time. The problem is simply that all three SCUs on the bus experience a fault simultaneously several times per day. We have an identical system at another site that has one SCU and it experiences the same problem. These problems do not occur when we switch back to using the old HPUX machines with their controller cards.

We are using software developed in C++. I will try capturing the traffic with NI spy.

Thanks for your help,

Allan

[Koninika]

Hi Allan,
Sorry I misunderstood you there.
Reviewing your problem again seems like there could be a timing issue there. The new cards are much faster than the old ones. What you can try is to play around with the timing parameters.
Go back under Measurement and automation explorer, highlight the GPIB device under devices and interfaces, right click and select properties and then go to advanced and edit the bus timing and parallel poll duration. Try various options and see if that helps out some.
Good luck!
Koninika
National Instruments.

[AllanH]

Can you suggest some configuration settings that might work? I've tried a few combinations with no luck.

Here are my current settings:

Bus timing: 2 microseconds
Cable length for high speed: disabled
Primary address: 29
System controller: yes
Assert REN: no
Enable auto polling: no
Enabled CIC protocol: no
I/O timeout: 10 sec
parallel poll duration: default
terminate read on EOS: no
send EOI with EOS on write: no
8-bit EOS compare: no
send EOI at End of Write: yes
EOS byte: 0

Thanks again,

Allan

[Koninika]

Hello Allan,
It is difficult to say what timing setting will be applicable to your system. But I would like you to try doing SPY capture from the system where it works and then compare it to the SPY capture from the system which is failing. You can check if you have a timing issue if you see the time for each call setting. Also the SPY capture lists the errors you encounter so you can see where the error occurred.
Try out the SPY capture and let me know how it works for you.
Good luck and have a great day!
Koninika
National Instruments.

[AllanH]

The system that works is an 11 year old HPUX system using an HP brand HP-IB controller card. The system having the problem is a new Windows 2000-based system using the NI PCI-GPIB card. I do not have the capability of doing a SPY capture on the old system.

I will try doing a SPY capture on the new system and see what I can determine from that.

Thanks,

Allan