Why can't I communicate with both of my GPIB devices?

[mrutter@contech]

I have 1 NI PCI GPIB board in my computer
I have 2 identical devices (keithley 237) connected one at address 15 one at address 16
and 2 different programs one program to connect with 15 and 1 to connect with 16
after I tell the 1st labview program to communicate with 15 then I tell the 2nd labview program to communicate with 16. then both communications come back with error device timeout. one at a time, I can communicate find with either device.

Why can't I communicate with two different GPIB devices at the same time?

[tbob]

It is certainly possible to communicate with two instruments, one at a time. Post your code, let us know your Labview version.

[Johannes Eble]

Hi mrutter,

first I would try some other addresses, e.g. 5 and 11. 15, 16 could be
a special address used by the controller.

Are there any other devices on the GPIB? If so then there could be
another instrument with the same ID. Do both (or all) devices show in
MAX, and can you communicate with them in MAX (e.g. query a *IDN? )

And are you sure that you *have* setup the correct addresses in the
Keithley instruments (look in the manuals if you don't know how to
change their addresses)? Some Keithley instrument are in local mode by
default, and you must explicitely set them in GPIB mode.

Sometimes, GPIB message termination can be an issue (although both the
GPIB cards and the instruments seem to get more tolerant of this).

Also, you must be very careful not to mix different versions of
LabView drivers and LabView version. I once had three different
LabView versions on a PC, it was terrible. We had to format the
harddisk and start afresh since we couldn't remove all of the older
versions.

I first try all GPIB instruments in MAX. If all devices appear there,
and I can query commands in MAX, then I usually write my own LabView
drivers with the standard GPIB-read and -write VIs. I write them with
my oldest LabView version (say 5.1), then I am sure they will work
with all versions up to 7.1

Actually I once had a project where *no* GPIB instrument worked 100%.
We could never really find out what was the problem (it was probably a
bad PC <-> Win98 <-> GPIB card <-> LabView 5.1 constellation).


Good luck. 🙂


Johannes



On Fri, 12 Nov 2004 13:49:31 -0600 (CST), "mrutter@contech"
wrote:

>I have 1 NI PCI GPIB board in my computer
>I have 2 identical devices (keithley 237) connected one at address 15
>one at address 16
>and 2 different programs one program to connect with 15 and 1 to
>connect with 16
>after I tell the 1st labview program to communicate with 15 then I
>tell the 2nd labview program to communicate with 16. then both
>communications come back with error device timeout. one at a time, I
>can communicate find with either device.
>
>Why can't I communicate with two different GPIB devices at the same
>time?