New GPIB-USB-HS interface has intermittent Code 6, 'GPIB Send' errors where older GPIB-USB worked fine

[Photon999]

I am using LabView 8.6 on a PC running Windows XP.  I have a series of LabView VI's that I have used for years with a borrowed (older) GPIB-USB interface connecting to my instruments.  I recently gave the older interface back to my friend and purchased a GPIB-USB-HS (the only option from NI).  Today I tried to run one of the programs, which talks to a few instruments, in particular a Keithley 238 SMU, on which it slowly increases the current value while taking a measurement on another instrument.  The program sets a value from 0, increasing by 1mA steps up to 200mA - it has always worked fine before.  Now, it runs for a while (varies from e.g. 6 to 54 steps) and then has an error (Code 6, GPIB Send) and gets tied up at that point.  I tried the VI that just sends the current value to the Keithley 238, and it does the same thing - I can run it multiple times but eventually it times out with the same error (code 6).

I have tried removing instruments and GPIB cabling, restarted muliple times etc, to no avail.  I wonder if there is an issue with the interface trying to run in high-speed mode? - the data sheet mentions a HS mode with noninterlocked handshake, could this be the problem, and how would I turn off the high speed mode if it is?  What else could be the issue - short of a bad GPIB-USB-HS interface?

thanks, Paul

[Dennis_Knutson]

In MAX, you have the option to change the bus timing. You could try the slowest setting. Sometimes this is necessary for older instruments.

[Photon999]

I tried modifying the advanced settings yesterday, setting Bus Timing to 2microseconds (slowest), also I played with the HS488 cable length, tried it off or 15meters.  I just repeated the test with Bus Timing 2us, HS488 cable length off.  I get the same problem, works for a while and then I get the code 6, GPIB send error which ties up the VI.

I am running an older version of NI-488-2 (version 2.5.2), which says it supports the USB-GPIB-HS interface (in Read Me file).  I have the newer version disk (version 2.7.3) that came with the interface - is there any reason that this should work better?  (I prefer to change as little as possible when things are all working...)

thanks

[Photon999]

OK, I updated NI-488-2 to version 2.7.3, still the same problem.  I also tried the interface plugged into an external USB hub rather than directly into the computer, again, no difference.  Is there something else I can try or change?

[Albert.Geven]

this device uses the max power, so it is good that you tested with external power.

Also the bustiming on 2Us is OK.

I would also switch off automatic polling (parallel is probably switched off, do it also for serial polling)

How long does it take before the system fails?

[Photon999]

OK, I switched off automatic polling and it behaves the same.  It takes a different (random) number of times running the routine to send a value to the Keithley 238 SMU, sometimes one of the first values, sometimes over one hundred values are sent before it goes wrong.  With the old (GPIB-USB) interface I never once saw this problem.

[Albert.Geven]

I have no further ideas, except that the 238 is very old and maybe you are too fast, soemtimes, How fast are you communicating?

I mean incrementing in how many ms?

[Photon999]

I am only incrementing every 3 or 4 seconds - that should not seem to be an issue.

[Albert.Geven]

I agree