GPIB EABO error with Agilent 3458A Multimeter

[MaartenDecat]

Hi,<meta http-equiv="CONTENT-TYPE" content="text/html; charset=utf-8"><meta name="GENERATOR" content="OpenOffice.org 2.0 (Linux)"><meta name="AUTHOR" content="System User"><meta name="CREATED" content="20080723;10445900"><meta name="CHANGEDBY" content="System User"><meta name="CHANGED" content="20080723;11084900">

We are trying to retrieve single measurements from the Agilent 3458A Multimeter over a GPIB (IEEE 488.2) connection. This does not give the results you should expect, for some reason. For testing, we are using the NI-488.2 Communicator to communicate with the device.

The problem started when asking for the device id. When querying

ID?

the device returns

HP3458A

1.999952011E+00

1.999946624E+00

1.999946771E+00

1.999947701E+00

[...and so on...]

1.999964573E+00

This is followed by an error from the NI-488.2 Communicator:

iberr = EABO

EABO indicates that an I/O operation has been canceled, usually due to a timeout condition after a GPIB read. Before reading from the instrument, verify that the GPIB command you are sending is understood by your device and instructs it to place data in its output buffer. For information on your device's command syntax, consult the instrument manufacturer's user documentation.

This is probably due to the fact that the multimeter is in continuous mode and constantly returns sample values which are received at our end of the line. So then we tried the 'Making Single Readings' example on page 83 of the manual by sending the commands:

RESET

TARM HOLD

DCV 10

TARM SGL

The last line is a query, and the result is:

1.999975245E+00

This result is returned after a long time (>10sec), which is way too long for pratical use. Again, the result is followed by the error:

iberr = EABO

EABO indicates that an I/O operation has been canceled, usually due to a timeout condition after a GPIB read. Before reading from the instrument, verify that the GPIB command you are sending is understood by your device and instructs it to place data in its output buffer. For information on your device's command syntax, consult the instrument manufacturer's user documentation.

The fact that only one sample is returned, is correct. But what about the EABO error that keeps showing? We don't seem te get rid of it, whatever we try.

Any help, please?

Best regards,

Maarten Decat

[NathanT]

Hello Maarten,
       Have you considered trying out one of the Instrument Drivers for your instrument that are available on IDNet?  They implement most of the instrument functions in an API that is much easier to use than the programming manual.  Which development environment are you planning on using?

Cheers,

NathanT

[MaartenDecat]

Hi,

We tried those drivers. We are working with Matlab Drivers (.mdd, using the Instrument Control toolbox) and those are most easy to implement ourself using the GPIB commands. We've implemented all other device drivers like that and this multimeter is the only device that won't work. For this device, the Matlab drivers do not seem to be used a lot, most of the google hits are IVI or LabView it seems.

I've tried to use the IVI driver and use it as a Matlab driver using the method in the Matlab help method. That gave some other errors and I'm not experienced with those drivers, so the Matlab drivers still looked the easy way out of this problem...

Any idea what the EABO error means?

[MaartenDecat]

Problem solved with the help of Agilent.

Apparently, the device was finished before the 488.2 standard and therefor the 3458A is not compliant with that standard. It is needed to send the END ON or END ALWAYS command to the device after resetting. This commands the device to work with EOI as we would like it to. For more information about this command you can refer to page 176 of the 3458A's user's guide.

[demux]

Greetings Maarten,

I think this article is a good place to start troubleshooting. There are many more articles on the National Instruments website but these all refer to the National Instruments driver.

Kind Regards,
Wouter

[demux]

Hi again Maarten,

It seems we were typing at the same moment  during our previous post. Therefore my reply after you solved the problem yourself.

Thank you for sharing this information with us and all the best with the further development of your application.

Kind Regards,
Wouter
National Instruments
Applications Engineering

[NathanT]

I am glad to hear that things are resolved now.  It generally never hurts to contact the manufacturer.

Cheers,

NathanT

[MaartenDecat]

Indeed, and their response was suprisingly fast! And now the answer is on the internet too, the best solution.

Thanks for the cooperation!

Maarten

[Amioo]

You saved my career