Max GPIB Cable Length

[GoBlue]

I know the GPIB spec calls for a maximum length of all cables to be 20 meters.  I'm wondering if anyone has experience with going beyond that, to say 25 meters, without using GPIB extenders? 

Thanks for your help.

[Collin M]

25m may be pushing it. Do you know which GPIB transceivers are at both ends? Are there any additional loads along the cable? You should be able to test this as long as both ends of the cable are terminated (have GPIB devices attached). As you noted this is out of spec so we can't make any promises. Do you have a oscilloscope to capture and analyze some signals?

[GoBlue]

I actually don't have the setup yet, I'm in the process of getting everything together now.  Like everyone else these days, I'm on a tight schedule and I'm trying to avoid any hiccups, while trying to save money...  I'll have one long cable going to one instrument, and then two short cables going to two other instruments on the bus. 

Thanks!

[Collin M]

If this is for a production system I would be against using a cable configuration that is out of spec. If you wanted to set something up temporarily or in a lab somewhere it may be ok.

Adding GPIB devices and cables typically improves the signal quality as the additonal cable and devices add capacitance. However, at some point point adding more cable may cause reflections that will cause errors. I think this would be an issue for two devices close to each other at one end of a long cable, and then other devices at the other end of that long cable. When the two devices that are close to each other attempt to communicate, the reflections from the far end may cause a problem because the reflections will take a relatively long time to appear at the original devices.