Timeout when reading through GPIB bus isolator

[ChrisTT]

Hello all,

I'm having some trouble with an optical GPIB bus isolator (IOtech Isolator488, brand new) and an Andeen-Hagerling AH2500A capacitance bridge (about 15 years old). (We use the isolator in separating a low temperature experiment from mains noise).

When talking to the bridge through the isolator, bus timeouts sporadically occur when trying to read data from the bridge (about one in 200 reads will time out). When directly connected to the GPIB controller, the error does not appear. Also, even when the bridge is on the controller side of the Isolator488, switching on the isolator suffices to produce the error.

Strange thing is, we perform serial polling to wait until the bridge sets its message available (MAV) bit in the status byte to one. So I'm pretty sure the data is in its output buffer before the bridge is adressed to talk.

(We have double-checked all cables. Inserting various delays between commands does not help either.)

I'm grateful for any suggestions on how to troubleshoot. It's kind of annoying having to wait for a random error to occur 😞

Could it possibly be a defective or not fully standard-compliant GPIB module on side of the bridge? All other devices work fine with the isolator. Also, we have heard of a later revision of the bridge working with the isolator.


Message Edited by ChrisTT on 04-07-2008 10:16 AM

Message Edited by ChrisTT on 04-07-2008 10:19 AM

[Steven_T]

Hello Chris,

The only way to know for sure is by using GPIB analyzers to see the handshaking, control, and data line transitions.  Since communication works fine when the isolator is not on the bus, I would start with the isolator.  The best test would be to use the bridge on the other side of isolator and use an analyzer on each side.  This way you can monitor the line transitions on each side.  The isolator should propogate all signals as the come in (except serial polls which are usually a special case). 

You will either find that the isolator or the instrument is misbehaving by not putting the lines in the proper state.  Use your NI-488.2 manual to find out which node should be controlling eah of the GPIB lines.

Controller - IFC, ATN
Talker - DAV
Listener - NRFD, NDAC 

Steven T.

[ChrisTT]

Hi Steven,

thanks a lot for your reply! Unfortunately, it seems we don't have a access to a bus analyzer here. We have, however, tested against a newer version of the bridge which does not produce the error.

The nice people at Andeen-Hagerling told us that there exists a firmware upgrade which will probably correct the problem.

Best regards, Chris

Message Edited by ChrisTT on 04-16-2008 02:25 AM

[ChrisTT]

Just a quick update for those who might read this later: The new firmware has indeed resolved the problem. 🙂

Message Edited by ChrisTT on 05-06-2008 08:16 AM

[Steven_T]

Chris,

Thats good news!  Its also alot easier of a fix than tracing analyzer captures for a few days 🙂

Steven T.