I have two instruments (at primary address 3 and 9) which have been daisy chained together and to an old HP unix computer. All have communicated without problems for years. (One instrument is an Acroamatics PCM decommutator. The other is an HP 3488a switch.) I am attempting to control the two instruments from a newer HP unix box (C100 with HP-UX 10.20) with a C program via a National Instruments EISA-GPIB controller. I have no problem communicating with either of the two instruments if only one is powered on at a time, but if both are powered on, errors occur.
Today, I installed a new PCI-GPIB+ controller/analyzer into a PC and added it to the GPIB bus so I could monitor
what was going on. I tried talking to the HP3488a with the other instrument turned off and then tried doing communicating with the HP3488a with the other instrument turned on. Here's the series of commands I issued
ibsic (controller)
ibsre (controller)
ibdev (HP3488a)
ibclr (HP3488a)
ibwrt "RESET" (HP3488a)
The one major difference I noticed was the status of the EOI line. With both instruments powered on the EOI line was being rapidly asserted and unasserted when the ibwrt "RESET" was issued and continued to do so after completion of the mini-program. This did not happen when only the 3488a was present. If I unplug the GPIB cable from one of the instruments, the EOI light on the analyzer stops blinking. If I plug the GPIB cable back onto the instrument, the EOI light flashes again.
Am I right in assuming that one of the two instruments is attempting to assert this line and the other is trying to unassert it? Is this the result of incorrect EOS and EOI settings o
n my part when I issue the ibdev command, or is there something wrong with one or both of the instruments? Unfortunately both instruments are quite old and I have limited documentation . I'd really appreciate any ideas or advice anyone may have.