NI-PXI-GPIB - communication - Change in data received compared to actual data

[Karthik_DP]

Thanks for your immediate reply. Answering to your queries, There are 14 Instruments connected in the GPIB bus, with obviously different addresses. The GPIB cable length to the nearest Instrument (1.5m) from the NI-GPIB card does not exceed 2m, with the overall length (10m) well within 20m. Similarly, at any instant, no 2 instruments are separated by a cable distance of 4m. There are no Bad or Loose cables, which we have physically and electrically verified. We have tried the tests after replacement of: 1. NI-GPIB card 2. GPIB cable Still the issue is open. The system is a closed system with local air cooling provision. For the past (more than 5 years), these checks were found to get cleared, without any EABO issue. Pls. check my earlier post with attachments for the data defects. The measured data from the instrument got captured in the NI spy, where as during the GPIB communication, the EABO (iberr value 6) error had occured. Pls. suugest us the possible causes of EABO (iberr value 6). The system was fully functional in the past, so there is no question of 'ever it worked at all'. The Instrument with which we are facing issue is 585C (Pls. check my earlier posts for more info). I can post you the details of the instruments used in our system. The full system is an In-house unit. Adequate care and proper handling of instruments, cables and units are strictly followed, as it is a precise test system. Pls. project us the possible causes, based on this feedback or let me know, if you require additional inputs.

[Albert.Geven]

The first thing I would try is to slow down the bus speed.

Default it is fast but the slowest system on the bus determines the speed anyhow.

I would go for 1 mircosecond speed.

[Dennis_Knutson]

I agree with Albert. It's a simple thing to try.

I don't believe there is anything wrong with the controller. You don't mention swapping out the failing instrument. Try that. Isolate the instrument by disconnecting all others. If the test passes, connect the others one at a time.

The NI-Trace file you posted did not have the error. Post one that does - not just an image. A timeout error where you get some data indicates the response is taking longer than expected or the EOI is not being asserted before the timeout period. As fully loaded as your system is, a marginal cable or instrument would be my guess.