at-gpib Interrupts

[WA7KGX]

I have an AT-GPIB ASSY 180625-01 s/n 3748  2C0  IRQ10 DMA 5 running under Fedora 386 Linx 16 with the

Linux GPIB 3.2.16 package.  This is on an Abit VH6T motherboard with an 800 MHz PIII and 1.5 G RAM.

This combination is able to send commands to a Racal-Dana model 1992 nanosecond universal counter.

It also reads data from the counter with no apparent errors.  The counter does not strain the GPIB connection

with high speed data.

I downloaded the ATE7.ZIP package from the FTP site and ran IBDIAG under FreeDOS 1.0.  It reports

that DMA and interrupt is not working.  It does pass the board if interrupt and DMA are set to "no".

I also tried similar tests on a GPIB-PCII/IIA ASSY 181065-01 Rev E1 with the same results.

Test passes only if IRQ and DMA set to "no".  I got these results using a BX6 motherboard.

Either card works with the BX6 motherboard.  Plugging thr 8 bit card into the VH6T motherboard

prevents the VH6T from starting up, even when I tried a variety of card settings.

Is there something in FreeDOS that prevents card interrupt and DMA from working?

[WA7KGX]

It appears the AT-GPIB **IS** generating interruots in Linux, about one per byte read.

[WA7KGX]

Is "ate7" the correct software for testing this card???

I did not see any error messages claiming otherwise.

Yet, interrupts don't work in the diagnostic but apparently

do in Linux.

[JL@NI]

HelloWA7KGX,

Here is a good article to start with:

http://zone.ni.com/devzone/cda/tut/p/id/2874

Regards,

Jeff L.

[WA7KGX]

Reading that brings back memories of writing and debugging interrupt drivers for

the PDP-8, PDP-11, 6800, and 8086.  Some of the chips had timing requirements

that required NOPs and/or careful rearrangement of instructions.  And I could pull

an all-nighter at will.  Those were the days.

So --- what is the correct diagnostic to use on the AT-GPIB board described above?

"MainDECs will run on a beached whale" - comment made at a Unix conference.

[JL@NI]

Hello WA7KGX:

I was unable to find information about the product that you are referring to in your first post. This product seems to be obsolete, which typically means that information and support are very limited. If you were able to program interrupts in

Fedora 386 Linux 16 with the Linux GPIB 3.2.16 package, then I am not sure that this is a hardware issue. I recommend posing this question on a FreeDOS forum. People who have worked with FreeDOS will most likely be the best folks to help find an answer. In the meantime, I am asking the Product Support Engineer to see if there is anything else we can do to resolve your question.

Regards,

Jeff L.