#CPU board 3.0/3.1
2 messages in this thread
Here's a question that is apparently stumping all the experts out there…
I just purchased the dkb 3128 memory expansion board for my 4000/040. In the
manual it warns that the memory board will not work correctly if the CPU card
is rev 3.0 instead of rev 3.1 or higher. Sure enough, I have a CPU card rev
3.0! And sure enough, the computer says the expansion board is defective —
just like dkb said would happen with rev 3.0.
After contacting a repair facility and being told that they did not have the
spec sheet on the revision, the repair person suggested I contact dkb and have
them fax the repair person the info on the revision.
I contacted dkb, but they do not have any documentation on the revision. They
said that they thought it was a major board revision. However, they did say
they recalled reading about a "work-around" a few years ago. The person at dkb
said that they thought that I might be able to find it by looking for
information on a fix for the "3640."
I have not been able to find any such information.
Does anyone out there know anything about this. It is very frustrating to have
a memory expansion board and not be able to use it!
Any help would be GREATLY appreciated.
—Clinton—
MSTie #873 (cruisin' on AutoPilot v1.80)
Clinton: The difference between a 3.0 vs. 3.1 A3640 card is that U209 is
upgraded to a new PAL and another chip on the card has one of it's pins cut. So
cutting the pin yourself might not be much of a problem, but getting the
upgraded PAL could prove to be difficult. Here is some other information I got
from Dave Haynie when I was discussing the problems with him:
"It is the existence of the new U209, plus the one other change (a lifted pin
on another IC) that makes a processor board "Rev 3.1". If both these changes
haven't been made, the board is still 3.0. Technically speaking, the problem is
where that card is sampling the STERM* (synchronous termination) signal from
the A4000's local bus. Apparently, though I don't know just why, prior to Rev
3.1 (eg, the new U209 PAL), the board was sampling STERM* too close to where it
might change. This wasn't a Zorro III problem per se, though the old Buster
(Rev I) had the slowest STERM* of the three devices driving it (the others
being RAMSEY and GARY). The new Buster (Rev K) speeds this up, but it can't
eliminate the problems on the '040 card itself."
You might try contacting SMG (Service Management Group) to see if they have any
information on the processor board updates. They might even have a few of the
upgraded U209 PAL's lying around. Good luck!
// John – (TechnoJunkie at heart)
\X/ jgager@BMI.NET