#A2630 crashing????!!???
7 messages in this thread
Help!
I am having a problem with a newly purchased A2630 board. It is GURUing at
seemingly random intervals, sometimes even before it gets into the
startup-sequence. Many gurus will fail or go wild and then go into another
guru. I occasionally get pyrotechnics, and the only fairly consistent part
is that the error usually lists an address in the $200000-$230000 range.
The type of guru is pretty inconsistent. This ONLY happens when I choose
Amiga with 68030 (or let it default to that)….the 68000 works fine.
My hardware setup: Amiga 2000, 1 meg chip ram, GVP Impact (Series One) SCSI
8/HD with a Seagate 82meg drive and 4 megs of SIMMS in the HD controller,
and the A2630 with 4meg 32 bit ram (factory direct from CBM, ordered
through CATS).
Now, I know there's a problem. I'd like to isolate the problem as much as
possible before I start talking to CBM about it. Hopefully I'll be able to
test the board on another 2000 tomorrow, but in the meantime, any hints? I
think that address range is where the first autoconfig memory goes, but I
can't find my hardware memory map; that would be the A2630's 4meg, right?
Sigh…I wanted to have SPEED, and all it's causing is CRASHES.
Any help appreciated.
tmm/Richard Cunningham
tmm/Richard Cunningham
Die and go to BBS Hell 🙂
Just a thought… someone more knowledgeable will have to confirm this, but
is it possible that there is a conflict between where the 16 bit ram and
the 32 bit ram are being autoconfig'd? I think you change a jumper on the
GVP (or somehow disable autoconfig) to set its 4 meg of 16 bit so that it
will not autoconfig and then addmem it into a known free memory location.
good-luck..
– Whapping from the North Coast Michael
If it is properly defined autoconfig, the OS sends the address for the
board to live at to the autoconfig device. There cannot be overlap unless
the board is really poorly designed. GVP had been at this too long to have
that large a mistake. -Greg
Richard,
A couple of thoughts. You don't mention what rev motherboard you have
in your 2000. There was a slight problem in some earlier 4.x motherboards
that had a slight DMA conflict with accellerator boards. I don't remember
the exact details, other than there is a tech service bulletin about the
problem and the fix is relatively simple – it involves soldering a small
resistor between two pins on the U601 chip. Your authorized service tech
should have the details.
I also seem to remember vaguely some problems with early GVP SCSI boards
and the CBM accellerators, but I'm not really sure which GVP products were
involved. You may want to contact GVP and ask them.
I thought it was a 1K 1/4W resistor between pins 11 and 20 of U605 on rev
4.4 motherboards. It was actually for a Signetics 74ALS245 that was slow in
bringing the address strobe (!AS) back to its inactive state. -Greg
Greg,
I trust your memory on this much more than mine. You even know which
end of a soldering iron to use! <grin>
I know I had the mod done to my Amiga, but I left the details to my
service tech. Brought my machine in and told him to apply the 'tech note'
concerning possible DMA/accellerator problems. He knew to what I was
referring, and did the magic voodoo stuff. 🙂 Me, I stick to software!
…BobR
I had a similar problem a year ago and found out the problem is that the
A2630 is much more noise and voltage level sensitive than just the 68000.
In my case it was an overloaded power supply, or one that was very near to
the limit, changing boards (and eliminating one) to drop the power draw 4
Amps turned a truly intermittant beast into a well behaving kitten.
You only listed a few boards, so I don't think you would be taking your
system unless you have other boards that you didn't list. My A2630 broke
when I added a bridgeboard, PC hardcard, and parallel port.
-Greg