zik@bruce.cs.monash.OZ.AU (Michael Saleeba) (12/18/90)
I recently bought an A2620 card secondhand and have had no real problems with
it except when I tried to run SetPatch with kickstart in RAM or WB2.02. When
I did this the card tends to guru after random periods of time.
I thought the card might be faulty so I removed it and tried the system
without the card. Sure enough WB2.0 worked fine. Next I tried the card back
in with the A2088 bridgeboard removed. CRASH. Finally I tried the A2620 in
and my Commodore 2Mb RAM (16 bit) expansion board out. Amazingly the system
now ran SetPatch and WB2.0 with no problems, and on the 68020!
This all leads me to belive that there is a bus contention problem between
the A2620 and the RAM card. Is this a known problem? If not, does anyone
have any idea which card might be causing the problem? And what solution
could be found?
Thanks in advance,
________ _
|_____ / | | "I don't want the world - I just want your half."
/ / _ | | __ Name: Michael Saleeba
/ / |_| | |/ / At: Monash University
/ / _ | / E-mail: zik@bruce.oz.au
/ / | | | \ Snail mail: 19 Evelyn Street,
/ /____ | | | |\ \ Clayton, 3168,
/_______| |_| |_| \_\ Victoria, Australia.daveh@cbmvax.commodore.com (Dave Haynie) (12/20/90)
In article <3482@bruce.cs.monash.OZ.AU> zik@bruce.cs.monash.OZ.AU (Michael Saleeba) writes: >This all leads me to belive that there is a bus contention problem between >the A2620 and the RAM card. Is this a known problem? If not, does anyone >have any idea which card might be causing the problem? And what solution >could be found? There's certainly no bus contention problem between the cards. Some of the very early A2052 cards may have had slightly marginal timing, which could be a problem when used with 68020 and 68030 systems. This goes way back, but I gather there were three types of A2052s made. The first used 74LS245s as data bus drivers. These proved to be too slow, so they changed them to 74F245s. Only, that proved to be a problem, because during reads, you wound up with the DRAM bus, still in 3-state, connected to an enabled 74F245. While the LS parts tend to float high when the inputs are undriven, the F parts jump around and even oscillate. So this was remedied by adding pullup resistors to the A2052 side of the data bus. Far as I know, that was a complete fix. Anyway, that's the 2052 story as I've heard it. Before panicing about the A2052, though, you should check your system power. Both the A2620 and the A2052 take a bit of juice to run. It's certainly possible that the combination is bringing your system voltage down to marginal levels. With both cards installed, have you or an EE type friend should try to measure the +5V supply, not on the motherboard, but at the back top end of each card. If it's noticably different than 5V, adjustment of your power supply may solve your troubles. > / / _ | | __ Name: Michael Saleeba -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "I can't drive 55" -Sammy Hagar