Re: Buggie powersupply doesn't like RamFAST w/1MB

[spec@vax2.concordia.ca (Mitchell Spector)]

In article spec@vax2.concordia.ca (Mitchell Spector) writes...
>
>   This is very odd, I hope someone can figure this one out. Tonight I
>replaced my stock Apple 60W supply with one of Stephen Buggie's 150W
>adapted PC powersupplies. To make this brief, as soon as I switch on
>my IIgs, the RamFAST reports "Fatal Memory Error =09" and refuses to
>operate (even if I try a warm-boot or invoke the GS self-test). Around
>9 out of 10 times this happens, if I keep switching the main power on
>and off, *eventually* the RamFAST starts up my hardrive and everything
>runs normally. I've run the 'RFTEST.SYSTEM' diagnostic program and copied
>files from different partitions, the 1MB of cache is fine.
> 
>   As soon as I return my stock 60W Apple supply, the RamFAST no longer
>has this problem. It works perfectly 10 out of 10 times when I switch
>on the power. Interestingly enough, I think it's the 1MB of cache
>configuration, since switching in a 256K cache RamFAST doesn't seem to
>have this problem. Why would a higher powered supply affect a 1MB cache
>board and not a 256K cache board? Or rather why should it affect my
>RF at all?
> 
>I certainly can't go on like this, cycling my power-switch a dozen or
>so times every time I start my system. Can someone come up with a reason
>as to WHY the Buggie supply is messing up my 1MB RamFAST?

   I've managed to solve the above problem. I swapped the pair of 1MBx4
DRAMs into a Sequential Systems RamFAST SCSI, then plugged that into my
IIgs (replacing the CV-Technologies board). The problem has now completely
*disappeared*, I no longer get a 'Fatal Memory Fault =09' error upon
switching-on the power. As well, I'm able to keep 1MB of cache installed.

   My theory is either one of two things: A) My CV-Technologies board
had a slight hardware defect which was intensified by the new supply.
B) The minor changes found in the Sequential board (Z180 over Hitachi
processor, etc) were enough where the board wasn't picky about the new
supply. If theory 'A' is correct, perhaps this was also the cause of
my 15/64 ZipGS locking up the machine, although I wouldn't bet on it.
Regardless, it's nice to have back-up cards sitting around! ;-)

Mitchell Spector
spec@vax2.concordia.ca