[Scgeeks] Re: RESOLVED Re: RAM voltages

[Twayne]

> "Twayne" <nobody at devnull.spamcop.net> wrote in message
> news:g828j7$efk$1 at news.spamcop.net...
>>   You CAN replace 2.5V RAM with 2.6V RAM and vice versa, as long as
>> they are non-ECC and non-EPP.
>
> I figured enough, you can never have perfect voltage control from the
> PS AC/DC converters, but I didn't open my mouth in case I was wrong
> since I don't have any experience with operating RAM at off-nominal
> voltages ;-) Most SS electronics will actually function ok up to at
> least +/- 0.3 volts, in my experience, but they can be temperature
> dependent if operating at a lower voltage than is standard. I've seen
> electronics boxes work fine in ambient conditions and lower than spec
> voltages, but when you drop the temp to some "magic" point, like -15
> or -20 C, they cease to operate because some component isn't getting
> enough juice.

Yeah, and the "SSTL" had me curious too but I never got a definition for 
that.  But a system that worked great here and not "there" is exactly 
what I was worried about; I agree.

>
>  So, we were able to pick a
>> PC2700-200Meg 512 Meg RAM stick, which ran at 2.5V, and I ordered
>> it. That way I could just add it TO the 256 that was already there;
>> that will be enough RAM for this machine.
>
> What the heck is a "200Meg 512 Meg Ram stick? I assume it's 512 of
> RAM, but why the stupid and confusing name?

Apparently at least Crucial and the SIW program I used add the FSB max 
speed to the RAM type.  PC3200-200 and PC3200-400 for example indicate 
the max FSB speed on the board.  Since this machine is 133 I can see why 
either would work.  e.g. the SIW indicated 256 Meg, Memory Type SDRAM 
DDR-400, which I admit is a little redundant, but Crucial didn't blink 
an eye at it; they returned the same data.
   BTW, if you're interested, their RAM investigator is a downloadable 
app; I grabbed it for future reference; you don't HAVE to do it online.
_Aside:_  Just for grins, since the BIOS allowed it, I tried increasing 
the FSB speed to the next setting to see what happened.  What happened 
was, I had to look for the BIOS erase jumper to get booted again!<G>  It 
wasn't labeled, either; had to guess by its location on the mobo. Then I 
noticed the label: It's UNDER the header base!  Duhhh!

>
> But then I also noticed the BIOS
>> is in a chip/socket so maybe there isn't any upgrading it.  Hmm,
>> something else to look into!  BIOS is new territory to me; if I
>> can't flash it, I know nothing about it.
>
> The BIOS in my old PC was stored the same way. You had to open the
> box and short out some pins by moving the connector cap to erase the
> existing BIOS, then you flashed the new BIOS to the EEPROM (I *think*
> it's EEPROM). Very scary process the first time you do it!

Yeah, now that I've identified the mobo maybe I can figure out if/how it 
can be flashed.  Couldn't hurt; it's got a 2002 date on it!  Watch it 
turn out to be a unique BIOS!

It's been an "interesting" exercise if nothing else.  I've got it fully 
rebuilt now and it's actually not too bad.  Pauses from anything that 
causes a massive prefetch reorg are a little annoying, but all are less 
than 10 seconds at least, and the page file use is high until I get the 
new RAM.  Least I hope the new RAM helps that<g>.

Cheers,

Twayne