"Kyle Brant" <> wrote in message news:bqhhq1$21fmjr$...
> "L David Matheny" <> wrote in message
> news:bqguqc$22scpj$...
> | "Gibby" <> wrote in message
> | news: m...
> | > I was sucessessfully running three (3) Crucial 256MB, 168 Dimm,
> | > Sdram PC133, 32Meg x 64, Unbuffered, Non-parity, 7.5ns,
> | > 3,3V, CL=3 sticks in my FIC PA=2013 Rev. 2 E-O036,
> | > K6=3/450, MB. I purchased the same as above, but CL=2,
> | > and I can only get two sticks in the first two slots to run
> | > successfully. I have tried all 3 sticks in the first two slots
> | > and system runs fine. But when I try putting the 3 sticks
> | > (Crucial# CT32M64S4D7E.16LT )in together (in all 3
> | > slots), I either don't get a good memory count, or if it does
> | > (just once) I can't get the complete Bios bootup (it stops) -
> | > I don't get to where it searches HD for Windows 98SE.
> | > Question: Why will 3 CL=3's work and not 3 CL=2's
> | > (only two will work at one time)? Perhaps my BIOS
> | > memory settings are wrong for CL=2. But why do two
> | > work but not three?
> | > Thanks for any help. Gibby
> | >
> | Timing is tighter for CL=2. There's probably too much
> | capacitance in the circuit board traces and the signal rise
> | times get sloppy with three sticks just because there are
> | more traces to drive.
> |
> I think you are warm with this analysis. It might be better said
> that the sdram device output loading on the memory interface
> and sdram circuits is higher with 3 versus 2 unbuffered output
> sdram sticks. The capacitance loading is probably substantially
> the same, as the capacitance (residual) is attributable to the
> mobo runners or traces layout. All of the mobo memory
> traces get driven no matter how many sticks you are using.
>
Yes, it's overall loading on the memory interface drive circuitry
that causes the timing problems, but surely the SDRAM circuit
board traces have some capacitance, too.
|