A
[email protected]
- Jan 1, 1970
- 0
OK, so we have this guy who's pushing his multi-cap value idea onto
our board. There's gonna be dozens of values. 68pf, 47pf, 680pf, you
get the idea. The idea being to cover the entire spectrum from DC to
2GHz+. My take on this is that every current loop into the BGA package
where the CPU lives is a seperate entity. Wouldn't the location of
each cap influence the L it sees, and therefore its resonant point?
How can the guy be so sure of which caps to use before the layout is
even begun?
And wouldn't the location of each cap be important?
It's for a CPU in a 672 pin BGA, on a 2x3 inch PCB. It runs quite fast
and draws several amps,
our board. There's gonna be dozens of values. 68pf, 47pf, 680pf, you
get the idea. The idea being to cover the entire spectrum from DC to
2GHz+. My take on this is that every current loop into the BGA package
where the CPU lives is a seperate entity. Wouldn't the location of
each cap influence the L it sees, and therefore its resonant point?
How can the guy be so sure of which caps to use before the layout is
even begun?
And wouldn't the location of each cap be important?
It's for a CPU in a 672 pin BGA, on a 2x3 inch PCB. It runs quite fast
and draws several amps,