Re: Subject: [SI-LIST] : Power Distribution along a connector

Jon Keeble (jkeeble@alpha.net.au)
Thu, 02 Sep 1999 12:03:40 -0700

The answer is - it depends.

What is the "benefit" of the "plane on the board and the plane on the =
backplane"?

Let's assume there are two planes: GND and PWR.

Broadly, planes do two things:
(1) distribute power, hopefully with lo DC resistance and LO hi
frequency impedance
(2) provide a well understood. continuous path for signal current to
return to source

The effect of a "break" in the power plane
(1) on power distribution will be small if the DC resistance is low,
and PWR-GND bypassing is adequate
(2) on signal return will depend on the where the return currents are
i.e. in which plane. If they are in the GND plane then there will be no
effect. If they are in the PWR plane, then these return currents will
attempt to pass through the "bridge" created by the fuse, which will
effectively lengthening the paths of the signals among other undesirable
effects.

One solution is to layup the board so all signals in the backplane
connector are on inner layers, sandwiched by two GND planes.

-----Original Message-----
From: Chris Bobek <cbobek@cadence.com>
To: Si-list <si-list@silab.eng.sun.com>
Date: Saturday, August 28, 1999 7:06 AM
Subject: [SI-LIST] : Power Distribution along a connector

I am designing a 6U VME card that has power and grounds (both 3.3V
and
5V) distributed along the length of two connectors. So, if the
board is
a rectangle, you have a row of power pins along the long side (right
side) of the board.

There is a requirement to provide fuses for each power source. So,
that
means there has to be an intermediary plane that the connector pins
bond
to that runs the length of the board. Then, at some point, let's
say in
the middle, a fuse bridges this plane to the power plane.

My question is, will adding this fuse negate all of the benfits we
have
of the plane on the backplane and the plane on the board, since all
positive current has to flow through this little fuse?

Does anyone have a better idea?


Thanks,

Chris

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****