[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
FW: 3.3V Decoupling
- To: Mailing List Recipients <pci-sig-request@znyx.com>
- Subject: FW: 3.3V Decoupling
- From: "Taylor, Mike" <mike@buntypost.dundee.ncr.com>
- Date: Fri Nov 22 08:40 GMT 1996
- Encoding: 41 TEXT
- Resent-Date: Fri Nov 22 08:40 GMT 1996
- Resent-From: pci-sig-request@znyx.com
- Resent-Message-Id: <"M0-BN1.0.id5.XUMbo"@dart>
- Resent-Sender: pci-sig-request@znyx.com
----------
From: pci-sig-request[SMTP:pci-sig-request@znyx.com]
Sent: 21 November 1996 17:56
To: Mailing List Recipients
Subject: Re: 3.3V Decoupling
part1 (text/plain)
------------------------------
> Date: Thu, 21 Nov 1996 15:50:35 -0500
> From: Jerry Sabath <sabath@atla3.agfa.com>
> Subject: 3.3V Decoupling
> To: Mailing List Recipients <pci-sig-request@znyx.com>
> Section 4.4.2.1 of the PCI Spec Revision 2.1 states:
>
> Additionally, all +3.3V pins (even if they are not actually
> delivering power) . . . on the PCI edge connector provide
> an AC return path, and must have plated edge fingers and be
> coupled to the ground plane on the add-in board . . . to
> ensure that they continue to function as efficient AC
> reference points . . ."
>
> Can someone explain why the specification does not allow
> unplated fingers? Would a board with unplated +3.3V fingers
> have significantly different characteristics than an
> unpopulated PCI slot?
The decoupling of the 3.3V pins is so that they act as an AC ground
for YOUR board... There aren't enough ground pins on the PCI bus
alone to provide this function...
-jrp
This assumes that motherboards which do not provide 3v3 to the slots
follow a similar decoupling scheme!!!
PCI 2.1 covers plugin cards, but does not appear to cover AT style
motherboards or LPX style and their riser cards. (or have I missed it?).
Mike.
9 ,