[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Power sequencing (3.3 V/5 V)
- To: Mailing List Recipients <pci-sig-request@znyx.com>
- Subject: Power sequencing (3.3 V/5 V)
- From: Pavel.Peleska@mch.scn.de
- Date: Tue, 17 Dec 96 18:37:22 GMT (Original MEZ)
- Resent-Date: Tue, 17 Dec 96 18:37:22 GMT (Original MEZ)
- Resent-From: pci-sig-request@znyx.com
- Resent-Message-Id: <"rtTUe2.0.Oa2.R_jjo"@dart>
- Resent-Sender: pci-sig-request@znyx.com
Hallo PCIers,
the PCI 2.1 Spec. states at page 142, chap. 4.3.4.2. Power/Sequencing:
" There is no specified sequence in which the power rails are activated or
deactivated. They may come up and down in any order......"
I interpret this that there is basically no requirement concerning the
sequence nor the time it can e.g. take to supply any one of the power
voltages. For example the 3.3 V power source can supply its power after 1
ms while it may take the 5 V power source e.g. 1 minute to power up.
Especially if one power source never comes up, this may happen if separate
DC/DC converters are used with one of them defective, supplying only one
voltage complies to the PCI spec.
Unfortunately some SCSI and Ethernet devices that operate with 3 V PCI
signaling but need 5 V for the core, require power sequencing with 5 V
having to come up before the 3.3 V are up. Demage to the I/O buffers may be
caused by not applying the power in the right order. Supposingly this
requirement can always be fulfilled in the PC, but causessubstantial
problems in embedded systems.
Question:
- Is it correct that devices that are PCI 2.1 compliant have to be power
sequencing tolerant?
- Does this mean that devices that require power sequencing cannot claim to
be 2.1 compliant?
Any comment appreciated,
Pavel Peleska Tel: ++49-89-722-41253
Siemens AG Fax: ++49-89-722-28502
SN EBG1 Email: Pavel.Peleska@mch.scn.de
Hofmannstr. 51
81359 Munich
Germany
¡