[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: PLL in PCI mode.
- To: "PCI SIG Mailing List" <pci-sig@znyx.com>
- Subject: RE: PLL in PCI mode.
- From: "Ingraham, Andrew" <Andrew.Ingraham@hp.com>
- Date: Mon, 29 Jul 2002 08:36:18 -0400
- Resent-Date: Mon, 29 Jul 2002 05:45:12 -0700
- Resent-From: pci-sig@znyx.com
- Resent-Message-ID: <eAylX.A.riB._cTR9@electra>
- Resent-Sender: pci-sig-request@znyx.com
- Thread-Index: AcI2yYOyEzeeEHzcRruUHSZG9wCspQALsTya
- Thread-Topic: PLL in PCI mode.
There are at least three good reasons why a PLL doesn't work in the 33MHz PCI Environment.
(1) The clock frequency can be anywhere from 0Hz to 33.333MHz. No PLL can do that.
When's the last time I saw one stop its clock? I don't know ... but I wouldn't put it past a laptop to throttle the clock speeds down occasionally to save power.
But it doesn't matter if you've seen one or not. If your device does use a PLL, and doesn't meet this spec, it would be "creative marketing" and a lot of lying to claim that it complies with PCI.
(2) The clock needn't even be periodic. It just needs to meet the min/max cycle time and min pulsewidth requirements. For whatever reason, the system COULD use a gated clock, a logic signal, skip cycles, or whatever.
(3) In 33MHz PCI mode, unlike 66MHz mode, Spread Spectrum dithering is unconstrained. A PLL that works on one system's SSC could fail miserably on another, because the dithering frequency happens to be lower, or it has wider modulation (bigger jitter).
Andy