[SI-LIST] Re: PCI Express AC Coupling
- From: Chris Johnson <cjohnson@xxxxxxx>
- To: Joseph.Schachner@xxxxxxxxxx
- Date: Thu, 11 Mar 2010 11:59:34 -0500
Thanks to all who replied, especially Joe. I'm designing a Virtex-5
based PCI Express card, and I intend to use Xilinx's canned PCI Express
IP. I'm trying to avoid getting into the really low level bus details
as much as possible, since the interface is just a small part of the
design. I will of course make sure that all of the differential pairs
are correctly impedance controlled, and now I know how to handle the AC
coupling.
If anyone has any experience with the Xilinx LogiCORE Endpoint Block
Plus IP, I'd love to hear any caveats you might have.
Chris
Joseph.Schachner@xxxxxxxxxx wrote:
> >From the PCI Express Base Spec rev 2.1, table 4-9 "2.5 and 5.0 GT/s
> Transmitter Specifications":
>
> C TX AC Coupling Capacitor 75 (min) 200 (max) nF All Transmitters
> shall be AC coupled. The AC coupling is required either within the media
> or within the transmitting component itself.
>
> and, section 4.3.5.1 says:
> 4.3.5.1. Channel AC Coupling Capacitors
> Each Lane of a PCI Express Link must be AC coupled. The minimum and maximum
> values for the
> capacitance is given in Table 4-9. Capacitors must be placed on the
> Transmitter side of an interface
> that permits adapters to be plugged and unplugged. In a topology where
> everything is located on a
> 5 single substrate, the capacitors may be located anywhere along the
> channel. External capacitors are
> assumed because the values required are too large to feasibly construct
> on-chip.
>
>
> Related: in section 4.3.6.3 is places an upper limit on channel
> capacitance to ground.
>
> Those are the only occurrences of the string "AC coupl" in the base spec.
>
> So, the AC coupling is not required at both ends of a lane, it is required
> at the TX side of a line. Therefore there is no guarantee that RX (from
> card to motherboard) is AC coupled on the motherboard. There is only a
> guarantee that TX (from motherboard to card) is AC coupled on the
> motherboard. The CEM spec also says this clearly (in 4.6.1).
>
> About the clock: in Gen 1, the only specs on the clock were in the Card
> Electromechanical (CEM) spec. In Gen2, there are lots of timing specs on
> the clock in the base spec, but not voltage - which is the same as for
> Gen1. So, looking in PCI EXPRESS CARD ELECTROMECHANICAL SPECIFICATION, REV
> 2.0, we find:
>
> 2.1. Reference Clock
> 2.1.1. Low Voltage Swing, Differential Clocks
> To reduce jitter and allow for future silicon fabrication process changes,
> low voltage swing,
> differential clocks are being used, as illustrated in Figure 2-1. The
> nominal single-ended swing for
> each clock is 0 to 0.7 V and a nominal frequency of 100 MHz ±300 PPM. The
> clock has a defined
> crossover voltage range and monotonic edges through the input threshold
> regions as specified in
> Chapter 4.
>
> and in Table 2-1 "REFCLK DC Specifications and AC Timing Requirements" it
> says
>
> VCROSS Absolute crossing point voltage +250 (min) +550 (max) mV
> VMAX Absolute Max input voltage +1.15 V
> VMIN Absolute Min input voltage - 0.3 V
>
> So: the clock is not AC coupled. It is guaranteed not to be AC coupled on
> the motherboard.
>
> Chris, I don't know what you need this info for, but I think chances are
> you actually need to read the PCI Express Base specification and the PCI
> Express CEM spec. They can be downloaded from here:
> http://www.pcisig.com/specifications/pciexpress/
>
> --- Joe S.
>
>
>
>
> From: "Michael Rose" <mrose@xxxxxxxxxxxx>
>
>
>
> To: "Chris Johnson" <cjohnson@xxxxxxx>, <si-list@xxxxxxxxxxxxx>
>
>
>
> Date: 03/10/2010 05:49 PM
>
>
>
> Subject: [SI-LIST] Re: PCI Express AC Coupling
>
>
>
> Sent by: si-list-bounce@xxxxxxxxxxxxx
>
>
>
>
>
>
>
>
> Chris, I thought PCIe required AC coupling on both ends(?). Generally, I
> like to place the coupling caps on the RX end since the channel
> attenuates the slew rate making the impedance discontinuity a bit less
> severe. In some cases, AC coupling on both ends has the advantage of
> tolerating a short circuit through the interconnect without damaging the
> driver.
>
> Mike
>
> -----Original Message-----
> From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
> On Behalf Of Chris Johnson
> Sent: Wednesday, March 10, 2010 4:37 PM
> To: si-list@xxxxxxxxxxxxx
> Subject: [SI-LIST] PCI Express AC Coupling
>
> Could someone tell me if any of the TX (from motherboard to card), RX
> (from card to motherboard), and REFCLK (from motherboard to card) PCI
> Express differential pairs are guaranteed to be AC coupled on the
> motherboard?
>
> Also, is there significant signal integrity downside to AC coupling on
> the plug in card, when the signals are already AC coupled on the
> motherboard, assuming that the AC coupling is right at the I/Os to the
> chip on the plug in card?
>
> Thanks,
>
> Chris
>
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>
> or to administer your membership from a web page, go to:
> //www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
>
> List technical documents are available at:
> http://www.si-list.net
>
> List archives are viewable at:
> //www.freelists.org/archives/si-list
>
> Old (prior to June 6, 2001) list archives are viewable at:
> http://www.qsl.net/wb6tpu
>
>
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>
> or to administer your membership from a web page, go to:
> //www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
>
> List technical documents are available at:
> http://www.si-list.net
>
> List archives are viewable at:
> //www.freelists.org/archives/si-list
>
> Old (prior to June 6, 2001) list archives are viewable at:
> http://www.qsl.net/wb6tpu
>
>
>
>
>
>
>
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
