Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [10GBT] More on the PAM12 emissions



Scott,

The analysis I showed is independent of the power backoff issue. The spiky
PSD problem will occur at any line length if the THP filter coefficients are
small/zero. At 0m, even if we backoff the power by 10dB, who's to say that
the PSD averaging is only going to be done over 0.5ms? In any case, if we
exchange THP filter coefficients a la HDSL2, are you then going to propose a
mechanism to reject THP filters that are "too small", because we have this
spiky PSD problem in the PAM12 encoding?

As far as Dr.Ungerboeck's fixed precoder proposal, the task force has not
agreed to that. Personally, I believe that it ties receivers up to a
specific implementation (specific receive filter, etc.) and it is not
flexible. Therefore, I still prefer the FIR approach that is in our PAM8
proposal and in Jose's presentation.

Regards,
Sailesh Rao.
srao@phyten.com

>From: Scott Powell <spowell@broadcom.com>
>Reply-To: spowell@broadcom.com
>To: STDS-802-3-10GBT@listserv.ieee.org
>Subject: Re: [10GBT] More on the PAM12 emissions
>Date: Mon, 26 Jul 2004 12:31:29 -0700
>
>Sailesh,
>   Are you proposing that we drop power backoff for short cables ?  I
>believe
>both proposals on the table recognize the importance of backing off the Tx
>power in short reach applications.  Also, as Dr. Ungerboeck pointed out
>quite well in ungerboeck_1_0704.pdf, there is really no need to "zero out"
>the THP coefficients.
>
>- Scott
>
>-----Original Message-----
>From: stds-802-3-10gbt@ieee.org [mailto:stds-802-3-10gbt@ieee.org] On
>Behalf
>Of sailesh rao
>Sent: Monday, July 26, 2004 11:18 AM
>To: STDS-802-3-10GBT@listserv.ieee.org
>Subject: [10GBT] More on the PAM12 emissions
>
>
>Hi All,
>
>I'm attaching the corrected glenn.m code with the proper repetition rate
>(4224 symbols). For those of you who don't have access to matlab, I'm also
>attaching the output of this code with 4224 (5us) and 422224 (0.5ms) as pdf
>plots. This illustrates the actual energy buildup we can expect in the
>PAM12
>framing scheme, as depicted on slide 22 of powell_1_0704.pdf.
>
>Another issue with the PAM12 proposal is that the bit-to-symbol mapping on
>slide 24 of powell_1_0704.pdf will cause a 0.47dB increase in the transmit
>PSD power when the THP coefficients are zero'ed out (e.g., at short line
>lengths). In contrast, in the PAM8 proposal, if the THP coefficients are
>zero, the transmit PSD power will decrease by 0.03dB.
>
>Therefore, I would like to officially raise the PAM12 emissions penalty to
>1.3dB from the 0.8dB I had been quoting. In this case, the total EMI
>penalty
>of  PAM12 is between 4.5dB and 5.1dB over the existing cabling
>infrastructure. It is between 3.1dB and 4.5dB for new cabling.
>
>Regards,
>Sailesh Rao.
>srao@phyten.com
>
>_________________________________________________________________
>Planning a family vacation? Check out the MSN Family Travel guide!
>http://dollar.msn.com

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/