Hello,
Curtis (Thanks!) noticed a logical mistake in that if a redriver TX consumes
the past, explicit convolution with that should not be necessary, which I fixed
in associated slides. This makes the upstream IR a lot simpler for final RX. I
also added a comment about why passing a delta-only impulse response may be too
restricting for a nonlinear redriver TX. Both are on slide 3.
Regards,
Alaeddin.
Attachment:
Symmetric Processing of IBIS AMI Redrivers and Handling AMI_Init only Models with Errata and Comments - Slide 3.pdf
Description: Symmetric Processing of IBIS AMI Redrivers and Handling AMI_Init only Models with Errata and Comments - Slide 3.pdf