Channel Ad Hoc telecon minutes February 22, 2006 Fulvio presented Intel sim results to determine appropriate EIT crosstalk level limits. Charles confirmed that Fulvio's Y axis should be Vpp instead of mVpp. Fulvio recommends to move the ICR toward M20 values, and keep the 15mVpp EIT aggressor level. Charles noted on slide 15, the M20 channel has an extra 11mV of self interference than the ITTC23 channel, and proposed that we add the 11mV of sine amplitude to the EIT channel limit when using the ITTC23 channel to balance the aggressor's effect with different channels. Joe asked that we evaluate 15mVpp limit with possible adder for self interference. Joe presented sim results with ITTC_20dB_returnloss channel. ITTC_20dB channel refers to the return loss, and was selected to match the new tougher return loss spec. As jitter DCD is changed the other jitter components the TJ is kept constant at 0.28UI. Joe used the old cap_like pkg model to reduce variables, the new pkg model is more return loss pessimistic. Thru_worst channel is the six shooter board with bad cables, and represents a "high confidence" channel that fails Joe's sim with any amount of DCD. Adam mentioned his displeasure with the present unintelligent method of determining borderline channel definitions. Charles volunteered to create some physical channels with different loss and reflection characteristics to iterate with Joe's simulator to determine pathological combinations. Rich asked what if the channel crosstalk budget is consumed by the other TX on the chip. EIT value Joe thinks it should be <=25mV Fulvio thinks it may be reasonable to go >20mV but he needs to resimulate with the ITTC20dB channel. TX simulator rise time should be defined. Joe, Charles, and Rich will work on creating channels that border the working space. They will use the email reflector for working discussion, then present in Denver. -Shannon