Channel Ad Hoc telecon minutes April 19, 2006 upcoming meetings: 1. Channel Ad Hoc meeting May 3 2. Austin interim May 23-25 full days 3. San Diego plenary July 17-22 presentation/discussions: spagna_c1_0406.pdf key conclusions: 1. ittc_20db channel works with interference up to 12mV 2. sinusoidal aggressor correlates with white noise 3. ittc_20db_returnloss reference channel suitable for EIT 4. relationships found between RL, and ILD (possibly integrals under limit curves) need to be quantified to show how they may help fill in the variables that are being caught by EIT, ILD, ICR, and RL specs. valliappan_c1_0406.pdf key conclusions: 1. With worst case conditions, all channels failed. KR channel failures in order of specs most violated were FA, ICR, IL, ILD, RL 2. With nominal conditions, 50% of the channels passed. mellitz - FA limit was increased for KX4, so slide 6 FA fails may not be accurate. moore - much resistance seen trying to create common spec for KX, KX4 and KR spagna - augmentation of channel definition, backplane designer needs better understanding of salient design points. mellitz - looking into ILD^2 integral between limit line and data to adjust ICR due to RL and self interference abler - this simulation was overly conservative Fulvio presented 24mVpp EIT agressor limit, Howard presented 16mVpp EIT agressor limit. magesh - fulvio's values may be a optomistic, he may have used single sided white noise abler - Are sim conditions adequately defined for all the sim teams? ??? - possibly use crosstalk amplitude relative to TX ampltitude to reduce excessive conservatism dambrosia - channels can be created for simulation that can't occurr in a real system telang_c1_0406.pdf key conclusions: 1. EIT power required is -140dBm/Hz, Noise/Com NC1126A gives -110dBm/Hz, so is adequate sawyer - look into the UFX7128 which goes 10M-10G with +/-3dB flatness moore - Vivek should drive the specs with white noise, including the equipment setup. power levels necessary, and crest factor if it is important. moore - next channel ad hoc meeting in two week's time moore - meeting results show no closure on spec, we need a tighter spec for either the RX or TX or channel