Here is a shotgun blast of constructed channels to look at possible ways of
specifying channel ripple vs channel loss.

Each channel name includes 2 numbers.  The first is 100x the mTC value which
would be computed using the method given in 69A.  The second is 100x the
magnitude of the worst case ILD relative to ILDmax and ILDmin.  Thus 
m_82_ripple_90.s4p has attenuation near the limit and ripple very near the
limit, and m_40_ripple_29 represents a short channel with near minimal ripple.

The channels were constructed by merging S-parameters measured on the Avago
Technologies Six-shooter board with ideal transmission line interconnect with
characteristic impedances which differ significantly from 50 Ohm.  

The merging was done using a proprietary Avago Technologies spice simulator. 
The method has been checked against measurement and found to be as accurate as 
the measurements on which it was based.

These channels meet all the specs given in annex 69B except equation 69B-6,
fitted attenuation less than Amax.  Many of the channels fail that 
equation even though IL < Amax at all frequencies.  See for instance 
m_40_ripple_29.


2006 February 22 Charles Moore
