Re: trace length

[mellitz@eagle.ColumbiaSC.NCR.COM]

Tobias,

Is this the straw that... ?
Take the case where all cards in a system went from 1.5 to 1.7 inches.   There
are certain system design cases that would increase the one way delay by up to
150 ps.  Some systems depend on the this delay and may count it a few times
i.e.  for reflections.  This could add up to 1/2 ns for a few trips around the
block.  So technically it is depended on.  There are certain things that you
could do to make this transparent.  Take a black box approach.  Just make it
electrically equivalent to the spec at the pin.  Controlling or lowering
effective  epsilon_r or load capacitance may be options.

... Richard Mellitz, NCR
On Nov 20, 11:22am, Tobias Stumber wrote:
> Subject: trace length
> Has anyone experiences with pci signal trace lengths which
> are longer than the limit of 1.5 inches ?
> Seems like in our layout some traces will become 1.7 inches
> (if we do not change our whole placement).
> Will this automatically cause problems ? ( I think/hope not,
> but eventually anyone of you knows !)
> 
> Have a nice day
> 					Tobias Stumber
> 					Robert Bosch GmbH
> 					Germany
> 
> mailto:tobias.stumber@fr.bosch.de
> 
>-- End of excerpt from Tobias Stumber

2È¶