long "tails" to rising/falling waveform

Hi all,

We are currently facing a problem with rising and falling waveforms. In
some models, there are long "tails" to rising/falling waveform, which when
simulated in ICX cause massive amplitude fluctuations. In order to
alleviate this problem, we took out the rising/falling ramp tables and
relied entirely on dv/dt information. The results were better but we
suspect contain more optimism than the real world. This brought us to the
issue of modifying the ramp tables such that the tails disappear. But by
doing that, we may skew the ramp relative to clamps and pull-up/down data.
Which we suspect would cause even bigger problems.

To all who may have dealt with this problem, what is the best way to modify
the ramp tables? Secondly, how desirable is it to just rely on the dv/dt
information. Dv/dt gives us just two points along the ramp. Hence, could
be intuitively less desirable than rising/falling waveforms. We noticed
that the models provided by Mentor Graphics have just the dv/dt info and not
ramp tables. Would't this give us more optimism than may exist in real
life?

Thanks,

 
Received on Mon Aug 28 10:21:59 2000