Re: Ramp Rates

John,

Regarding quesiton 1, if 5V/2nS is the true ramp rate in compliance with
the NOTES ON DATA DERIVATION METHOD, replacing this with 2.5V/1nS is NOT
equivalent (assuming that the subparameter R_load has not changed) for
several reasons: (1) It would be in violaton of the specified measurement
conditions. (2) A different measurement condition that is not in compliance
with the stated conditions (e.g. a lower resistance load) would most likely
NOT give a 1nS risetime response, but something closer to 2nS. (3) The dV
value is actually redundant since it can be derived from the I/V characteristics
and test load DC conditions directly. If it is in conflict with the derived
value, I would use the derived value for consistency with I/V curve usaage
elsewhere in the IBIS model.

It is my interpretation to assume that dt_r and dt_f are the informative
portions of the [Ramp] specification and are to be used directly. Even
though the [Ramp] is given as a ratio, it should NOT be scaled. However,
I believe the frequency response approximation assumes Tr is the 10-90%
value, so the only scaling would be to multiply dt_r by (.8/.6) to
extrapolate to the 10-90% value, and similarly for dt_f.

Bob Ross
Interconnectix, Inc.

> Dear IBISians,
> 1. Can anyone clarify the convention for the dV/dT values for the ramp
> characteristics. It is of course necessary to extract rise or fall time in
> order to get dV/dT, but is the value for dV necessarily the actual voltage
> transition when the output is being measured with a standard load ? In
> other words, if a device had a 20%->80% transition time of 2nS with a
> transition voltage of 5V, would 2.5V/1nS be allowed as an equally valid
> entry as 5V/2nS. The reason I ask is that when looking at the line
> formula-based in the frequency domain, the upper frequency of concern is
> 0.35/Tr which is a useful figure to be able to derive easily and
> approximately.

> 2. Any information on IBIS model availability, now or in the near future,
> free or for sale, would be most welcome.

> Thanks in advance for your help.

> Best Regards,

> John

> --------------------------------------------------------------------------
> John Berrie, Chief Application Engineer
> Zuken-Redac Limited,
> Green Lane,Tewkesbury,
> Glos. GL20 8HE, England
> Tel: (44) 01684 294161
> Fax:(44) 01684 299754
> E-Mail johnb@redac.co.uk
> --------------------------------------------------------------------------
Received on Mon Mar 20 19:58:18 1995