Re[2]: BIRD 35.2 - MULTI-STAGED OUTPUTS vs. V-t table

From: Bob Ross <bob@icx.com>
Date: Thu Nov 07 1996 - 10:58:00 PST

To All:

Thanks Arpad for your comment. I had actually replied to Dileep's
question below but forgot to copy to the reflector or to myself. So
I am reconstructing a similar response.

Multi-staged switching [Driver Schedule] keyword schedules other [Models].
These scheduled [Models] have all the properties needed to function
independently including their own V/T tables. In fact, that would
probably be the recommended for models that are scheduled to provide
more predictable results for different simulators. So in this sense,
the [Driver Schedule] is in addition to the [Rising Waveform] and
[Falling Waveform] tables because it calls [Model]s which contain
these tables, and the composite response is the summation of the
individual staged responses of each [Model].

For downward compatibiltiy and consistency, the [Model] under which
[Driver Schedule] is positioned needs to still be a functioning
model with overall I/V tables and [Ramp] and possible V/T characteristics.
In this sense the [Driver Schedule] is an alterative to the V/T
tables and V/I tables contained under this [Model] keyword
because these tables are not used. So there is no conflicting
information. However, simulators which do not support [Driver Schedule]
or need it for a particular application could still use the [Model] if the
[Driver Schedule] keyword and table are removed or ignored.

Bob Ross
Interconnectix

> Date: Thu, 07 Nov 96 08:27:00 PST
> From: Arpad Muranyi <Arpad_Muranyi@ccm.fm.intel.com>
> Message-ID: <Thu, 07 Nov 96 08:31:38 PST_8@ccm.fm.intel.com>
> To: ibis@vhdl.org
> cc: dileep@contec.Apsimtech.COM
> Subject: Re[2]: BIRD 35.2 - MULTI-STAGED OUTPUTS vs. V-t table
> Status: R

> Text item:

> Dileep and Bob,

> I believe the main reason this BIRD was written was because Jon Powell showed us
> a few cases for which the V-t curves are NOT adequate. I am only including
> short excerpts from his description here. (I am shure he would be willing to
> send the full document to anyone interested).

> 1) Decreased drive capability. In this example the driver has two output
> stages. The first stage is High Current Low Voltage (100 mA at 3 V), the second
> stage is lower driver to 5 V.

> 2) TTL-CMOS staged output. In this example the driver has two staged outputs,
> the first is a standard TTL and the secone is standard 5 V CMOS.

> Arpad
> ================================================================================

> Hi Bob,
> Thanks for filling us in on the background.

> >I would expect this method to be used with models that
> >have [Rising Waveform] and [Falling Waveform] tables to
> >better control the individual shapes.

> My understanding was that the multi-stage switching specification
> was an ALTERNATE and may be a more detailed way of specifying the
> waveforms. Your statement seems to imply that the multi-stage
> specification is in ADDITION to the waveform specification.
> If that is the case, I do not see how one can utilize both
> pieces of information and reconstruct a waveform in the simulator
> and also I do not see how one can develop a model to isolate
> these two effects.

> Dileep Divekar
Received on Thu Nov 7 11:07:08 1996

This archive was generated by hypermail 2.1.8 : Fri Jun 03 2011 - 09:52:29 PDT