[IBIS-Users] RE: Question about differential pair buffer

Ramiro,

1)  I think I need to clarify something to you first.
IBIS has many different buffer types.  It can have push/pull
types as well as open_XXX types.  So it is perfectly legal
to have models without pullup, or pulldown IV tables.  However,
in order to get a signal, something will have to pull up or
down in the rest of the circuit, just as in real life, but
this doesn't have to be in the buffer itself.  It could be
a termination resistor or similar device somewhere in the
circuit.

2)  Looking at your drawings, the first one on the top, it
seems that the output transistor is in an emitter follower
configuration, hanging off the top supply rail.  This looks
like a pullup to me.  The two series transistors on the bottom,
with "bias" look like a pulldown.  So you do have both pullup
and pulldown in this buffer.

3)  In terms of IBIS, however, what you need to consider is
what these are doing.  If you have a transistor that is always
on, acting as a terminator, or something similar, you would
need to put its IV data in the clamp tables.  In IBIS terminology
anything that is not switched goes into the clamp tables.  Everything
else that is turned on or off as the buffer changes states will
go into the pullup and/or pulldown tables.  So in your case I would
put those BJT-s on the top into the [Pullup] IV tables, and the
MOSFET-s on the bottom into the [GND Clamp] tables.

4)   The IBIS parser will not give you an error if you don't have
pullup or pulldown tables just because they are missing.  These
are optional keywords in IBIS.  However, you will get errors which
are related.  The Vt curves and IV curves work together, and there
are checks for which they both need to be there in certain configurations.
If you do not satisfy some of these requirements, you will get errors.

5)  Regarding figure 2, I do not understand your labeling.  What is
Vos  (or is it Vds?) stand for?  Is it the supply voltage (Vcc or Vdd)?
However, I have to caution you that you have an extra issue here
because your output transistors are BJT-s.  They tend to influence
each other much more than MOSFET-s, so what you do on one output
will have a relatively strong effect on the other output.  What this
means is that in reality, IBIS should have an I-V surface instead
of an IV curve for these [Pullup] tables, output current as a function
of the voltage of both outputs.  Conventional IBIS cannot do this,
and you may need to find your own compromises if you want to make
models with the normal IBIS keywords, or you will have to find ways
to model this device with the *-AMS extensions of IBIS 4.1.

6)  Regarding figure 3, I wonder about that short across the source...
and also about the labeling I asked you already.

7)  I would also recommend that you look at my presentation on how
to characterize differential buffers at:

http://www.eda.org/pub/ibis/summits/oct03/muranyi.pdf

I hope this will help you in your modeling.

Arpad
========================================================================

-----Original Message-----
From: RODRIGUEZ_LECONA_RAMIRO [mailto:rlecona@siu.buap.mx] 
Sent: Monday, June 13, 2005 6:43 AM
To: Muranyi, Arpad
Cc: ibis-info@eda.org; ibis-users@eda.org
Subject: Re: Question about differential pair buffer

Hi Arpad, 

      Attached I send you the figures that show the buffer schematic and 
procedures I am using to extract the IBIS curves for this differential 
buffer and I have a few questions: 

      If I don't specify the pullup curve, how does  the IBIS simulator know 
how to drive a high state?
      If I don't specify the pullup curve, Will the IBIS parser give error 
messages? because it can't match the pullup curve with the rising waveform.
      Are these the suitable procedures to extract the IBIS curves? 

      Thank you. 

      Regards, 

     Ramiro

|------------------------------------------------------------------
|For help or to subscribe/unsubscribe, email majordomo@eda.org
|with just the appropriate command message(s) in the body:
|
|  help
|  subscribe   ibis       <optional e-mail address, if different>
|  subscribe   ibis-users <optional e-mail address, if different>
|  unsubscribe ibis       <optional e-mail address, if different>
|  unsubscribe ibis-users <optional e-mail address, if different>
|
|or email a written request to ibis-request@eda.org.
|
|IBIS reflector archives exist under:
|
|  http://www.eda.org/pub/ibis/email_archive/  Recent
|  http://www.eda.org/pub/ibis/users_archive/  Recent
|  http://www.eda.org/pub/ibis/email/          E-mail since 1993