Subject: RE: [IBIS-Users]DC Sweep for the pullup IV curve
From: Muranyi, Arpad (arpad.muranyi@intel.com)
Date: Wed Aug 20 2003 - 08:55:41 PDT
Allen,
Even though the IBIS spec requires those ranges you mention,
it does not forbid to provide the entire -Vcc to 2Vcc range
for the clamps. In my IBIS models I always give the whole
range, but I also make sure that I am not double counting for
some effects (on die terminations, etc).
Also, when you measure the data, you will have to sweep it
the full range anyway, because you have to provide a difference
curve for the pullup and pulldown tables. In order be able to
subtract the clamp curve from the pullup or pulldown, you have
to have the whole range covered by both. If you have the data,
you might as well put it in the IBIS model too.
Arpad
===============================================================
-----Original Message-----
From: Allen Chao [mailto:achao@synopsys.com]
Sent: Monday, August 18, 2003 4:00 PM
To: Muranyi, Arpad; ibis-users@eda.org
Subject: RE: [IBIS-Users]DC Sweep for the pullup IV curve
Hi Arpad,
You are right, the current source is composed by 2 pull-up transistors in
series.
Now, for the clamping IV curves, do we always sweep from -Vcc to Vcc (for
gnd clamp) and sweep from Vcc to 2Vcc(for power clamp) at the high-impedance
output? or there are other ways to do it? 'cause our output has 3 modes, low
speed, full speed and high speed, however there is only one mode (stand-by)
where the output is in high-impedance. In different modes(LS/FS/HS), there
are different circuitries connected to the output, using the above method,
we will get the same power clamp/gnd clamp IV curves for all 3 modes. I am
not sure if this is right. Also, I think you are right about the parasitic
diodes being the clamping devices. (Since we don't really have a specific
diode for clamping and there is still some clamping effect in high-impedance
state.)
Thanks,
Allen
-----Original Message-----
From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On
Behalf Of Muranyi, Arpad
Sent: Monday, August 18, 2003 12:40 PM
To: ibis-users@eda.org
Subject: RE: [IBIS-Users]DC Sweep for the pullup IV curve
Allen,
OK, there is no pulldown, because the external resistor pulls the
signal to aero volts. But what is that "current source" made out
of? The designs I have seen so far implement current sources with
transistors, except they are sized and biased in such a way that
their IV curve will look more like a current source.
What does a current source's IV curve look like? By definition a
current source provides a constant current regardless of what the
voltage across its terminals is, i.e. the IV curve is a horizontal
line. In theory this should extend to all voltages, however, in
real life (unless it is a specialized circuit design) it will be
polarized (i.e. work only for one voltage polarity, say positive).
So implementing a current source can be done by a huge transistor,
which has a very steep linear region (coming out of the origin),
and then using a fairly low bias voltage you can operate in its
saturation region, which looks like a constant current source.
Note, however, that this transistor is still connected to some
sort of a supply rail.
So regarding your buffer, I would guess that it has this kind of
a pullup transistor, and whether the designer wants to call it a
pullup or not is a different question.
Having said all this, you can make an IBIS model of this buffer
using the "open source" model type, which has a pullup, but no
pulldown IV curve. Don't worry about it being a current source,
the shape of the IV curve will describe that very well.
Regarding clamps, I have a hard time to believe that the above
transistor doesn't have any parasitic diodes (unless it is not
MOSFET) and that there is absolutely no ESD protection in this
device. It may be the case, but usually it is not. So I would
expect that you would still see some currents outside the rails
when you 3-state the device for this reason. These currents
will go into the IBIS clamp IV tables.
Arpad Muranyi
Intel Corporation
=================================================================
-----Original Message-----
From: Allen Chao [mailto:achao@synopsys.com]
Sent: Monday, August 18, 2003 8:04 AM
To: Muranyi, Arpad; ibis-users@eda.org
Subject: RE: [IBIS-Users]DC Sweep for the pullup IV curve
Hi Arpad,
Thanks for your fast reply... I just consulted further with the buffer
designer here. He said that this output is a high-speed data output and
instead of using Vcc (power-supply voltage) for the power clamp diode and
the pull-up transistor, we use a current source to drive the output to 0.4v
for signal high and simply turn off the transistors between current source
and the output for the low signal. (since the output is also connected to
another resistor and ground, the output will be grounded for the low
signal). As I described, there are not really pullup/pulldown transistors
and clamping devices.
My question is, can we still create an IBIS model for this output?
Thanks in advance,
Allen Chao
|------------------------------------------------------------------
|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
|------------------------------------------------------------------
|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
This archive was generated by hypermail 2b28 : Wed Aug 20 2003 - 08:59:45 PDT