This FAQ is a result of various ‘real life’ questions
that have been received with regards to S2IBIS3. This list is, by no means, a
comprehensive one and is only an effort to assemble questions and problems
users face while trying to get S2IBIS3 working. Please email us if you don’t fine
what you are looking for. If you are looking for S2IBIS3, you can download it here.
A5.
I am getting parser errors related to the rising and falling waveforms like:
“
ERROR - Model PCI33V: The [Rising Waveform]
with [R_fixture]=50 Ohms and [V_fixture]=3.3V
has TYP column DC endpoints of 1.03V and 3.30v, but
an equivalent load applied to the model's I-V tables yields
different voltages ( 1.43V and 3.30V),
a difference of 21.08% and 0.05%, respectively.
AND
ERROR - Model PCI33V: The [Falling Waveform]
with [R_fixture]=50 Ohms and [V_fixture]=0V
has TYP column DC endpoints of 0.00V and 2.17v, but
an equivalent load applied to the model's I-V tables yields
different voltages ( 0.53V and 2.38V),
a difference of 28.48% and 11.26%, respectively.
”
“Running Java Exception in thread "main"
java.lang.NoClassDefFoundError: s2ibis3 java
version "1.4.2_09" Java(TM) 2 Runtime Environment, Standard
Edition (build 1.4.2_09-b05) Java
HotSpot(TM) Client VM (build 1.4.2_09-b05, mixed mode) An
error occured running the Java program. If the Java version above is less than
1.4, please install Java 2, ver 1.4.”
“Exception in thread "main"
java.lang.ArrayIndexOutOfBoundsException: 56at
getSpiceVIData.getSpiceVIData(s2ispice.java:1381) at
generateVICurve.generateVICurve(s2ispice.java:1600) at
analyzePin.analyzePin(s2ianalyz.java:1017) at
analyzeComponent.analyzeComponent(s2ianalyz.java:1431) at
s2ibis3.main(s2ibis3.java:102) java version "1.4.2_02" Java(TM)
2 Runtime Environment, Standard Edition (build 1.4.2_02-b03) Java HotSpot(TM)
Client VM (build 1.4.2_02-b03, mixed mode) An
error occured running the Java program. If the Java version above is less than
1.4, please install Java 2, ver 1.4.”
“Exception in thread "main" java.lang.ClassFormatError: s2ibis3
(Illegal constant pool type)
at java.lang.ClassLoader.defineClass0(Native Method)
at java.lang.ClassLoader.defineClass(Unknown Source)
at java.security.SecureClassLoader.defineClass(Unknown Source)
at java.net.URLClassLoader.defineClass(Unknown Source)
.... ....
at java.lang.ClassLoader.loadClassInternal(Unknown Source)
java version "1.4.2_09"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2_09-b05)
Java HotSpot(TM) Client VM (build 1.4.2_09-b05, mixed mode)
An error occured running the Java program.
If the Java version above is less than 1.4, please install Java 2, ver 1.4.
A1.Non-monotonic behavior appears to be an issue. What corrective course of
action is usually taken to prevent or resolve this concern?
The IBIS standard specifies that the pullup and pulldown curves
contain pullup and pulldown data ONLY, i.e. in the region where the clamp
diodes are active the current due to the clamp diodes must be subtracted from
the pullup/pulldown current. This is where the non-monotonic curves come from
-- at the extremes of the I/V curves where you are subtracting a large diode
current from the combined diode/on-state-IV curve. In practice, a simulator
sums the two currents together (power clamp and pullup or gnd clamp and
pulldown) thereby making the result monotonic.
IBIS simulators like to see smooth monotonic curves. Often times
the non-monotonic points are not in the “active region” of the device and can
be smoothed out without affecting the accuracy of the IBIS file. When the
non-monotonic points appear in the ‘active region’ of the device, then there
can be accuracy issues. It should be noted that monotonic issues in IBIS models
are highly dependent on the actual SPICE netlist. If the device exhibit
non-monotonic behavior, the IBIS models will show non-monotonic behavior.
A2.What is meant by non-monotonic ?
Non-Monotonic behavior is when a
signal goes negative (or positive) relative to its preceding coordinates. It can be also explained as
a change in the sign of the slope during a transition. You
should refer to this document to get a better understanding for monotonic vs
non monotonic behavior. http://www.eda.org/pub/ibis/summits/jan02/bell2.pdf
A3.How can we generate
IBIS Ver - 4.1 models? (is there to any free tool generate Version 4.1 models?)
Currently – I am not aware of any
free tool that would be able to generate 4.1 models – though S2IBIS3 can
generate 3.2 models – you can hand edit these models to make them compatible
with 4.1
A4.What is the difference between IBIS ver - 3.2 and IBIS ver - 4.1 ?
IBIS version 3.2 adds support for
an electrical board description format, multi-staged buffers or buffers that
may use multiple I-V tables and diode transient times, among other features.
IBIS version 4.0 extends the maximum number of points permitted in V-T tables
(1000 Vs 100), supports the inclusion of independent validation data tables and
adds more parameters for expressing “databook” criteria for evaluating buffer
performance. IBIS 4.1 is the new multilingual extension to 4.0. It now
supports external models that includes VHDL AMS, Verilog AMS and SPICE ( 3F5 )
A5.
I am getting parser errors related to the rising and falling waveforms like:
“
ERROR - Model PCI33V: The [Rising Waveform]
with [R_fixture]=50 Ohms and [V_fixture]=3.3V
has TYP column DC endpoints of 1.03V and 3.30v, but
an equivalent load applied to the model's I-V tables yields
different voltages ( 1.43V and 3.30V),
a difference of 21.08% and 0.05%, respectively.
AND
ERROR - Model PCI33V: The [Falling Waveform]
with [R_fixture]=50 Ohms and [V_fixture]=0V
has TYP column DC endpoints of 0.00V and 2.17v, but
an equivalent load applied to the model's I-V tables yields
different voltages ( 0.53V and 2.38V),
a difference of 28.48% and 11.26%, respectively.
”
What could be wrong? What can I do to get rid of these errors and warnings?
Generally, one of the following is
the problem with your model (not in any particular order):
You should look at the output waveforms to see if they have completed transitions.
If not, the simulation times should be increased to allow complete VT tables.
Check what is your R_load set to. Generally, it should be the same as R_fixture.
Similarly, V_fixture should be the same as the terminal voltages of your buffer.
Check that you have specified for HSPICE (or the simulator you are using) to use the
rising/falling waveforms, otherwise it defaults to using only the ramp data.
The keywords for the simulations to use the waveforms for HSPICE are: ramp_fwf = 2
and ramp_rwf = 2 to use 2 rising and 2 falling waveform.
The default values are 0, so it uses none of the waveforms.
The checker matches the endpoints from the rising and falling waveforms with intersection
of the load line (created using V_fixture and R_fixture) and the VI table.
Hence check if your VI table is generated properly. It might be possible that the rising and
falling waveforms are generated accurately but the VI tables are not. This could happen
if there is double counting of the clamp currents in the VI table or if the clamping
diodes are not set up properly in the circuit and are draining more current than expected.
If you are using an old version of the checker, then update your checker as there was a
problem with the old IBIS checker as described here:
B1.Describe the behind
the scene process of translating Spice simulation output files to IBIS format specifications.
s2ibis uses the following scheme
for determining what tables need to be generated for a particular pin:
i.For POWER, GND, and NC pins, no tables are generated.
ii.For each Input pin, s2ibis automatically runs SPICE jobs for [GND_clamp]
and [POWER_clamp] tables.
iii.For each Output (*Not* I/O, 3-state, or Open_drain) pin, s2ibis
generates [Pulldown] and [Pullup] tables only. Clamp simulations are not even
attempted for Output pins. [Ramp] tables are generated by causing the output
to transition while attached to the IBIS- specified load, reading the 20% and 80%
points of the transition, and calculating the slope in between these points.
iv.For each I/O or 3-state pin, s2ibis generates SPICE simulations for
[Pulldown], [Pullup], [GND_clamp], [Power_clamp], and [Ramp] tables.
Generation of the [Ramp] tables uses exactly the same procedure as described
above under Ouput pins. Generation of the [Pulldown], [Pullup], [GND_clamp],
and [Power_clamp] tables involves a current partitioning scheme.
First s2ibis runs the
[GND_clamp] and [Pulldown] simulations. The [Pulldown] table is generated as
follows:
pulldown(-Vcc->2Vcc):subtract
the associated ground clamp table from the pulldown table
The [GND_clamp] table is not
modified.
The [Pullup] and [POWER_clamp]
tables are created using an analogous procedure.
Then s2ibis runs the
[POWER_clamp] and [Pullup] simulations. The [Pullup] table is generated as
follows:
pullup(-Vcc->2Vcc):subtract
the associated power clamp table from the pullup table
The [POWER_clamp] table is not
modified.
B2.It is recommended that the V-I
tables consist of no more than 100 data points. How is this controlled in the
setup file?
The latest version of S2IBIS3
(V1.1) has an improved algorithm.
B3.What is the difference between s2ibis2 & s2ibis3 ?
s2ibis3 is similar to s2ibis2. The
main difference to it is that it is written in JAVA as compared to C and
Lex and Yacc. s2ibis3 has its own command language, which is similar to
the IBIS language in many respects, and can be easily extended to include new
functionality. S2ibis3 also removes a few bugs from s2ibis2. Please let me know
if you have anything particular in mind so I can address that.
B4.Do you know why the
IBIS example file "ex1" has 3 instances of the string "out"
in the first [Pin] statement, when the specification only says there should be
2 instances of that string (pin name and signal name)? It’s a little unclear as
to why the SPICE node name would be distinct from the signal name. It is also
not clear why the spec (ver 3.2) doesn't mention that the Spice Node name is a
necessary part of the [Pin] statement.
Spice node name is what is
mentioned in the spice netlist. This could be different (or same) to what is
described in the data book. To be able to perform simulations, the simulation
engine would need the spice netlist name. In ex 1, it happens that the pin
name, signal name and the node name in the netlist is same. Refer to ex4
(multipow) and you will notice that the signal name and node names for some
pins are different. In the spec, the signal name is given to be RAS0# (for pin
1) but it might be possible that in the netlist, the node is named RAS0 (or
some other name).
Its not a part of the .ibs file
that’s why the spec (ver 3.2) doesn't mention that the Spice Node name is a
necessary part of the [Pin] statement.. It's only a part of the .s2i file. We
need it in the .s2i file because that’s what we use to generate the simulation
files.
B5.Is there a document discussing the .s2i file content?
Its available with the installation files - the name is -
s2ibis3.txt
B6.Do you have an example file that is compatible with Pspice. I’ve looked
through the examples and they appear to be for Hspice or spectre.
I don’t have anything that is
compatible with Pspice. If you do find something, please let me know if you can
share them so I can put on the s2ibis3 website that would help s2ibis3 pspice
users.
B7.When using s2ibis3, what is the
IBIS version output file? Are there any possible compatibility issues if I take
this output and attempt to link it to other files which might be IBIS 4.1?
Currently, s2ibis3 produces IBIS
version 3.2 models. There are no potential compatibility issues if you want to
hand edit it to be able to introduce some version 4.1 features in the model.
(as long as they are correct)
B8.I have CYGWIN. Would I be able to
run S2IBIS3 on it?
S2IBIS3 has not been tested with
CYGWIN. You might face issues when trying to configure it with CYGWIN. You can
instead try to run S2IBIS3 on windows (XP) in much the same way. Please refer
to the readme.txt file for more information
B9.Can s2ibis3 be used to generate version 2.1 IBIS data?
Yes. Currently S2IBIS3 can create IBIS models ver 3.2 or below.
If you intend to make ver 2.1 models, simply do not include keywords that are ver 3.2 (or higher) specific,
and make sure that you specify in the .s2i setup file that you want ver 2.1 model.
B10.I am wondering if s2ibis3 support linux platform?
S2IBIS3 has not been tested with
linux. It has only been specifically tested only on unix and Windows XP machine. Having said that,
many users have reported that they have been able to use S2IBIS3 on linux. All you need is a linux compatible
spice simulation engine and Java.
B11.Is there a sample .s2i file that we can look at?
Yes. There are several. A sample .s2i file can be viewed here. There are
several others included in the example directories included in the s2ibis3 package.
B12.Is there a real life example of how to use s2ibis3?
Yes. There are several. A sample s2ibis3 run on a Micron SDRAM part
(MT48LC32M16A2TG) can be downloaded here. The spice models can be
downloaded from Micron website here or saved locally here. There are
several other examples included in the example directories in the s2ibis3 package.
“Running Java Exception in thread "main"
java.lang.NoClassDefFoundError: s2ibis3 java
version "1.4.2_09" Java(TM) 2 Runtime Environment, Standard
Edition (build 1.4.2_09-b05) Java
HotSpot(TM) Client VM (build 1.4.2_09-b05, mixed mode) An
error occured running the Java program. If the Java version above is less than
1.4, please install Java 2, ver 1.4.”
error. What’s wrong? I have the correct version
of java installed.
You need to use the “-root” command
to tell s2ibis3 where your class files are.The “-root “command specifies the s2ibis3
executable (that are in the directory /java where you have installed s2ibis3)
Hence, if "../../s2i_install_dir/java" is where your s2ibis3 class
files are located, type: “s2ibis3 –root
../../s2i_install_dir -s2ibis3 buffer.s2i” (without the quotes) to tell s2ibis3
where the “/java” directory is located.
C2. I am
getting the
“Exception in thread "main"
java.lang.ArrayIndexOutOfBoundsException: 56at
getSpiceVIData.getSpiceVIData(s2ispice.java:1381) at
generateVICurve.generateVICurve(s2ispice.java:1600) at
analyzePin.analyzePin(s2ianalyz.java:1017) at
analyzeComponent.analyzeComponent(s2ianalyz.java:1431) at
s2ibis3.main(s2ibis3.java:102) java version "1.4.2_02" Java(TM)
2 Runtime Environment, Standard Edition (build 1.4.2_02-b03) Java HotSpot(TM)
Client VM (build 1.4.2_02-b03, mixed mode) An
error occured running the Java program. If the Java version above is less than
1.4, please install Java 2, ver 1.4.”
error. (the line
numbers could be different if you are running different versions of the
program). What could be wrong? I am running the correct version of java.
This error could be caused because
the VI data from the spice run could not be collected neatly. If you are
observing that there are .out files created by the run, it is possible that
there are more than 1 VI tables in one or more .out files. This could be
because of the fact that you have 2 .temp statements in your spice files.
Remove one of the .temp statement, run s2ibis3 again.
C3.I am using
Windows XP. I am having the same error as described in C1 above. What should I
do?
You should be getting the errors as
shown below. Use the “-root” command with s2ibis3.bat (as desribed in C1 above)
to resolve your problem.
C4. I am
getting the
“Exception in thread "main" java.lang.ClassFormatError: s2ibis3
(Illegal constant pool type)
at java.lang.ClassLoader.defineClass0(Native Method)
at java.lang.ClassLoader.defineClass(Unknown Source)
at java.security.SecureClassLoader.defineClass(Unknown Source)
at java.net.URLClassLoader.defineClass(Unknown Source)
.... ....
at java.lang.ClassLoader.loadClassInternal(Unknown Source)
java version "1.4.2_09"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2_09-b05)
Java HotSpot(TM) Client VM (build 1.4.2_09-b05, mixed mode)
An error occured running the Java program.
If the Java version above is less than 1.4, please install Java 2, ver 1.4.
”
What could be wrong? I am running the correct version of java.
This error is not related to s2ibis3 files. This problem might be because you downloaded the
s2ibis3 files as ASCII files rather than binary files as described here. For verification purposes, the byte size of the latest version of s2ibis3 (s2ibis3_v1_1.zip) is 226,063 bytes
C5.We always generated IBIS models previously
using s2ibis2 and I'm migrating to s2ibis3. Under the old s2ibis2, we used a line in our
.s2i file:
[Spice command] spice3_opt %s %s
In s2ibis2, the first %s becomes each .spi file (i.e. x001.spi) and the 2nd %s becomes each .out file (i.e. x001.out).
However, s2ibis3 just hangs there when it gets to this line. How can I use my own command to run in s2ibis3?
The s2ibis2 command:
"hspice %s >%s 2>%s"
is in C format. S2ibis3 is written in Java - hence you would need to put this in JAVA format.
You can have (for example)
or any other spice command that works for you. I would also try to comment out the [spice command] line from the .s2i file and check to see if the
default works.
C6.
Will s2ibis3.csh run for java version above 1.4.X or does it only work only for 1.4.X.
I tried installing S2IBIS3 on a linux machine with java version 1.5.0 and I am getting Error
"The installed Java is not version 1.4.X. Please install Java 2, ver 1.4."
S2IBIS3.csh has been modified to work with JAVA 1.5.X. Please download the new version of s2ibis3.csh from
here.
