Subject: RE: [IBIS-Users] Question about warning message]
From: Tom Dagostino (tom@teraspeed.com)
Date: Tue Nov 19 2002 - 10:48:21 PST
I totally agree about the quality of SPICE models. I've seen too many SPICE
based IBIS models that do not predict the actual behavior of the silicon.
And I'm not just talking about clamps. Too often the pullup or pulldown
measurements made on a die will fall outside the range predicted by SPICE.
So, is -Vcc to 2Vcc enough? Consider the 3.3V, 5 Volt compliant receiver on
the end of an unterminated net with a 5 volt CMOS driver. Seems the part
could see almost +10 Volts.
Tom Dagostino
Device Modeling Division
Teraspeed Consulting Group LLC
2926 SE Yamhill St.
Portland, OR 97214
(503) 430-1065
tom@teraspeed.com
http://www.teraspeed.com
-----Original Message-----
From: owner-ibis-users@server.eda.org
[mailto:owner-ibis-users@server.eda.org]On Behalf Of Fred Balistreri
Sent: Tuesday, November 19, 2002 10:15 AM
To: Lynne Green
Cc: ibis-users@server.eda.org; scott@teraspeed.com
Subject: Re: [IBIS-Users] Question about warning message]
I believe the original reason for testing from -vcc to 2vcc
had to do with transmission line theory. The possible maximum
reflection with an open circuit becomes the 2*voltage of the
driving line given that the reflection coeficient becomes
infinite. Oddly enough once there are diodes and esd circuitry
reaching such voltages becomes difficult. In point of fact
this is what we wish to determine. So while setting these limits
is general and perhaps should be the default it is sometimes
not necessary. In fact often times it is not possible or only
possible through the magic of simulation. As pointed out the
wording in the IBIS spec makes it mandatory. However a part
will pass the parser even if the mandatory voltages are not
achieved. I also am not voluntering to write a bird but perhaps
one is in order.
I do not think the model maker should be fideling with the SPICE
netlist unless they are the ones who know the design. At the
same time I agree with Kim that better attention needs to be
paid to the SPICE models. The IC vendors need to realize what
the application of the model is for. Instead what seems to
happen is that the SPICE model gets thrown over the wall to
a field organization that simply passes it on to the outside
world as is or creates an IBIS model from it.
Best Regards,
Lynne Green wrote:
>
> Hi, Scott,
>
> Having table values that cover -Vcc to 2Vcc is useful for DC convergence
> for beginning a
> transient (time-domain) simulation. Fortunately, SI simulators are
> smart enough to know
> the DC operating point must lie between the reflection limits.
> ("General purpose" simulators
> often do not assume a voltage range, so they can literally search "far
> and wide". This
> makes them more sensitive to numeric overflow in model extrapolation.)
>
> The easiest thing to do is to "clamp" the I-V table values outside the
> range of "reasonable"
> operation. +/-100 Amps is a good limit, or any other value beyond the
> expected currents
> during maximum overshoots.
>
> As noted by previous posters, the correct thing is to add resistance to
> the SPICE diodes
> for both output driver transistors and clamp diodes.
>
> Best regards,
> Lynne
>
> -----Original Message-----
> From: Scott McMorrow [mailto:scott@teraspeed.com]
> Sent: Monday, November 18, 2002 10:31 PM
> To: ibis-users@eda.org
> Subject: Re: [IBIS-Users] Question about warning message]
>
> Fred,
>
> Damn! You actually read the spec and are absolutely correct. This is
> something that really needs to be corrected, as it it quite
> inappropriate for modern devices.
>
> regards,
>
> scott
>
> \
> --
> Scott McMorrow
> Teraspeed Consulting Group LLC
> 2926 SE Yamhill St.
> Portland, OR 97214
> (503) 239-5536
> http://www.teraspeed.com
>
> Fred Balistreri wrote:
>
> The reason IBIS models go from -Vcc to 2Vcc is because that is what it
> says in the spec. The following is a direct quote from the spec,
> immediately
> following the section that describes the testing voltage range:
>
> "These voltage ranges must be spanned by the IBIS data. Data derived
> from
> lab measurements may not be able to span these ranges as such and so may
> need to be extrapolated to cover the full range. This data must not be
> left for the simulator to provide."
>
> If we don't want people to follow the spec then I would argue that the
> spec needs modification. More importantly because of the spec ibis tools
> all
> conform to the voltage testing as prescribed, included s2ibis which a
> lot
> of people use. So I don't think the statement about people blindly
> following
> the voltage requirement is true. It may be more of the case that
> historically
> this is what we have done, which can be traced back to where it all
> started,
> the specification itself.
>
> Best Regards,
>
> "Flora, Matthew" wrote:
>
> Hi All,
>
> The non-monotonic warnings are generated by the IBIS Golden Parser
> because it is
> not summing the I-V table together before checking for monotonicity. I
> believe
> there is a pending bug (71) fix to the IBIS Golden Parser which will sum
> the
> tables before the monotonicity check.
> http://www.vhdl.org/pub/ibis/bugs/ibischk/bug71
>
> It has been argued that the extreme currents are acceptable because a
> part
> pushed to those extremes may actually give extreme current in the
> instant before
> it melts. I believe this is because people are blindly following the
> -VCC to
> 2VCC voltage range recommendation for I-V tables and so are trying to
> come up
> with currents well outside of the normal operating region for some
> parts. I
> don't know if the IBIS spec actually say it, but I believe the intention
> is that
> the I-V tables should include all of the voltages where "interesting"
> things
> happen. If the part was never designed to handle voltages out at -VCC
> or 2VCC,
> then the tables need not include such voltages. I believe this was more
> common
> when more people were doing bench measurements to make the tables and
> they
> limited the voltage range to actual voltages that got close to melt
> down, but
> not beyond.
>
> Just my two pence,
> Matthew Flora
>
>
> -----Original Message-----
> From: Angulo, John
> Sent: Monday, November 18, 2002 10:28 AM
> To: 'ibis-users@eda.org'
> Subject: RE: [IBIS-Users] Question about warning message
>
> The following is posted on behalf of Tom Dagostino at
> Teraspeed Consulting Group. Please send any private replies
> to tom@teraspeed.com only.
>
> ----------------------------------------------------------------
>
> I'll disagree on this. The extreme currents are likely OK but the
> non-monotonic warnings may cause the simulations to not
> converge properly or
> in some simulators cause the model not to load at all. If
> the sum of the IV
> curves do not have monotonic issues you are OK. If the sum
> of the curves do
> have monotonic issues I would investigate before I would say it is OK.
>
> Tom Dagostino
>
> Device Modeling Division
>
> Teraspeed Consulting Group LLC
>
> 2926 SE Yamhill St.
>
> Portland, OR 97214
>
> (503) 430-1065
>
> tom@teraspeed.com
>
> http://www.teraspeed.com
>
> -----Original Message-----
> From: owner-ibis-users@server.eda.org
> [mailto:owner-ibis-users@server.eda.org]On Behalf Of
> Abdulrahman A. Rafiq
> Sent: Friday, November 15, 2002 11:32 PM
> To: Eric Hsu; ibis-users@server.eda.org
> Subject: Re: [IBIS-Users] Question about warning message
>
> Eric,
>
> The warning message of "non-monotonic" is perfectly
> acceptable and can be
> ignored. As for the warning message of extrreme currents
> present, it also is
> acceptable. In fact if you open the .ibs file and comment our all the
> extreme +ve and -ve currents in the I-V tables and then rerun
> ibischks, this
> warning message will not show up.
>
> In short you needn't worry too much about these warnings.
>
> Hope this helps.
>
> Regards,
>
> Abdulrahman Rafiq
>
> Eric Hsu <ehsu@netlogicmicro.com> wrote:
>
> Hi All,
>
> When running ibischk3 to check ver2.1 model, if there are warning
> messages complaining about "non-monotonic"and "extreme
> currents present"
> around hundred amp level, is it possible and acceptible?
>
> Thanks
>
> ~eric
>
> |------------------------------------------------------------------
> |For help or to subscribe/unsubscribe, email majordomo@eda.org
> |with just the appropriate command message(s) in the body:
> |
> | help
> | subscribe ibis
> | subscribe ibis-users
> | unsubscribe ibis
> | unsubscribe ibis-users
> |
> |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
>
> ************************************
> ABDULRAHMAN AK RAFIQ
> Scientific Engineering Consultants, Inc
> Phone: (805) 708-1976
> Email: arafiqsi@yahoo.com
> URL: www.geocities.com/arafiqsi/
> ************************************
>
> --------------------------------------------------------------
> --------------
> --
> Do you Yahoo!?
> Yahoo! Web Hosting - Let the expert host your site
>
> ------=_NextPart_000_008A_01C28D83.9D8DBFE0
> Content-Type: text/html;
> charset="us-ascii"
> Content-Transfer-Encoding: quoted-printable
>
> <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
> <HTML><HEAD>
> <META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
> charset=3Dus-ascii">
>
> <META content=3D"MSHTML 5.50.4916.2300" name=3DGENERATOR></HEAD>
> <BODY>
> <DIV><SPAN class=3D434591623-16112002><FONT face=3DArial
> color=3D#0000ff =
> size=3D2>I'll=20
> disagree on this. The extreme currents are likely OK but the =
> non-monotonic=20
> warnings may cause the simulations to not converge properly
> or in some=20
> simulators cause the model not to load at all. If the
> sum of the =
> IV curves=20
> do not have monotonic issues you are OK. If the sum of
> the curves =
> do have=20
> monotonic issues I would investigate before I would say it is
> OK. =20
> </FONT></SPAN></DIV>
> <DIV> </DIV>
> <DIV class=3DSection1>
> <P style=3D"MARGIN: 0in 0in 0pt"><SPAN=20
> style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Tom Dagostino</SPAN></P>
> <P style=3D"MARGIN: 0in 0in 0pt"><SPAN=20
> style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Device Modeling =
> Division</SPAN></P>
> <P style=3D"MARGIN: 0in 0in 0pt"><SPAN=20
> style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Teraspeed
> Consulting Group =
>
> LLC</SPAN></P>
> <P style=3D"MARGIN: 0in 0in 0pt"><SPAN=20
> style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">2926 SE Yamhill =
> St.</SPAN></P>
> <P style=3D"MARGIN: 0in 0in 0pt"><SPAN=20
> style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">Portland, OR =
> 97214</SPAN></P>
> <P style=3D"MARGIN: 0in 0in 0pt"><SPAN=20
> style=3D"FONT-SIZE: 10pt; FONT-FAMILY: Arial">(503)
> 430-1065</SPAN></P>
> <P style=3D"MARGIN: 0in 0in 0pt"><SPAN=20
> style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
> Arial">tom@teraspeed.com</SPAN></P>
> <P style=3D"MARGIN: 0in 0in 0pt"><SPAN=20
> style=3D"FONT-SIZE: 10pt; FONT-FAMILY: =
> Arial">http://www.teraspeed.com></P></DIV>
> <BLOCKQUOTE>
> <DIV class=3DOutlookMessageHeader dir=3Dltr align=3Dleft><FONT =
> face=3DTahoma=20
> size=3D2>-----Original Message-----<BR><B>From:</B>=20
> owner-ibis-users@server.eda.org =
> [mailto:owner-ibis-users@server.eda.org]<B>On=20
> Behalf Of </B>Abdulrahman A. Rafiq<BR><B>Sent:</B> Friday,
> November =
> 15, 2002=20
> 11:32 PM<BR><B>To:</B> Eric Hsu; =
> ibis-users@server.eda.org<BR><B>Subject:</B>=20
> Re: [IBIS-Users] Question about warning message<BR><BR></FONT></DIV>
> <P>Eric,=20
> <P>The warning message of "non-monotonic" is perfectly
> acceptable and =
> can be=20
> ignored. As for the warning message of extrreme currents
> present, it =
> also is=20
> acceptable. In fact if you open the .ibs file and comment
> our all the =
> extreme=20
> +ve and -ve currents in the I-V tables and then rerun
> ibischks, this =
> warning=20
> message will not show up.=20
> <P>In short you needn't worry too much about these warnings.=20
> <P>Hope this helps.=20
> <P>Regards,=20
> <P>Abdulrahman Rafiq=20
> <P> <B><I>Eric Hsu
> <ehsu@netlogicmicro.com></I></B> wrote:=20
> <BLOCKQUOTE=20
> style=3D"PADDING-LEFT: 5px; MARGIN-LEFT: 5px; BORDER-LEFT:
> #1010ff 2px =
> solid">Hi=20
> All,<BR><BR>When running ibischk3 to check ver2.1 model,
> if there =
> are=20
> warning<BR>messages complaining about "non-monotonic"and
> "extreme =
> currents=20
> present"<BR>around hundred amp level, is it possible and=20
> =
> acceptible?<BR><BR>Thanks<BR><BR>~eric<BR><BR>|---------------
> -----------=
> ----------------------------------------<BR>|For=20
> help or to subscribe/unsubscribe, email
> majordomo@eda.org<BR>|with =
> just the=20
> appropriate command message(s) in the body:<BR>|<BR>| help<BR>| =
> subscribe=20
> ibis <OPTIONAL e-mail address, if different><BR>| subscribe =
> ibis-users=20
> <OPTIONAL e-mail address, if different><BR>| unsubscribe ibis =
> <OPTIONAL=20
> e-mail address, if different><BR>| unsubscribe ibis-users
> <OPTIONAL =
> e-mail=20
> address, if different><BR>|<BR>|or email a written request to=20
> ibis-request@eda.org.<BR>|<BR>|IBIS reflector archives exist=20
> under:<BR>|<BR>| http://www.eda.org/pub/ibis/email_archive/ =
> Recent<BR>|=20
> http://www.! eda.org/pub/ibis/users_archive/ Recent<BR>|=20
> http://www.eda.org/pub/ibis/email/ E-mail since=20
> =
> 1993</BLOCKQUOTE><BR><BR>************************************<
> BR>ABDULRAH=
> MAN=20
> AK RAFIQ<BR>Scientific Engineering Consultants,
> Inc<BR>Phone: (805)=20
> 708-1976<BR>Email: arafiqsi@yahoo.com <BR>URL: =
> www.geocities.com/arafiqsi/=20
> <BR>************************************
> <P><BR>
> <HR SIZE=3D1>
> Do you Yahoo!?<BR><A=20
> =
> href=3D"http://rd.yahoo.com/hosting/mailsig/*http://webhosting
> .yahoo.com"=
>
> Yahoo!=20
>
> Web Hosting</A> - Let the expert host your =
> site</BLOCKQUOTE></BODY></HTML>
>
> ------=_NextPart_000_008A_01C28D83.9D8DBFE0--
> |------------------------------------------------------------------
> |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
>
>
>
>
> --
> Scott McMorrow
> Teraspeed Consulting Group LLC
> 2926 SE Yamhill St.
> Portland, OR 97214
> (503) 239-5536
> http://www.teraspeed.com
> |------------------------------------------------------------------
> |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
-- Fred Balistreri fred@apsimtech.comhttp://www.apsimtech.com |------------------------------------------------------------------ |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 : Tue Nov 19 2002 - 10:57:02 PST