From: owner-ibis-users@ (ibis-users) To: ibis-users-digest@eda.org Subject: ibis-users V1 #28 Reply-To: Sender: owner-ibis-users@ Errors-To: owner-ibis-users@ Precedence: bulk ibis-users Friday, August 15 2003 Volume 01 : Number 028 ---------------------------------------------------------------------- Date: Fri, 1 Aug 2003 15:20:09 -0700 From: "Abril, Giovanni" Subject: RE: [IBIS-Users] C_comp value in IBIS model This is a multi-part message in MIME format. - ------_=_NextPart_001_01C3587B.11D2CF3A Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Thanks Weston. =20 Is there a way for me to vary the capacitance in accordance with the = spec then? The way you explained it, I should have the same C_comp = value for typ, min and max, an extracted value, right? I'll assume ICX = does it this way, since you work for Mentor. If the measured value = turns out to be say 6pF, how would I cover a spec of 2.5pF? - -----Original Message----- From: Beal, Weston [mailto:weston_beal@mentorg.com] Sent: Friday, August 01, 2003 3:16 PM To: 'Kim Helliwell'; Abril, Giovanni Cc: ibis-users@eda.org Subject: RE: [IBIS-Users] C_comp value in IBIS model The effect of C_comp is only double counted if you are using a = substandard simulator. All the major simulators that I know of expect = C_comp to be specified with the same (static approximation) value as = what is included in the waveforms. Somewhere in the simulator this is = reconciled. Usually it is done is such a way that the capacitive effect = is calculated out of the waveforms and then the lumped capacitor is = present at the pin during the transient simulation.. =20 As Arpad noted, the capacitive effect is usually impossible to remove = from the SPICE models used to derive IBIS data. It is part of the = transistor models. Arpad and others have proposed methods of finding = the best static value of the die capacitance to use for C_comp and this = value needs to be there. If you put the value of 0 for C_comp then the = simulator will try to back out the capacitance and put no capacitance at = the pin. At first this seems OK because the waveforms still have all = the capacitive effect. The problem appears when a reflected signal hits = this pin or if this pin is hi-Z. There would be no capacitive loading = and you would then have invalid simulations. =20 C_comp is your friend. Embrace it. ( and enhance it ) =20 Regards, Weston =20 - -----Original Message----- From: Kim Helliwell [mailto:kimgh@apple.com] Sent: Friday, August 01, 2003 2:45 PM To: Abril, Giovanni Cc: ibis-users@eda.org Subject: Re: [IBIS-Users] C_comp value in IBIS model If the waveforms already include the effects of C_comp, putting=20 the C_comp value in the IBIS model will double-count the effect=20 of C_comp, which is probably not what you want.=20 I feel certain that the right thing to do is re-run the spice sims=20 with C_comp removed from the simulations, and put the correct=20 typ/min/max values of C_comp directly in the IBIS model=20 If I were your customer trying to use this model, I'd be really unhappy=20 to find that C_comp was (silently) being double counted. And I'd=20 be suspicious of your model if C_comp were 0 with no explanation.=20 At the very least, if you don't want to re-run things, you should set=20 C_comp to 0 but put in a comment that the reason it's zero is that=20 the effects of C_comp are included in the rising/falling waveforms.=20 But this is definitely a distant second-best, in my opinion.=20 Kim=20 On Friday, August 1, 2003, at 02:13 PM, Abril, Giovanni wrote:=20 Hello IBIS gurus,=20 I've created an IBIS model using a spice netlist from the vendor, and = I'm wondering what to do with C_comp. My rising and falling waveforms = include the effect of C_comp, but what should I put in the IBIS model? = I have a spec sheet that says min C_comp is 2.8pF, max is 10pF. = Currently I have 5.0pF for typ, 2.8pF min and 10pF max.=20 Thanks,=20 Giovanni=20 Kim Helliwell=20 Apple Computer=20 kimgh@apple.com=20 408 974 9936=20 - ------_=_NextPart_001_01C3587B.11D2CF3A Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Thanks=20 Weston.
 
Is=20 there a way for me to vary the capacitance in accordance with the spec=20 then?  The way you explained it, I should have the same C_comp = value for=20 typ, min and max, an extracted value, right?  I'll assume ICX does = it this=20 way, since you work for Mentor.  If the measured value turns out to = be say=20 6pF, how would I cover a spec of 2.5pF?
-----Original Message-----
From: Beal, Weston=20 [mailto:weston_beal@mentorg.com]
Sent: Friday, August 01, = 2003 3:16=20 PM
To: 'Kim Helliwell'; Abril, Giovanni
Cc:=20 ibis-users@eda.org
Subject: RE: [IBIS-Users] C_comp value in = IBIS=20 model

The=20 effect of C_comp is only double counted if you are using a substandard = simulator.  All the major simulators that I know of expect C_comp = to be=20 specified with the same (static approximation) value as what is = included in=20 the waveforms.  Somewhere in the simulator this is = reconciled. =20 Usually it is done is such a way that the capacitive effect is = calculated out=20 of the waveforms and then the lumped capacitor is present at the pin = during=20 the transient simulation..
 
As=20 Arpad noted, the capacitive effect is usually impossible to remove = from the=20 SPICE models used to derive IBIS data.  It is part of the = transistor=20 models.  Arpad and others have proposed methods of finding the = best=20 static value of the die capacitance to use for C_comp and this value = needs to=20 be there.  If you put the value of 0 for C_comp then the = simulator will=20 try to back out the capacitance and put no capacitance at the = pin.  At=20 first this seems OK because the waveforms still have all the = capacitive=20 effect.  The problem appears when a reflected signal hits this = pin or if=20 this pin is hi-Z.  There would be no capacitive loading and you = would=20 then have invalid simulations.
 
C_comp is your friend.  Embrace it.  ( and enhance = it=20 )
 
Regards,
Weston
 
-----Original Message-----
From: Kim Helliwell=20 [mailto:kimgh@apple.com]
Sent: Friday, August 01, 2003 = 2:45=20 PM
To: Abril, Giovanni
Cc:=20 ibis-users@eda.org
Subject: Re: [IBIS-Users] C_comp value = in IBIS=20 model

If the waveforms already include the effects of C_comp, putting =

the C_comp value in the IBIS model will double-count the effect =

of C_comp, which is probably not what you want.


I feel certain that the right thing to do is re-run the spice = sims

with C_comp removed from the simulations, and put the correct =

typ/min/max values of C_comp directly in the IBIS model


If I were your customer trying to use this model, I'd be really = unhappy=20

to find that C_comp was (silently) being double counted. And I'd =

be suspicious of your model if C_comp were 0 with no explanation. =


At the very least, if you don't want to re-run things, you should = set=20

C_comp to 0 but put in a comment that the reason it's zero is = that

the effects of C_comp are included in the rising/falling = waveforms.=20


But this is definitely a distant second-best, in my opinion. =


Kim


On Friday, August 1, 2003, at 02:13 PM, Abril, Giovanni wrote: =


Hello IBIS gurus,


I've created an IBIS model using a = spice netlist=20 from the vendor, and I'm wondering what to do with C_comp.  My = rising=20 and falling waveforms include the effect of C_comp, but what should = I put in=20 the IBIS model?  I have a spec sheet that says min C_comp is = 2.8pF, max=20 is 10pF.  Currently I have 5.0pF for typ, 2.8pF min and 10pF=20 max.


Thanks,


Giovanni



Kim Helliwell

Apple Computer

kimgh@apple.com

408 974 9936


- ------_=_NextPart_001_01C3587B.11D2CF3A-- |------------------------------------------------------------------ |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 ------------------------------ Date: Fri, 1 Aug 2003 15:52:42 -0700 From: Kim Helliwell Subject: Re: [IBIS-Users] C_comp value in IBIS model - --Apple-Mail-8--133216226 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=ISO-8859-1; format=flowed OK, I sit corrected. Sorry to Giovanni for the misinformation... Kim On Friday, August 1, 2003, at 03:05 PM, Muranyi, Arpad wrote: > Kim, > =A0 > I hate to argue with you, but your response is quite wrong. > =A0 > It is not possible, or the least very hard to remove the effects > of die capacitance from a SPICE model.=A0 So the waveforms that > were generated from a SPICE simulation will always include the > effects of C_comp.=A0 Also, putting such a waveform and a C_comp > value into an IBIS model will not double count C_comp.=A0 This > is actually how IBIS simulators expect to get the data, because > they are the ones who do the compensation to avoid double counting. > =A0 > So Giovanni did exactly the right thing. > =A0 > The only thing he needs to watch out for is that if he does a > SPICE vs. IBIS correlation, he should have the value for C_comp > in the IBIS model that was measured on the SPICE model, and not > the data sheet number, because the two may not be the same value. > For correlation work.=A0 However, if the IBIS model needs to be > according to the data book, using the data book values is the right > thing to do.=A0 (Remember, the SPICE model may not have all the > layout related extra capacitances in the circuit, which is the > case most often when someone has a pre layout model). > =A0 > Arpad Muranyi > Intel Corporation > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > =A0 > -----Original Message----- > From: Kim Helliwell [mailto:kimgh@apple.com] > Sent: Friday, August 01, 2003 2:45 PM > To: Abril, Giovanni > Cc: ibis-users@eda.org > Subject: Re: [IBIS-Users] C_comp value in IBIS model > > If the waveforms already include the effects of C_comp, putting > the C_comp value in the IBIS model will double-count the effect > of C_comp, which is probably not what you want. > > I feel certain that the right thing to do is re-run the spice sims > with C_comp removed from the simulations, and put the correct > typ/min/max values of C_comp directly in the IBIS model > > If I were your customer trying to use this model, I'd be really = unhappy > to find that C_comp was (silently) being double counted. And I'd > be suspicious of your model if C_comp were 0 with no explanation. > > At the very least, if you don't want to re-run things, you should set > C_comp to 0 but put in a comment that the reason it's zero is that > the effects of C_comp are included in the rising/falling waveforms. > > But this is definitely a distant second-best, in my opinion. > > Kim > > On Friday, August 1, 2003, at 02:13 PM, Abril, Giovanni wrote: > > Hello IBIS gurus, > > I've created an IBIS model using a spice netlist from the vendor, and=20= > I'm wondering what to do with C_comp.=A0 My rising and falling = waveforms=20 > include the effect of C_comp, but what should I put in the IBIS=20 > model?=A0 I have a spec sheet that says min C_comp is 2.8pF, max is=20 > 10pF.=A0 Currently I have 5.0pF for typ, 2.8pF min and 10pF max. > > Thanks, > > Giovanni > > > Kim Helliwell > Apple Computer > kimgh@apple.com > 408 974 9936 > > Kim Helliwell Apple Computer kimgh@apple.com 408 974 9936 - --Apple-Mail-8--133216226 Content-Transfer-Encoding: quoted-printable Content-Type: text/enriched; charset=ISO-8859-1 OK, I sit corrected. Sorry to Giovanni for the misinformation... Kim On Friday, August 1, 2003, at 03:05 PM, Muranyi, Arpad wrote: Courier = NewKim, =A0 Courier NewI hate to argue with you, but your response is quite = wrong. =A0 Courier NewIt is not possible, or the least very hard to remove the = effects Courier Newof die capacitance from a SPICE model.=A0 So the waveforms = that Courier Newwere generated from a SPICE simulation will always include = the Courier Neweffects of C_comp.=A0 Also, putting such a waveform and a = C_comp Courier Newvalue into an IBIS model will not double count C_comp.=A0 = This Courier Newis actually how IBIS simulators expect to get the data, = because Courier Newthey are the ones who do the compensation to avoid double = counting. =A0 Courier NewSo Giovanni did exactly the right thing. =A0 Courier NewThe only thing he needs to watch out for is that if he does = a Courier NewSPICE vs. IBIS correlation, he should have the value for = C_comp Courier Newin the IBIS model that was measured on the SPICE model, and = not Courier Newthe data sheet number, because the two may not be the same = value. Courier NewFor correlation work.=A0 However, if the IBIS model needs to = be Courier Newaccording to the data book, using the data book values is the = right Courier Newthing to do.=A0 (Remember, the SPICE model may not have all = the Courier Newlayout related extra capacitances in the circuit, which is = the Courier Newcase most often when someone has a pre layout model). =A0 Courier NewArpad = Muranyi Courier NewIntel = Corporation Courier = New=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =A0 Tahoma-----Original Message----- From: Kim Helliwell [mailto:kimgh@apple.com] Sent: Friday, August 01, 2003 2:45 PM To: Abril, Giovanni Cc: ibis-users@eda.org Subject: Re: [IBIS-Users] C_comp value in IBIS model If the waveforms already include the effects of C_comp, putting the C_comp value in the IBIS model will double-count the effect of C_comp, which is probably not what you want. I feel certain that the right thing to do is re-run the spice sims with C_comp removed from the simulations, and put the correct typ/min/max values of C_comp directly in the IBIS model If I were your customer trying to use this model, I'd be really unhappy to find that C_comp was (silently) being double counted. And I'd be suspicious of your model if C_comp were 0 with no explanation. At the very least, if you don't want to re-run things, you should set C_comp to 0 but put in a comment that the reason it's zero is that the effects of C_comp are included in the rising/falling waveforms. But this is definitely a distant second-best, in my opinion. Kim On Friday, August 1, 2003, at 02:13 PM, Abril, Giovanni wrote: ArialHello IBIS = gurus, ArialI've created an IBIS model using a spice netlist from the vendor, and I'm wondering what to do with C_comp.=A0 My rising and falling waveforms include the effect of C_comp, but what should I put in the IBIS model?=A0 I have a spec sheet that says min C_comp is 2.8pF, max is 10pF.=A0 Currently I have 5.0pF for typ, 2.8pF min and 10pF max. ArialThanks, ArialGiovanni Kim Helliwell Apple Computer kimgh@apple.com 408 974 9936 Kim Helliwell Apple Computer kimgh@apple.com 408 974 9936 - --Apple-Mail-8--133216226-- |------------------------------------------------------------------ |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 ------------------------------ Date: Fri, 1 Aug 2003 16:05:45 -0700 From: "Beal, Weston" Subject: RE: [IBIS-Users] C_comp value in IBIS model This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. - ------_=_NextPart_001_01C35881.70831173 Content-Type: text/plain; charset="iso-8859-1" Giovanni, The best solution would be to use your favorite C_comp calculation algorithm to calculate the typical value. Then use the datasheet values for min and max. When I'm in a hurry or have a non-critical model I usually set the typical value to something a little bit below the average of the min and max values from the data sheet. There have been a couple of presentations at IBIS summits about how to find values for C_comp. You can check the archives for those presentations. Regards, Weston - -----Original Message----- From: Abril, Giovanni [mailto:giovanni.abril@ngc.com] Sent: Friday, August 01, 2003 3:20 PM To: Beal, Weston; kimgh@apple.com Cc: ibis-users@eda.org Subject: RE: [IBIS-Users] C_comp value in IBIS model Thanks Weston. Is there a way for me to vary the capacitance in accordance with the spec then? The way you explained it, I should have the same C_comp value for typ, min and max, an extracted value, right? I'll assume ICX does it this way, since you work for Mentor. If the measured value turns out to be say 6pF, how would I cover a spec of 2.5pF? - -----Original Message----- From: Beal, Weston [mailto:weston_beal@mentorg.com] Sent: Friday, August 01, 2003 3:16 PM To: 'Kim Helliwell'; Abril, Giovanni Cc: ibis-users@eda.org Subject: RE: [IBIS-Users] C_comp value in IBIS model The effect of C_comp is only double counted if you are using a substandard simulator. All the major simulators that I know of expect C_comp to be specified with the same (static approximation) value as what is included in the waveforms. Somewhere in the simulator this is reconciled. Usually it is done is such a way that the capacitive effect is calculated out of the waveforms and then the lumped capacitor is present at the pin during the transient simulation.. As Arpad noted, the capacitive effect is usually impossible to remove from the SPICE models used to derive IBIS data. It is part of the transistor models. Arpad and others have proposed methods of finding the best static value of the die capacitance to use for C_comp and this value needs to be there. If you put the value of 0 for C_comp then the simulator will try to back out the capacitance and put no capacitance at the pin. At first this seems OK because the waveforms still have all the capacitive effect. The problem appears when a reflected signal hits this pin or if this pin is hi-Z. There would be no capacitive loading and you would then have invalid simulations. C_comp is your friend. Embrace it. ( and enhance it ) Regards, Weston - -----Original Message----- From: Kim Helliwell [mailto:kimgh@apple.com] Sent: Friday, August 01, 2003 2:45 PM To: Abril, Giovanni Cc: ibis-users@eda.org Subject: Re: [IBIS-Users] C_comp value in IBIS model If the waveforms already include the effects of C_comp, putting the C_comp value in the IBIS model will double-count the effect of C_comp, which is probably not what you want. I feel certain that the right thing to do is re-run the spice sims with C_comp removed from the simulations, and put the correct typ/min/max values of C_comp directly in the IBIS model If I were your customer trying to use this model, I'd be really unhappy to find that C_comp was (silently) being double counted. And I'd be suspicious of your model if C_comp were 0 with no explanation. At the very least, if you don't want to re-run things, you should set C_comp to 0 but put in a comment that the reason it's zero is that the effects of C_comp are included in the rising/falling waveforms. But this is definitely a distant second-best, in my opinion. Kim On Friday, August 1, 2003, at 02:13 PM, Abril, Giovanni wrote: Hello IBIS gurus, I've created an IBIS model using a spice netlist from the vendor, and I'm wondering what to do with C_comp. My rising and falling waveforms include the effect of C_comp, but what should I put in the IBIS model? I have a spec sheet that says min C_comp is 2.8pF, max is 10pF. Currently I have 5.0pF for typ, 2.8pF min and 10pF max. Thanks, Giovanni Kim Helliwell Apple Computer kimgh@apple.com 408 974 9936 - ------_=_NextPart_001_01C35881.70831173 Content-Type: text/html; charset="iso-8859-1"
Giovanni,
 
The best solution would be to use your favorite C_comp calculation algorithm to calculate the typical value.  Then use the datasheet values for min and max.  When I'm in a hurry or have a non-critical model I usually set the typical value to something a little bit below the average of the min and max values from the data sheet.
 
There have been a couple of presentations at IBIS summits about how to find values for C_comp.  You can check the archives for those presentations.
 
Regards,
Weston
 
-----Original Message-----
From: Abril, Giovanni [mailto:giovanni.abril@ngc.com]
Sent: Friday, August 01, 2003 3:20 PM
To: Beal, Weston; kimgh@apple.com
Cc: ibis-users@eda.org
Subject: RE: [IBIS-Users] C_comp value in IBIS model

Thanks Weston.
 
Is there a way for me to vary the capacitance in accordance with the spec then?  The way you explained it, I should have the same C_comp value for typ, min and max, an extracted value, right?  I'll assume ICX does it this way, since you work for Mentor.  If the measured value turns out to be say 6pF, how would I cover a spec of 2.5pF?
-----Original Message-----
From: Beal, Weston [mailto:weston_beal@mentorg.com]
Sent: Friday, August 01, 2003 3:16 PM
To: 'Kim Helliwell'; Abril, Giovanni
Cc: ibis-users@eda.org
Subject: RE: [IBIS-Users] C_comp value in IBIS model

The effect of C_comp is only double counted if you are using a substandard simulator.  All the major simulators that I know of expect C_comp to be specified with the same (static approximation) value as what is included in the waveforms.  Somewhere in the simulator this is reconciled.  Usually it is done is such a way that the capacitive effect is calculated out of the waveforms and then the lumped capacitor is present at the pin during the transient simulation..
 
As Arpad noted, the capacitive effect is usually impossible to remove from the SPICE models used to derive IBIS data.  It is part of the transistor models.  Arpad and others have proposed methods of finding the best static value of the die capacitance to use for C_comp and this value needs to be there.  If you put the value of 0 for C_comp then the simulator will try to back out the capacitance and put no capacitance at the pin.  At first this seems OK because the waveforms still have all the capacitive effect.  The problem appears when a reflected signal hits this pin or if this pin is hi-Z.  There would be no capacitive loading and you would then have invalid simulations.
 
C_comp is your friend.  Embrace it.  ( and enhance it )
 
Regards,
Weston
 
-----Original Message-----
From: Kim Helliwell [mailto:kimgh@apple.com]
Sent: Friday, August 01, 2003 2:45 PM
To: Abril, Giovanni
Cc: ibis-users@eda.org
Subject: Re: [IBIS-Users] C_comp value in IBIS model

If the waveforms already include the effects of C_comp, putting

the C_comp value in the IBIS model will double-count the effect

of C_comp, which is probably not what you want.


I feel certain that the right thing to do is re-run the spice sims

with C_comp removed from the simulations, and put the correct

typ/min/max values of C_comp directly in the IBIS model


If I were your customer trying to use this model, I'd be really unhappy

to find that C_comp was (silently) being double counted. And I'd

be suspicious of your model if C_comp were 0 with no explanation.


At the very least, if you don't want to re-run things, you should set

C_comp to 0 but put in a comment that the reason it's zero is that

the effects of C_comp are included in the rising/falling waveforms.


But this is definitely a distant second-best, in my opinion.


Kim


On Friday, August 1, 2003, at 02:13 PM, Abril, Giovanni wrote:


Hello IBIS gurus,


I've created an IBIS model using a spice netlist from the vendor, and I'm wondering what to do with C_comp.  My rising and falling waveforms include the effect of C_comp, but what should I put in the IBIS model?  I have a spec sheet that says min C_comp is 2.8pF, max is 10pF.  Currently I have 5.0pF for typ, 2.8pF min and 10pF max.


Thanks,


Giovanni



Kim Helliwell

Apple Computer

kimgh@apple.com

408 974 9936


- ------_=_NextPart_001_01C35881.70831173-- |------------------------------------------------------------------ |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 ------------------------------ Date: Wed, 6 Aug 2003 11:25:12 +0200 From: erik.van.der.ven@philips.com Subject: [IBIS-Users] How to determine Ccomp This is a multipart message in MIME format. - --=_alternative 0033F68DC1256D7A_= Content-Type: text/plain; charset="us-ascii" Dear IBIS collegues, What is the best way to determine Ccomp? Basically there are two ways: applying a ramp to the output terminal and calculate Ccomp from the slew-rate, or connecting an inductor to the output and calculate the Ccomp from the resonace frequency. Which method gives the best results? > always, or application dependent? How is this done for differential outputs. Outputs like LVDS have internal feedback loops to control the common level. Must something be done with the "other" output? Is the peaking a reliable measure for the value of Ccomp, or may the added inductor shift the poles and zeros of the feedback buffer due this load causing a peaking that is not just the parallel connection of Ccomp and the test inductor? With kind regards, Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334 - --=_alternative 0033F68DC1256D7A_= Content-Type: text/html; charset="us-ascii"
Dear IBIS collegues,

What is the best way to determine Ccomp?
Basically there are two ways: applying a ramp to the output terminal and calculate Ccomp from the slew-rate, or connecting an inductor to the output and calculate the Ccomp from the resonace frequency.
Which method gives the best results? > always, or application dependent?
How is this done for differential outputs. Outputs like LVDS have internal feedback loops to control the common level. Must something be done with the "other" output? Is the peaking a reliable measure for the value of Ccomp, or may the added inductor shift the poles and zeros of the feedback buffer due this load causing a peaking that is not just the parallel connection of Ccomp and the test inductor?

With kind regards,


Ir. Erik van der Ven
Room DB1032
Business Line Networking Infrastructure
Philips Semiconductors BV
Gerstweg 2
6534 AE Nijmegen
The Netherlands
Phone: +31-24-3534334 - --=_alternative 0033F68DC1256D7A_=-- |------------------------------------------------------------------ |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 ------------------------------ Date: Wed, 6 Aug 2003 12:50:04 -0700 From: "Muranyi, Arpad" Subject: RE: [IBIS-Users] C-Comp Sorry for the late response. The way this could be done is to simply repeat all C_comp parameters twice, one set for drive mode, and another for receive mode. One C_comp for receive mode would not be enough. However, there is more to this problem. The capacitance is not only dependent on whether the buffer drives or receives, but it is also voltage and frequency dependent. (See BIRD 79 and related presentations). The question in my mind is, BIRD 79 was voted down, because all those features will be doable with the *-AMS extensions to IBIS. The same is true for the C_comp drive/receive parameters. Is it worth adding these parameters to the existing style, or should we just go with the *-AMS style modeling and be able to describe anything freely? Arpad ============================================================ - -----Original Message----- From: Kim Helliwell [mailto:kimgh@apple.com] Sent: Friday, July 18, 2003 1:31 PM To: Robert Haller Cc: Roy Leventhal; anshuli@ti.com; ibis-users@eda.org Subject: Re: [IBIS-Users] C-Comp I'm just curious how this would be folded in with the 4-part C_comp (BIRD 76.1). Will we end up with 2 (or even 4) more C_comp values? Or does just 1 value adequately characterize the receiver? Kim On Friday, July 18, 2003, at 10:10 AM, Robert Haller wrote: > IBIS users, > > At today's IBIS open forum call, the issue was raised concerning the > limitations of Ccomp in the present IBIS spec. This issue is being > added to the next ibis open forum agenda technical discussion. > > I would like to solicit feedback from Model developers and users for > proposed changes, as well as voice my own opinion. > > Present syntax > > | typ min max > | > C_comp 3.90pF 3.10pF 4.80pF > > My personal experience has shown that I/O buffer often have different > C_comp when driving and receiving. To keep things simple, I would > suggest that a simple extension to the C_comp would be to specify a > C_comp driving and C_comp receiving for Model_type = I/O > > regards, > bob > -- > Robert J. Haller (rhaller@sisoft.com) > Principal Consultant > Signal Integrity Software Inc. > 6 Clock Tower Place, Suite 250 > Maynard, MA 01754 > Phone: (978) 461-0449, ext 15 > > > > Roy Leventhal wrote: >> Anshuli, >> Be aware that capacitive behavior is included in the V-T behavioral >> curves >> and/or the slew rates based on them. Unsophisticated providers of >> simulation >> software (hopefully, few remain) can thus mistakenly "double count" >> the >> effects of Ccomp unless they adjust for its presence in the circuit. >> So, why is Ccomp included? To account for the capacitive loading on >> the >> circuit, direct or reflected wave, and it is the total capacitive >> loading >> looking into the die pad. >> Best Regards, >> Roy > > > |------------------------------------------------------------------ > |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 > > Kim Helliwell Apple Computer kimgh@apple.com 408 974 9936 |------------------------------------------------------------------ |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 |------------------------------------------------------------------ |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 ------------------------------ Date: Wed, 6 Aug 2003 13:23:35 -0700 From: "Muranyi, Arpad" Subject: RE: [IBIS-Users] How to determine C_comp This is a multi-part message in MIME format. - ------_=_NextPart_001_01C35C58.9D156B6E Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I know of about six methods to measure C_comp: =20 =20 *RC time constant with a step function=20 *does not show voltage or frequency dependencies=20 *Time Domain forced saw tooth voltage=20 *can provide voltage dependent capacitance curves=20 *does not show frequency dependence=20 *TDR technique (Time Domain)=20 *does not show voltage or frequency dependencies=20 *Frequency Domain sweep with tank circuit=20 *can provide voltage dependent capacitance curves=20 *does not show frequency dependence(?)=20 *Frequency Domain imaginary current=20 *can provide voltage and frequency dependent capacitance curves=20 *shows frequency dependence very well=20 *Frequency Domain pole/zero method=20 *See BIRD 79 on IBIS web page for details=20 =20 Arpad Muranyi Intel Corporation =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =20 =20 - -----Original Message----- From: erik.van.der.ven@philips.com [mailto:erik.van.der.ven@philips.com] Sent: Wednesday, August 06, 2003 2:25 AM To: ibis-users@eda.org Cc: cor.duijndam@philips.com Subject: [IBIS-Users] How to determine Ccomp Dear IBIS collegues,=20 What is the best way to determine Ccomp?=20 Basically there are two ways: applying a ramp to the output terminal and = calculate Ccomp from the slew-rate, or connecting an inductor to the = output and calculate the Ccomp from the resonace frequency.=20 Which method gives the best results? > always, or application dependent? = How is this done for differential outputs. Outputs like LVDS have = internal feedback loops to control the common level. Must something be = done with the "other" output? Is the peaking a reliable measure for the = value of Ccomp, or may the added inductor shift the poles and zeros of = the feedback buffer due this load causing a peaking that is not just the = parallel connection of Ccomp and the test inductor?=20 With kind regards,=20 Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334 - ------_=_NextPart_001_01C35C58.9D156B6E Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I know of=20 about six methods to measure C_comp:
 
 
•RC=20 time constant with a step function
•does not show voltage or frequency=20 dependencies
•Time=20 Domain forced saw tooth voltage
•can provide voltage dependent capacitance = curves=20
•does not show frequency = dependence=20
•TDR=20 technique (Time Domain)
•does not show voltage or frequency=20 dependencies
•Frequency=20 Domain sweep with tank circuit
•can provide voltage dependent capacitance = curves=20
•does not show frequency = dependence(?)
•Frequency=20 Domain imaginary current
•can provide voltage and frequency dependent = capacitance=20 curves
•shows frequency dependence very = well
•Frequency=20 Domain pole/zero method
•See BIRD 79 on IBIS web page for details =
 
Arpad=20 Muranyi
Intel=20 Corporation
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D
 
 
-----Original Message-----
From: = erik.van.der.ven@philips.com=20 [mailto:erik.van.der.ven@philips.com]
Sent: Wednesday, August = 06, 2003=20 2:25 AM
To: ibis-users@eda.org
Cc:=20 cor.duijndam@philips.com
Subject: [IBIS-Users] How to = determine=20 Ccomp


Dear IBIS = collegues,

What is the = best way to=20 determine Ccomp?
Basically = there are two=20 ways: applying a ramp to the output terminal and calculate Ccomp from = the=20 slew-rate, or connecting an inductor to the output and calculate the = Ccomp from=20 the resonace frequency.
Which method=20 gives the best results? > always, or application dependent? =
How is this done for differential outputs. = Outputs like=20 LVDS have internal feedback loops to control the common level. Must = something be=20 done with the "other" output? Is the peaking a reliable measure for the = value of=20 Ccomp, or may the added inductor shift the poles and zeros of the = feedback=20 buffer due this load causing a peaking that is not just the parallel = connection=20 of Ccomp and the test inductor?

With=20 kind regards,


Ir. = Erik van der=20 Ven
Room DB1032
Business Line Networking Infrastructure
Philips = Semiconductors BV
Gerstweg 2
6534 AE Nijmegen
The = Netherlands
Phone:=20 +31-24-3534334 - ------_=_NextPart_001_01C35C58.9D156B6E-- |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 7 Aug 2003 09:29:12 +0200 From: erik.van.der.ven@philips.com Subject: Re: [IBIS-Users] How to determine Ccomp This is a multipart message in MIME format. - --=_alternative 00295778C1256D7B_= Content-Type: text/plain; charset="us-ascii" Hello Dekker, Thank you for your response. I meant determination of Ccomp by simulations, sorry. Best regards, Erik. Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334 "Dekker Graden" 08/06/03 06:11 PM Please respond to "Dekker Graden" To: Erik van der Ven/NYM/SC/PHILIPS@EMEA2 cc: Subject: Re: [IBIS-Users] How to determine Ccomp Classification: Hello Erik If your talking about getting Ccomp from measurements there is another way of doing it that I have found to be quite good using a TDR. The measurement method depends on if you have packaged or die only chips. Anyway.... you just use a TDR to shoot in a pulse... then use something like Iconnect (TDA systems software) to calculate what Ccomp is from the waveform you get. There are papers and application notes on http://www.tdasystems.com site that explain this way of getting Ccomp. The only down side here is that you tend to only get the basic Ccomp... about typical. Anyway... just a pointer to a different direction if its an option for you. The software and equipment is rather spendy. Dekker - ----- Original Message ----- From: erik.van.der.ven@philips.com To: ibis-users@eda.org Cc: cor.duijndam@philips.com Sent: Wednesday, August 06, 2003 2:25 AM Subject: [IBIS-Users] How to determine Ccomp Dear IBIS collegues, What is the best way to determine Ccomp? Basically there are two ways: applying a ramp to the output terminal and calculate Ccomp from the slew-rate, or connecting an inductor to the output and calculate the Ccomp from the resonace frequency. Which method gives the best results? > always, or application dependent? How is this done for differential outputs. Outputs like LVDS have internal feedback loops to control the common level. Must something be done with the "other" output? Is the peaking a reliable measure for the value of Ccomp, or may the added inductor shift the poles and zeros of the feedback buffer due this load causing a peaking that is not just the parallel connection of Ccomp and the test inductor? With kind regards, Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334 - --=_alternative 00295778C1256D7B_= Content-Type: text/html; charset="us-ascii"
Hello Dekker,


Thank you for your response. I meant determination of Ccomp by simulations, sorry.

Best regards,
Erik.

Ir. Erik van der Ven
Room DB1032
Business Line Networking Infrastructure
Philips Semiconductors BV
Gerstweg 2
6534 AE Nijmegen
The Netherlands
Phone: +31-24-3534334







"Dekker Graden" <tarka@teraspeed.com>

08/06/03 06:11 PM
Please respond to "Dekker Graden"

       
        To:        Erik van der Ven/NYM/SC/PHILIPS@EMEA2
        cc:        
        Subject:        Re: [IBIS-Users] How to determine Ccomp

        Classification:        




Hello Erik
 
If your talking about getting Ccomp from measurements there is another way of doing it that I have found to be quite good using a TDR. The measurement method depends on if you have packaged or die only chips. Anyway.... you just use a TDR to shoot in a pulse... then use something like Iconnect (TDA systems software) to calculate what Ccomp is from the waveform you get.
 
There are papers and application notes on http://www.tdasystems.com site that explain this way of getting Ccomp.
 
The only down side here is that you tend to only get the basic Ccomp... about typical.
 
Anyway... just a pointer to a different direction if its an option for you. The software and equipment is rather spendy.
 
Dekker
----- Original Message -----
From: erik.van.der.ven@philips.com
To: ibis-users@eda.org
Cc: cor.duijndam@philips.com
Sent: Wednesday, August 06, 2003 2:25 AM
Subject: [IBIS-Users] How to determine Ccomp


Dear IBIS collegues,

What is the best way to determine Ccomp?
Basically there are two ways: applying a ramp to the output terminal and calculate Ccomp from the slew-rate, or connecting an inductor to the output and calculate the Ccomp from the resonace frequency.
Which method gives the best results? > always, or application dependent?
How is this done for differential outputs. Outputs like LVDS have internal feedback loops to control the common level. Must something be done with the "other" output? Is the peaking a reliable measure for the value of Ccomp, or may the added inductor shift the poles and zeros of the feedback buffer due this load causing a peaking that is not just the parallel connection of Ccomp and the test inductor?

With kind regards,


Ir. Erik van der Ven
Room DB1032
Business Line Networking Infrastructure
Philips Semiconductors BV
Gerstweg 2
6534 AE Nijmegen
The Netherlands
Phone: +31-24-3534334

- --=_alternative 00295778C1256D7B_=-- |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 7 Aug 2003 11:15:43 -0400 From: "Mike LaBonte" Subject: RE: [IBIS-Users] How to determine C_comp The HSPICE manual gives a circuit example of the Frequency Domain imaginary current method in the Buffer Characterization section, by the way. Mike LaBonte - -----Original Message----- From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On Behalf Of Muranyi, Arpad Sent: Wednesday, August 06, 2003 4:24 PM To: ibis-users@eda.org Subject: RE: [IBIS-Users] How to determine C_comp I know of about six methods to measure C_comp: •RC time constant with a step function •does not show voltage or frequency dependencies •Time Domain forced saw tooth voltage •can provide voltage dependent capacitance curves •does not show frequency dependence •TDR technique (Time Domain) •does not show voltage or frequency dependencies •Frequency Domain sweep with tank circuit •can provide voltage dependent capacitance curves •does not show frequency dependence(?) •Frequency Domain imaginary current •can provide voltage and frequency dependent capacitance curves •shows frequency dependence very well •Frequency Domain pole/zero method •See BIRD 79 on IBIS web page for details Arpad Muranyi Intel Corporation ================================================ - -----Original Message----- From: erik.van.der.ven@philips.com [mailto:erik.van.der.ven@philips.com] Sent: Wednesday, August 06, 2003 2:25 AM To: ibis-users@eda.org Cc: cor.duijndam@philips.com Subject: [IBIS-Users] How to determine Ccomp Dear IBIS collegues, What is the best way to determine Ccomp? Basically there are two ways: applying a ramp to the output terminal and calculate Ccomp from the slew-rate, or connecting an inductor to the output and calculate the Ccomp from the resonace frequency. Which method gives the best results? > always, or application dependent? How is this done for differential outputs. Outputs like LVDS have internal feedback loops to control the common level. Must something be done with the "other" output? Is the peaking a reliable measure for the value of Ccomp, or may the added inductor shift the poles and zeros of the feedback buffer due this load causing a peaking that is not just the parallel connection of Ccomp and the test inductor? With kind regards, Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334 |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 7 Aug 2003 10:20:16 -0700 From: "Eric Hsu" Subject: [IBIS-Users] R/L/C_pin issue This is a multi-part message in MIME format. - ------_=_NextPart_001_01C35D08.2B261770 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi IBIS expert, When editing R/L/C_pin of ibis model, should I need to consider the = connection from die to board? such as flip-chip BGA case:"pad --> RDL = - --> bump --> package ( trace ) --> ball"? If true, what's the reasonable way to consider about ball portion? My = point is because it is the interface between chip and board. People may = take this portion as part of board trace. Best Regards, Eric Hsu Interface Technologies NetLogic Microsystems, Inc. 450 National Ave. Mountain View, CA 94043 650-961-6676 x198 This e-mail contains NetLogic Microsystems, Inc. Confidential = information - ------_=_NextPart_001_01C35D08.2B261770 Content-Type: application/ms-tnef; name="winmail.dat" Content-Transfer-Encoding: base64 eJ8+IhERAQaQCAAEAAAAAAABAAEAAQeQBgAIAAAA5AQAAAAAAADoAAEIgAcAGAAAAElQTS5NaWNy b3NvZnQgTWFpbC5Ob3RlADEIAQ2ABAACAAAAAgACAAEEgAEAEAAAAFIvTC9DX3BpbiBpc3N1ZQAu BQEFgAMADgAAANMHCAAHAAoAFAAQAAQAGwEBIIADAA4AAADTBwgABwAKABQAEAAEABsBAQmAAQAh AAAANEVDODRGNUNBNEUwNTc0REJGNjIxQkEzMDg1QTk3OEMAVAcBA5AGAHwHAAAvAAAACwACAAEA AAADACYAAAAAAAMANgAAAAAAQAA5AHAXJisIXcMBHgA9AAEAAAABAAAAAAAAAAIBRwABAAAALAAA AGM9dXM7YT0gO3A9bmxtcztsPU9SSU9OMy0wMzA4MDcxNzIwMTZaLTExODYAHgBwAAEAAAAQAAAA Ui9ML0NfcGluIGlzc3VlAAIBcQABAAAAFgAAAAHDXQgrJUD66UzZ2Usnqm81USCcZScAAB4AGgwB AAAACQAAAEVyaWMgSHN1AAAAAB4AHQ4BAAAAEAAAAFIvTC9DX3BpbiBpc3N1ZQACAQkQAQAAAIMC AAB/AgAATAMAAExaRnW/mELtAwAKAHJjcGcxMjUWMgD4C2BuDhAwMzPJDsZmZQ9ANDEB9wKkEwPj AgBjaArAc2V0/DAgBxMCgwBQBFURqQhVKweyAoB9CoF2CJB3a+kLgGQ0DGBjAFALAwu1gCBIaSBJ QkkF8MhleHAEkHQsCqIKhBkKgFdoCfAXYGRpdAULgGcH8C9ML0NfAnALgCBvZiBpYgUEACAEYmws IHNoyQhgbGQXECBuCeAbUPh0byAFoACBBIEb0BiwWxwCG5BjGSACICADUiDfGQAcwBvhBuALET8a 8BXguGggYQQgDvAFIC0RwGEFICBCR0EcAB8QZYw6IgqwG1AtLT4H8MRETCCzYnVtH8AgwqEKsGNr YWccwCgb0DZyANAcwCkhNAdAbCJ6PxfkSRogIrAKUBrgd/kR0HQnBCAcoglwHxACIMUBoGwcwHdh eRvbAaB/CGAFQCNyIgAJER1RHqBNXyZAJ9ALgAVAGmFiBZBhfnUSABowKNMcoiixBJBm8yLSKSB0 dwnhHAAfogBwgxtQHlMuIFBlbwtQ5xzAAMAmQWFrHgEfoAQgfyfVHwIKsQVAGhEeUyKkLo0X6kIH kAVAUmVnCxHqcxfbRQUQYxbgHsAj1aUqV1QFkGhuCPFnCJBnEZAX8wfAdEwz4TIgTZMN4ANgc3kw gGVtMRBJFxBuYy+VNDUSME6vJOAdUQdAEkB2L4ZNCGCzAjALcSBWFVEa4EMgAFw5NBAwD3AX8zY2 sC0AOTYxLTY2NzbgIHgxOTgX5QBAM2D5LXJlLQDAAxEcEThSB6LnNL81xhQhbmYcUQIwBzH7KjEC EHIAwB1COvYBQBf5BRUBAEGwAB4ANRABAAAAQgAAADw4Rjg0NDRBNkUyQzhGNTQxOEM3RDA0Q0ZF MEYxRTFFQzA0NjBDN0BvcmlvbjMubmV0bG9naWNtaWNyby5jb20+AAAAAwCAEP////8fAPMQAQAA ACgAAABSAP/4TAD/+EMAXwBwAGkAbgAgAGkAcwBzAHUAZQAuAEUATQBMAAAACwD2EAAAAABAAAcw cBcmKwhdwwFAAAgwINwqKwhdwwEDAN4/r28AAAMA8T8JBAAAHgD4PwEAAAAJAAAARXJpYyBIc3UA AAAAAgH5PwEAAABYAAAAAAAAANynQMjAQhAatLkIACsv4YIBAAAAAAAAAC9PPU5MTVMvT1U9RklS U1QgQURNSU5JU1RSQVRJVkUgR1JPVVAvQ049UkVDSVBJRU5UUy9DTj1FSFNVAB4A+j8BAAAAFQAA AFN5c3RlbSBBZG1pbmlzdHJhdG9yAAAAAAIB+z8BAAAAHgAAAAAAAADcp0DIwEIQGrS5CAArL+GC AQAAAAAAAAAuAAAAAwAZQAAAAAADABpAAAAAAB4AMEABAAAABQAAAEVIU1UAAAAAHgAxQAEAAAAF AAAARUhTVQAAAAAeADhAAQAAAAUAAABFSFNVAAAAAB4AOUABAAAAAgAAAC4AAAADAAlZAwAAAAsA ZoEIIAYAAAAAAMAAAAAAAABGAAAAAA6FAAAAAAAAAwB+gQggBgAAAAAAwAAAAAAAAEYAAAAAUoUA AHN5AQAeAH+BCCAGAAAAAADAAAAAAAAARgAAAABUhQAAAQAAAAQAAAA5LjAAAwDCgQggBgAAAAAA wAAAAAAAAEYAAAAAAYUAAAAAAAALAMeBCCAGAAAAAADAAAAAAAAARgAAAAADhQAAAAAAAAMAzIEI IAYAAAAAAMAAAAAAAABGAAAAABGFAAAAAAAAAwDRgQggBgAAAAAAwAAAAAAAAEYAAAAAEIUAAAAA AAADANiBCCAGAAAAAADAAAAAAAAARgAAAAAYhQAAAAAAAAsADYIIIAYAAAAAAMAAAAAAAABGAAAA AAaFAAAAAAAACwApAAAAAAALACMAAAAAAAMABhCBNvqqAwAHEOIBAAADABAQAAAAAAMAERAAAAAA HgAIEAEAAABlAAAASElJQklTRVhQRVJULFdIRU5FRElUSU5HUi9ML0NQSU5PRklCSVNNT0RFTCxT SE9VTERJTkVFRFRPQ09OU0lERVJUSEVDT05ORUNUSU9ORlJPTURJRVRPQk9BUkQ/U1VDSEFTRgAA AAACAX8AAQAAAEIAAAA8OEY4NDQ0QTZFMkM4RjU0MThDN0QwNENGRTBGMUUxRUMwNDYwQzdAb3Jp b24zLm5ldGxvZ2ljbWljcm8uY29tPgAAABii - ------_=_NextPart_001_01C35D08.2B261770-- |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 07 Aug 2003 18:18:47 -0700 From: Bob Ross Subject: [IBIS-Users] BIRD82.1 Clarification of Clamp Table Use This is a multi-part message in MIME format. - --------------090607040305050902040500 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit To IBIS Committee: BIRD82.1 by Lynne Green and Bob Haller is attached for review. Bob - -- Bob Ross Teraspeed Consulting Group LLC Teraspeed Consulting Group LLC 2926 SE Yamhill St. Device Modeling Division Portland, OR 97214 13610 SW Harness Lane 503-239-5536 Beaverton, OR 97008 http://www.teraspeed.com 503-430-1065 bob@teraspeed.com 503-246-8048 Direct - --------------090607040305050902040500 Content-Type: text/plain; name="BIRD82.1.txt" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="BIRD82.1.txt" **************************************************************************** **************************************************************************** Buffer Issue Resolution Document (BIRD) BIRD ID#: 82.1 ISSUE TITLE: Clarification of Clamp Table Use REQUESTOR: Lynne Green, Cadence, and Robert Haller, SiSoft DATE SUBMITTED: June 27, 2003, August 7, 2003 DATE ACCEPTED BY IBIS OPEN FORUM: Pending **************************************************************************** **************************************************************************** STATEMENT OF THE ISSUE: Double-counting of clamp currents is the result of confusion over how the clamp currents are summed by the tools. Model makers are often unsure if clamps are always active, or whether the [Power Clamp] works only with the [Pullup] and the [GND Clamp] only with the [Pulldown]. This is difficult, if not impossible, to detect with a parser. This issue can be resolved by clarification of the text under | Keywords: [Pulldown], [Pullup], [GND Clamp], [POWER Clamp] **************************************************************************** STATEMENT OF THE RESOLVED SPECIFICATIONS: Old text: | It is assumed that the simulator sums the clamp tables | together with the appropriate [Pullup] or [Pulldown] table | when a buffer is driving high or low, respectively. From this | assumption and the nature of 3-statable buffers, it follows | that the data in the clamping table sections are handled as | constantly present tables and the [Pullup] and [Pulldown] | tables are used only when needed in the simulation. | Proposed text: | It is intended that both clamp tables are summed together | and then added to the appropriate [Pullup] or [Pulldown] | table when a buffer is driving high or low, respectively. | | From this assumption and the nature of 3-statable buffers, | it follows that the data in the clamping table sections are | handled as constantly present tables and the [Pullup] and | [Pulldown] tables are used only when needed in the simulation. | **************************************************************************** ANALYSIS PATH/DATA THAT LED TO SPECIFICATION: Comments from members of the Quality committee, as well as others, indicated that some models double-count clamp currents. **************************************************************************** ANY OTHER BACKGROUND INFORMATION: Initially, it was suggested that this be detected by the parser, but it was decided that this was too difficult. This led to further discussion of why double-counting occurred, and how clarification of the spec could help address the issue. **************************************************************************** - --------------090607040305050902040500-- |------------------------------------------------------------------ |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 ------------------------------ Date: Fri, 8 Aug 2003 09:29:38 -0700 From: "deBurgh, Guy" Subject: [IBIS-Users] BIRD74.6 EMI Parameters To the IBIS Committee. BIRD 74 has been updated. Briefly the changes are: 1) Improved example, better clarification of syntax, and grammar corrections. The technical content has not been affected. Guy de Burgh Mentor Graphics **************************************************************************** ** **************************************************************************** ** BIRD ID#: 74.6 ISSUE TITLE: EMI Parameters REQUESTER: Guy de Burgh, Mentor Graphics DATE SUBMITTED: 3/19/02, 5/31/02, /9/16/02, 4/30/03, 5/21/03, 7/18/03, 8/08/03 DATE ACCEPTED BY IBIS OPEN FORUM: Pending **************************************************************************** ** **************************************************************************** ** STATEMENT OF THE ISSUE: Additional parameters required for EMI analysis. All these parameters are optional. **************************************************************************** ** STATEMENT OF THE RESOLVED SPECIFICATIONS: The following changes are proposed: **************************************************************************** ** Changes and additions to the IBIS Specification are shown by the |* lines. Examples are shown as they would be seen in a specification. |=========================================================================== == |=========================================================================== == | | Section 10 | | E M I P A R A M E T E R S | |=========================================================================== == | | There are two sections here: one for a [Component] and one for a [Model]. | |=========================================================================== == |=========================================================================== == | | This section describes the structure of the EMI parameters under a | top-level [Component] keyword. It is used to describe the EMI parameters | associated with a [Component]. The parameters must be surrounded by | the [Begin EMI Component] and [End EMI Component] keywords. | | The following set of keywords are defined: | | [Begin EMI Component] | [End EMI Component] | [Pin EMI] | [Pin Domain EMI] | | The following set of subparameters are defined: | | Domain | Cpd | C_Heatsink_gnd | C_Heatsink_float | |=========================================================================== == | | Keyword: [Begin EMI Component] | Required: No | Description: Marks the beginning of the Component EMI parameters. | Sub-Params: Domain, Cpd, C_Heatsink_gnd, C_Heatsink_float | | Domain indicates whether the component is a digital, analog | or part digital part analog. Analog circuits are more | susceptible to low-level noise. Analog circuits operate at | very low signal levels (mV or uV) and can contain high gain | amplifiers. In contrast, digital circuits operate at | relatively large signal levels (compared to analog circuits). | | The syntax for Domain is: | Domain Domain_value | | Where Domain_value is an enumerated argument, and is one of: | Digital, Analog, Digital_analog | | This subparameter is optional. If not entered the default | is Digital. | | Cpd is the power dissipation capacitance parameter. | Cpd (Power Dissipation Capacitance) is the internal parasitic | capacitance (e.g., gate-to-source and gate-to-drain | capacitance) plus the equivalent capacitance associated with | the through currents when both transistors (n-channel and | p-channel) are momentarily conducting. | | It is typically for CMOS devices, and helps provide a more | accurate estimation of the power bus current, and therefore | the noise voltage on the power bus. Knowing the high frequency | noise on the power bus (due to switching of digital circuits) | the radiation can be calculated. | | Sometimes Iccd (Dynamic power supply current) is found in | databooks. It is normally given for FACT families. Iccd is | specified in units of mA/MHz. | | Cpd can be calculated from Iccd by the equation | Cpd (nF) = Iccd (mA/MHz) / Vcc (V). | | The syntax for Cpd is: | Cpd = capacitance_value | | The units of capacitance_value are Farads. | | This subparameter is optional. If not entered the default | is 0.0F. | | C_Heatsink_Float and C_Heatsink_Gnd define the heatsink | capacitance and connection conditions. | C_Heatsink_Float indicates that the heatsink is floating, | and C_Heatsink_Gnd indicates that the heatsink is grounded. | | Internal currents inside a (high speed) IC can be closely | coupled onto a heatsink. As the heatsink is physically much | larger than the IC silicon chip and bond wires it is a more | efficient radiator. Knowing the capacitance of the heatsink | the radiated electric field can be estimated. | | Only one of these subparameters can be defined. It is not | legal to define both. It is legal to omit both. In this case | it means that a heatsink is not present. | | The subparameter takes one argument: the heatsink capacitance | | The syntax for Heatsink_cap is: | C_Heatsink_float = capacitance_value | C_Heatsink_gnd = capacitance_value | | The units for capacitance_value are Farads. | | This subparameter is optional. If not entered the default | is that the component does not have a heatsink. | |=========================================================================== == | | Keyword: [End EMI Component] | Required: No | Description: Marks the end of the Component EMI parameters. | |=========================================================================== == | | Example: | [Begin EMI Component] Domain Digital Cpd = 6.4pF C_Heatsink_gnd = 3.4pF [End EMI Component] | |=========================================================================== == | | Keyword: [Pin EMI] | Required: No | Description: Specifies the EMI parameters for a Pin. | Sub-Params: domain_name, clock_div | Usage Rules: Each line must contain three columns. | The first column must contain the pin name. This pin name | must match a pin name in the [Pin] keyword. (The pin name | is the first column in the [Pin] record.) | | The second column is the domain name. This specifies the clock | domain for that pin. This is used by [Pin Domain EMI]. | The field should be set to NA if unused. | | The default for domain_name is that the percentage of power | used is 100%. | | The third column is the clock division. This is the ratio | of the frequency at this pin to the reference pin. The | reference pin is always set to "1.0". The ratio is a floating | point number. The choice of the reference in does not matter | as this information is pin to pin ratios. It is suggested that | the pin with the maximum frequency is chosen as the reference. | The field should be set to NA if unused. | | The default for clock_div is 1.0 | | Column length limits are: | pin_name 5 characters max | domain_name 20 characters max | clock_div 5 characters max | | It is not a requirement to specify every pin. An undefined | pin will default to 100% power usage for Domain_name, | and 1.0 for clock_div. | |=========================================================================== = | | Keyword: [Pin Domain EMI] | Required: No | Description: Specifies the percentage of power used in each clock domain. | Sub-Params: percentage | Usage Rules: Each line must contain two columns. | The first column must contain the domain_name. This name must | match a domain name in the [Pin EMI] keyword. (The domain | name is the second column in that record.) | | The percentage represents a user definable percentage of the | power used by that domain. It is an integer in the range | 0 < percentage =< 100 | | Column length limits are: | domain_name 20 characters max | percentage 5 characters max | |=========================================================================== == | | Example: | [Begin EMI Component] Domain Digital Cpd = 6.4pF | [Pin EMI] domain_name clock_div 4 MEM 0.5 5 MEM 0.5 7 NA 0.5 | domain_name defaults to 100% 8 RIOG NA | clock_div defaults to 1.0 14 CPU 1.0 15 RIOG 0.5 | [Pin Domain EMI] percentage CPU 40 MEM 30 RIOG 30 | [End EMI Component] | |=========================================================================== == |=========================================================================== == | | This section describes the structure of the EMI parameters under a | top-level [Model] keyword. It is used to describe the EMI parameters | associated with a [Model]. The parameters must be surrounded by | the [Begin EMI Model] and [End EMI Model] keywords. | | The following set of keywords are defined: | | [Begin EMI Model] | [End EMI Model] | | The following set of subparameters are defined: | | Model_emi_type | Model_Domain | |=========================================================================== == | | Keyword: [Begin EMI Model] | Required: No | Description: Marks the beginning of the Model EMI parameters. | Sub-Params: Model_emi_type, Domain | | Model_emi_type indicates whether the model (for this pin) is a | ferrite or not. | | The syntax for Model_emi_type is: | Model_emi_type Model_emi_type_value | | Where Model_emi_type_value is an enumerated argument, and is | one of: | Ferrite, Not_a_ferrite | | If not entered (the default) the model is Not_a_ferrite. | | Model_Domain indicates whether the model is digital or analog. | This is only used if the [Component EMI] Domain is set to | Digital_analog. If the [Component EMI] Domain is set to | anything else Model_Domain is ignored. | | The syntax for Domain is: | Model_Domain Domain_value | | Where Domain_value is one of: | Digital, Analog | | If not entered the default is to use the [Component EMI] | Domain setting and it's default. | |=========================================================================== == | | Keyword: [End EMI Model] | Required: No | Description: Marks the end of the Model EMI parameters. | |=========================================================================== == | |--------------------------------------------------------------------------- - -- | [Begin EMI Model] Domain Analog Model_emi_type Ferrite [End EMI Model] | **************************************************************************** ** ANALYSIS PATH/DATA THAT LED TO SPECIFICATION: EMI is becoming increasingly important in PCB design. EMI simulation is not an exact science. The problem is too complicated for today's computers. However being able to identify which signals radiate more than others (relatively), and showing where the energy flows in a system does help in identifying potential EMI problems. No longer requiring absolute accuracy, but requiring relative accuracy, makes the analysis considerably easier, and possible. The parameters specified here are chosen for that purpose, and have been specifically chosen as they are readily available in data books in order to make it easier to create models. **************************************************************************** ** ANY OTHER BACKGROUND INFORMATION: Presentations on this topic were given at the IBIS Summit Meeting at DAC on June 21, 2001, and at DesignCon on January 28, 2002. **************************************************************************** ** |------------------------------------------------------------------ |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 ------------------------------ Date: Mon, 11 Aug 2003 11:00:37 +0200 From: erik.van.der.ven@philips.com Subject: [IBIS-Users] pullup/pulldown range LVDS This is a multipart message in MIME format. - --=_alternative 0031B6BCC1256D7F_= Content-Type: text/plain; charset="us-ascii" Dear IBIS users, For ECL the pullup/pulldown curves only have to cover the 0-2.2V range (relative to Vcc). This is because of the limited voltage swing, which is centered around a DC voltage far away from the supply rails. How about LVDS? It has a similar output swing. Is it needed to cover the "regular" voltage range of -3.3-6.6V (for 3.3V supply)? It's rather a waste of accuracy for the typical operation area (less points remain). With kind regards, Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334 - --=_alternative 0031B6BCC1256D7F_= Content-Type: text/html; charset="us-ascii"
Dear IBIS users,

For ECL the pullup/pulldown curves only have to cover the 0-2.2V range (relative to Vcc). This is because of the limited voltage swing, which is centered around a DC voltage far away from the supply rails.
How about LVDS? It has a similar output swing. Is it needed to cover the "regular" voltage range of -3.3-6.6V (for 3.3V supply)? It's rather a waste of accuracy for the typical operation area (less points remain).

With kind regards,


Ir. Erik van der Ven
Room DB1032
Business Line Networking Infrastructure
Philips Semiconductors BV
Gerstweg 2
6534 AE Nijmegen
The Netherlands
Phone: +31-24-3534334 - --=_alternative 0031B6BCC1256D7F_=-- |------------------------------------------------------------------ |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 ------------------------------ Date: Mon, 11 Aug 2003 10:42:02 -0400 From: "Allen Chao" Subject: [IBIS-Users] for the minimum and maximum data Hi IBIS experts, For the minimum and maximum datas, do we only consider the temperature and the voltage source (90% to 110%)? or we should consider the slow/fast corners for the resistors and mos transistors if we have those libraries available? Thanks in advance, Allen |------------------------------------------------------------------ |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 ------------------------------ Date: Mon, 11 Aug 2003 13:06:06 -0400 From: "Lynne Green" Subject: RE: [IBIS-Users] pullup/pulldown range LVDS This is a multi-part message in MIME format. - ------_=_NextPart_001_01C3602A.DA493B92 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Hello, Erik, =20 One would normally add one point at each end outside of the normal operating range (using the same current as the normal operating endpoint). This only "costs" two points of the 100 allowed in IBIS 3.2, so there is no significant loss in accuracy. =20 This does two things: 1) Allows the simulator tool to search the entire -Vcc,+2Vcc range when=20 performing convergence phase, resulting in faster convergence. 2) Allows the IBIS file to pass the IBIS parser. =20 | Example with 12mA end points in I-V data: - -3.3V -12m | dummy end point 0.4V -12m | end point of actual operating region | normal data, up to 96 more lines 1.6V 12m | end point of actual operating region 6.6V 12m | dummy end point =20 Best regards, Lynne =20 - -----Original Message----- From: erik.van.der.ven@philips.com [mailto:erik.van.der.ven@philips.com] Sent: Monday, August 11, 2003 2:01 AM To: ibis-users@eda.org Subject: [IBIS-Users] pullup/pulldown range LVDS Dear IBIS users,=20 For ECL the pullup/pulldown curves only have to cover the 0-2.2V range (relative to Vcc). This is because of the limited voltage swing, which is centered around a DC voltage far away from the supply rails.=20 How about LVDS? It has a similar output swing. Is it needed to cover the "regular" voltage range of -3.3-6.6V (for 3.3V supply)? It's rather a waste of accuracy for the typical operation area (less points remain).=20 With kind regards,=20 Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334 - ------_=_NextPart_001_01C3602A.DA493B92 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Message
Hello,=20 Erik,
 
One would normally add one point at each end outside of = the normal=20 operating
range=20 (using the same current as the normal operating endpoint).  This=20 only
"costs" two points of the 100 allowed in IBIS 3.2, so there is = no=20 significant loss
in=20 accuracy.
 
This=20 does two things:
1)=20 Allows the simulator tool to search the entire -Vcc,+2Vcc range when=20
   performing convergence phase, resulting in=20 faster convergence.
2)=20 Allows the IBIS file to pass the IBIS parser.
 
|=20 Example with 12mA end points in I-V data:
-3.3V   -12m   | dummy end = point
0.4V    -12m   | end point of actual = operating=20 region
          &nbs= p;         =20 | normal data, up to 96 more lines
1.6V     12m   | end point of = actual=20 operating region
6.6V     12m   | dummy end=20 point
 
Best=20 regards,
Lynne
 
-----Original Message-----
From: = erik.van.der.ven@philips.com=20 [mailto:erik.van.der.ven@philips.com]
Sent: Monday, August = 11, 2003=20 2:01 AM
To: ibis-users@eda.org
Subject: [IBIS-Users] = pullup/pulldown range LVDS


Dear IBIS users,

For ECL the=20 pullup/pulldown curves only have to cover the 0-2.2V range (relative to = Vcc).=20 This is because of the limited voltage swing, which is centered around a = DC=20 voltage far away from the supply rails.
How about LVDS? It has a similar output swing. Is it needed to = cover the=20 "regular" voltage range of -3.3-6.6V (for 3.3V supply)? It's rather a = waste of=20 accuracy for the typical operation area (less points remain).=20

With kind = regards,=20


Ir. Erik van der = Ven
Room=20 DB1032
Business Line Networking Infrastructure
Philips = Semiconductors=20 BV
Gerstweg 2
6534 AE Nijmegen
The Netherlands
Phone:=20 +31-24-3534334 =00 - ------_=_NextPart_001_01C3602A.DA493B92-- |------------------------------------------------------------------ |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 ------------------------------ Date: Wed, 13 Aug 2003 15:09:19 +0100 From: Gavin Cosgrave Subject: [IBIS-Users] IBIS Temperature range Hi all, I am generating an IBIS model for a part with a Vdd spec of 2.5 to 5.5 Volts. My question is whether I should generate two IBIS models, one with Voltage Range 4.5, 5, 5.5, and the other with 2.5, 2.7, 3? The reason I ask is that if someone is using a Vdd=5V system, and if I extract the model with voltage range 2.5, 5, 5.5, the min will be well below anything they ever see. Apologies if this question has already been answered, but I can't find anything in the archive. Thanks, Gavin. - -- Gavin Cosgrave, Analog Devices, Building 2200, Cork Airport Business Park, Cork, Ireland. Tel:+353-21-4917906 Internal:6021-7906 |------------------------------------------------------------------ |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 ------------------------------ Date: Wed, 13 Aug 2003 13:59:41 -0400 From: "Mike LaBonte" Subject: RE: [IBIS-Users] IBIS Temperature range In theory an IBIS model produced for one supply voltage works over a range of supply voltages. In practice, you will probably find that producing the same model using s2ibis2 with different supply voltages yields different models. I prefer not to use an IBIS model at any voltage other than the one it was produced for. Mike LaBonte - -----Original Message----- From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On Behalf Of Gavin Cosgrave Sent: Wednesday, August 13, 2003 10:09 AM To: ibis-users@eda.org Subject: [IBIS-Users] IBIS Temperature range Hi all, I am generating an IBIS model for a part with a Vdd spec of 2.5 to 5.5 Volts. My question is whether I should generate two IBIS models, one with Voltage Range 4.5, 5, 5.5, and the other with 2.5, 2.7, 3? The reason I ask is that if someone is using a Vdd=5V system, and if I extract the model with voltage range 2.5, 5, 5.5, the min will be well below anything they ever see. Apologies if this question has already been answered, but I can't find anything in the archive. Thanks, Gavin. - -- Gavin Cosgrave, Analog Devices, Building 2200, Cork Airport Business Park, Cork, Ireland. Tel:+353-21-4917906 Internal:6021-7906 |------------------------------------------------------------------ |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 |------------------------------------------------------------------ |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 ------------------------------ Date: Wed, 13 Aug 2003 16:37:50 -0400 From: "Allen Chao" Subject: RE: [IBIS-Users] IBIS model Error This is a multi-part message in MIME format. - ------=_NextPart_000_0007_01C361B9.3C85EDC0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Hi IBIS experts, I have the same problem as Muthu Lakshmi. I don't really understand how the parser calculated the DC end points from the IV table... If anyone can kindly show me how it works or know where to find the information on the internet, it will be greatly appreciated. Thanks in advance, Allen Chao -----Original Message----- From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On Behalf Of erik.van.der.ven@philips.com Sent: Monday, July 28, 2003 4:10 AM To: KMuthulakshmi@scmmicro.co.in; ibis-users@eda.org Subject: Re: [IBIS-Users] IBIS model Error Hello Muthu Lakshmi, I think there is something wrong with your IBIS model. The DC end points calculated from the I-V tables should match the V-T end points (logic levels). If the error is within 2% you should get a warning, otherwise an error. The difference you mention is very big. The DC end points are calculated from the I-V tables (therefore you don't see the end points of the V-T table). The I-V tables are derived by sweeping a voltage at the output and measuring the resulting current. By applying a certain load the voltages from the logic high and low level are calculated from the pull-up and pull-down curve (and of course they should be near the waveform end point voltages). The V-T data is used to describe the transition between the logic levels calculated from the I-V data. I would suggest that you contact Cypress to ask if this model is OK. With kind regards, Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334 KMuthulakshmi@scmmicro.co.in Sent by: owner-ibis-users@eda.org 07/28/03 05:32 AM To: ibis-users@eda.org cc: (bcc: Erik van der Ven/NYM/SC/PHILIPS) Subject: [IBIS-Users] IBIS model Error Classification: Hi All, I need IBIS expert's advice on the following. I get the following warning when checking the IBIS model. Warning - Model bp_usb_hs: The [Rising Waveform] with [R_fixture]=45 Ohms and [V_fixture]=0V has TYP column DC endpoints of 0.00V and 0.38v, but an equivalent load applied to the model's I-V tables yields different voltages ( 1.14V and 1.06V), a difference of 99.99% and 64.28%, respectively. Warning - Model bp_usb_hs: The [Falling Waveform] with [R_fixture]=45 Ohms and [V_fixture]=0V has TYP column DC endpoints of 0.00V and 0.38v, but an equivalent load applied to the model's I-V tables yields different voltages ( 1.14V and 1.06V), a difference of 99.99% and 64.28%, respectively. Warning - Model bp_usb_hs: The [Rising Waveform] with [R_fixture]=45 Ohms and [V_fixture_min]=0V has MIN column DC endpoints of 0.00V and 0.37v, but an equivalent load applied to the model's I-V tables yields different voltages ( 1.09V and 0.93V), a difference of 99.99% and 60.23%, respectively. Warning - Model bp_usb_hs: The [Falling Waveform] with [R_fixture]=45 Ohms and [V_fixture_min]=0V has MIN column DC endpoints of 0.00V and 0.37v, but an equivalent load applied to the model's I-V tables yields different voltages ( 1.09V and 0.93V), a difference of 99.99% and 60.23%, respectively. Warning - Model bp_usb_hs: The [Rising Waveform] with [R_fixture]=45 Ohms and [V_fixture_max]=0V has MAX column DC endpoints of 0.00V and 0.39v, but an equivalent load applied to the model's I-V tables yields different voltages ( 1.77V and 1.83V), a difference of 100.00% and 78.72%, respectively. Warning - Model bp_usb_hs: The [Falling Waveform] with [R_fixture]=45 Ohms and [V_fixture_max]=0V has MAX column DC endpoints of 0.00V and 0.38v, but an equivalent load applied to the model's I-V tables yields different voltages ( 1.77V and 1.83V), a difference of 100.00% and 79.26%, respectively. Please tell me the impact of this warning in simulation. I use XTK for the simulation. Eventhough the rise and fall VT curve values range from 0V to 0.38V in the IBIS model, I get 1.04V to 1.14V wave form. How is this related with this. I am using Cypress "CY7C65640" hub controller's IBIS model. Awaiting your earliest reply. Note: One IBIS parser gives the above warnings and another parser gives them as Errors. I do not know whether files can be attached with this mail. If it is OK, I can send the IBIS model. thanks & regards K.Muthu Lakshmi (KML) SCM Microsystems (India) Pvt. Ltd. email: KMuthulakshmi@scmmicro.co.in Web: www.scmmicro.com - ------=_NextPart_000_0007_01C361B9.3C85EDC0 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable
Hi=20 IBIS experts,
I have the same problem as Muthu Lakshmi. I don't = really=20 understand how the parser calculated the DC end points from the IV = table...=20 If anyone can kindly show me how it works or know where to find the = information=20 on the internet, it will be greatly = appreciated.
 
Thanks in advance,
 
Allen Chao 
 
 
-----Original Message-----
From: = owner-ibis-users@eda.org=20 [mailto:owner-ibis-users@eda.org]On Behalf Of=20 erik.van.der.ven@philips.com
Sent: Monday, July 28, 2003 = 4:10=20 AM
To: KMuthulakshmi@scmmicro.co.in;=20 ibis-users@eda.org
Subject: Re: [IBIS-Users] IBIS model=20 Error


Hello = Muthu=20 Lakshmi,

I think there = is=20 something wrong with your IBIS model. The DC end points calculated = from the=20 I-V tables should match the V-T end points (logic levels). If the = error is=20 within 2% you should get a warning, otherwise an error. The difference = you=20 mention is very big.

The DC end=20 points are calculated from the I-V tables (therefore you don't see the = end=20 points of the V-T table). The I-V tables are derived by sweeping a = voltage at=20 the output and measuring the resulting current. By applying a certain = load the=20 voltages from the logic high and low level are calculated from the = pull-up and=20 pull-down curve (and of course they should be near the waveform end = point=20 voltages). The V-T data is used to describe the transition between the = logic=20 levels calculated from the I-V data.

I would suggest that you contact Cypress to ask if this model = is=20 OK.


With kind = regards,=20


Ir. Erik van der = Ven
Room=20 DB1032
Business Line Networking Infrastructure
Philips = Semiconductors=20 BV
Gerstweg 2
6534 AE Nijmegen
The Netherlands
Phone:=20 +31-24-3534334







KMuthulakshmi@scmmicro.co.in=20

Sent by:
owner-ibis-users@eda.org=20

07/28/03 05:32 AM =

        =
        To: =    =20    ibis-users@eda.org
        cc:       =  (bcc:=20 Erik van der Ven/NYM/SC/PHILIPS)
        Subject:     =  =20  [IBIS-Users] IBIS model Error=20

       =20 Classification:        =20





Hi=20 All,

I need IBIS expert's advice on the=20 following. =

I get the following warning when = checking the IBIS=20 model.

Warning - Model bp_usb_hs: = The [Rising=20 Waveform]
     with [R_fixture]=3D45 Ohms and=20 [V_fixture]=3D0V =
     has = TYP column DC=20 endpoints of  0.00V and  0.38v, but
     an equivalent load applied to the = model's I-V=20 tables yields =
    =  different voltages (=20 1.14V and  1.06V), =
    =  a=20 difference of 99.99% and 64.28%, respectively.

Warning - Model bp_usb_hs: The [Falling Waveform] =
   =20  with [R_fixture]=3D45 Ohms and [V_fixture]=3D0V
     has TYP column DC endpoints of =  0.00V and=20  0.38v, but =
     an = equivalent load=20 applied to the model's I-V tables yields
   =20  different voltages ( 1.14V and  1.06V),
     a difference of 99.99% and 64.28%,=20 respectively. =

Warning - Model bp_usb_hs: = The [Rising=20 Waveform]
     with [R_fixture]=3D45 Ohms and=20 [V_fixture_min]=3D0V =
     has = MIN column DC=20 endpoints of  0.00V and  0.37v, but
     an equivalent load applied to the = model's I-V=20 tables yields =
    =  different voltages (=20 1.09V and  0.93V), =
    =  a=20 difference of 99.99% and 60.23%, respectively.

Warning - Model bp_usb_hs: The [Falling Waveform] =
   =20  with [R_fixture]=3D45 Ohms and [V_fixture_min]=3D0V
     has MIN column DC endpoints of =  0.00V and=20  0.37v, but =
     an = equivalent load=20 applied to the model's I-V tables yields
   =20  different voltages ( 1.09V and  0.93V),
     a difference of 99.99% and 60.23%,=20 respectively. =

Warning - Model bp_usb_hs: = The [Rising=20 Waveform]
     with [R_fixture]=3D45 Ohms and=20 [V_fixture_max]=3D0V =
     has = MAX column DC=20 endpoints of  0.00V and  0.39v, but
     an equivalent load applied to the = model's I-V=20 tables yields =
    =  different voltages (=20 1.77V and  1.83V), =
    =  a=20 difference of 100.00% and 78.72%, respectively.

Warning - Model bp_usb_hs: The [Falling Waveform] =
   =20  with [R_fixture]=3D45 Ohms and [V_fixture_max]=3D0V
     has MAX column DC endpoints of =  0.00V and=20  0.38v, but =
     an = equivalent load=20 applied to the model's I-V tables yields
   =20  different voltages ( 1.77V and  1.83V),
     a difference of 100.00% and 79.26%,=20 respectively. =

Please tell me the impact of this = warning in=20 simulation. I use XTK for the simulation. Eventhough the rise and fall = VT=20 curve values range from 0V to 0.38V in the IBIS model, I get 1.04V to = 1.14V=20 wave form. How is this related with this. I am using Cypress = "CY7C65640" hub=20 controller's IBIS model.=20

Awaiting your earliest = reply.

Note: One IBIS parser gives the above = warnings and=20 another parser gives them as Errors.
    =    I=20 do not know whether files can be attached with this mail. If it is OK, = I can=20 send  the IBIS model.=20

thanks & = regards
K.Muthu=20 Lakshmi (KML)

SCM Microsystems (India) Pvt. Ltd.
email:=20 KMuthulakshmi@scmmicro.co.in
Web: www.scmmicro.com =

- ------=_NextPart_000_0007_01C361B9.3C85EDC0-- |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 14 Aug 2003 08:36:05 +0200 From: Pascal GILGENKRANTZ Subject: Re: [IBIS-Users] IBIS model Error Hello Allen, I ran into the same a few weeks ago. I found the answer in the free "Visual IBIS Editor" (From Innoveda) in the help menu which is very well explained. -> Help Menu -> Verifying IBIS Models -> V/t and V/I Table Data are Mismatched Available from Mentor Web site : http://www.mentor.com/hyperlynx/visibis.cfm Pascal Gilgenkrantz STMicroelectronics > achao@synopsys.com wrote: > > Hi IBIS experts, > I have the same problem as Muthu Lakshmi. I don't really understand how the > parser calculated the DC end points from the IV table... If anyone can > kindly show me how it works or know where to find the information on the > internet, it will be greatly appreciated. > > 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 | 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 ------------------------------ Date: Thu, 14 Aug 2003 12:28:56 +0530 From: "Birendra Rana" Subject: RE: [IBIS-Users] IBIS model Error This is a multi-part message in MIME format. - ------_=_NextPart_001_01C36231.87B06F10 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Hi, =20 Such errors are encountered when the IBIS model is not generated properly. One reason of getting this kind of error, as per my experience, is when the ramp tables haven't captured the full waveform. The ibischk utility takes the 'not fully settled' final value of the ramp as the stable end value and compares with the IV table DC end point and flags an error or warning depending on how big is the difference. Because of this reason I have seen errors/warnings normally flagged for one end point. =20 Muthu Lakshmi's error reports however have large difference at the two end points which means the problem lies else where.=20 =20 Regards, =20 Birendra - -----Original Message----- From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org] On Behalf Of Allen Chao Sent: Thursday, 14 August 2003 2:08 AM To: erik.van.der.ven@philips.com; KMuthulakshmi@scmmicro.co.in; ibis-users@eda.org Subject: RE: [IBIS-Users] IBIS model Error Hi IBIS experts, I have the same problem as Muthu Lakshmi. I don't really understand how the parser calculated the DC end points from the IV table... If anyone can kindly show me how it works or know where to find the information on the internet, it will be greatly appreciated. =20 Thanks in advance, =20 Allen Chao=20 =20 =20 - -----Original Message----- From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On Behalf Of erik.van.der.ven@philips.com Sent: Monday, July 28, 2003 4:10 AM To: KMuthulakshmi@scmmicro.co.in; ibis-users@eda.org Subject: Re: [IBIS-Users] IBIS model Error Hello Muthu Lakshmi,=20 I think there is something wrong with your IBIS model. The DC end points calculated from the I-V tables should match the V-T end points (logic levels). If the error is within 2% you should get a warning, otherwise an error. The difference you mention is very big.=20 The DC end points are calculated from the I-V tables (therefore you don't see the end points of the V-T table). The I-V tables are derived by sweeping a voltage at the output and measuring the resulting current. By applying a certain load the voltages from the logic high and low level are calculated from the pull-up and pull-down curve (and of course they should be near the waveform end point voltages). The V-T data is used to describe the transition between the logic levels calculated from the I-V data.=20 I would suggest that you contact Cypress to ask if this model is OK.=20 With kind regards,=20 Ir. Erik van der Ven Room DB1032 Business Line Networking Infrastructure Philips Semiconductors BV Gerstweg 2 6534 AE Nijmegen The Netherlands Phone: +31-24-3534334=20 =09 KMuthulakshmi@scmmicro.co.in=20 Sent by:=20 owner-ibis-users@eda.org=20 07/28/03 05:32 AM=20 =20 To: ibis-users@eda.org=20 cc: (bcc: Erik van der Ven/NYM/SC/PHILIPS)=20 Subject: [IBIS-Users] IBIS model Error=20 Classification: =20 Hi All,=20 I need IBIS expert's advice on the following.=20 I get the following warning when checking the IBIS model.=20 Warning - Model bp_usb_hs: The [Rising Waveform]=20 with [R_fixture]=3D45 Ohms and [V_fixture]=3D0V=20 has TYP column DC endpoints of 0.00V and 0.38v, but=20 an equivalent load applied to the model's I-V tables yields=20 different voltages ( 1.14V and 1.06V),=20 a difference of 99.99% and 64.28%, respectively.=20 Warning - Model bp_usb_hs: The [Falling Waveform]=20 with [R_fixture]=3D45 Ohms and [V_fixture]=3D0V=20 has TYP column DC endpoints of 0.00V and 0.38v, but=20 an equivalent load applied to the model's I-V tables yields=20 different voltages ( 1.14V and 1.06V),=20 a difference of 99.99% and 64.28%, respectively.=20 Warning - Model bp_usb_hs: The [Rising Waveform]=20 with [R_fixture]=3D45 Ohms and [V_fixture_min]=3D0V=20 has MIN column DC endpoints of 0.00V and 0.37v, but=20 an equivalent load applied to the model's I-V tables yields=20 different voltages ( 1.09V and 0.93V),=20 a difference of 99.99% and 60.23%, respectively.=20 Warning - Model bp_usb_hs: The [Falling Waveform]=20 with [R_fixture]=3D45 Ohms and [V_fixture_min]=3D0V=20 has MIN column DC endpoints of 0.00V and 0.37v, but=20 an equivalent load applied to the model's I-V tables yields=20 different voltages ( 1.09V and 0.93V),=20 a difference of 99.99% and 60.23%, respectively.=20 Warning - Model bp_usb_hs: The [Rising Waveform]=20 with [R_fixture]=3D45 Ohms and [V_fixture_max]=3D0V=20 has MAX column DC endpoints of 0.00V and 0.39v, but=20 an equivalent load applied to the model's I-V tables yields=20 different voltages ( 1.77V and 1.83V),=20 a difference of 100.00% and 78.72%, respectively.=20 Warning - Model bp_usb_hs: The [Falling Waveform]=20 with [R_fixture]=3D45 Ohms and [V_fixture_max]=3D0V=20 has MAX column DC endpoints of 0.00V and 0.38v, but=20 an equivalent load applied to the model's I-V tables yields=20 different voltages ( 1.77V and 1.83V),=20 a difference of 100.00% and 79.26%, respectively.=20 Please tell me the impact of this warning in simulation. I use XTK for the simulation. Eventhough the rise and fall VT curve values range from 0V to 0.38V in the IBIS model, I get 1.04V to 1.14V wave form. How is this related with this. I am using Cypress "CY7C65640" hub controller's IBIS model.=20 Awaiting your earliest reply.=20 Note: One IBIS parser gives the above warnings and another parser gives them as Errors.=20 I do not know whether files can be attached with this mail. If it is OK, I can send the IBIS model.=20 thanks & regards K.Muthu Lakshmi (KML) SCM Microsystems (India) Pvt. Ltd. email: KMuthulakshmi@scmmicro.co.in Web: www.scmmicro.com=20 - ------_=_NextPart_001_01C36231.87B06F10 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Message
Hi,
 
Such errors are encountered when the IBIS model is not = generated=20 properly. One reason of getting this kind of error, as per my = experience,=20 is when the ramp tables haven't captured the full waveform. The = ibischk=20 utility takes the 'not fully settled' final value of the ramp as = the stable=20 end value and compares with the IV table DC end point and flags an error = or=20 warning depending on how big is the difference. Because of = this reason=20 I have seen errors/warnings normally flagged for one end=20 point.
 
Muthu Lakshmi's error reports however have large = difference at=20 the two end points which means the problem lies else where.=20
 
Regards,
 
Birendra
-----Original Message-----
From:=20 owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org] On = Behalf Of=20 Allen Chao
Sent: Thursday, 14 August 2003 2:08 = AM
To:=20 erik.van.der.ven@philips.com; KMuthulakshmi@scmmicro.co.in;=20 ibis-users@eda.org
Subject: RE: [IBIS-Users] IBIS model=20 Error

Hi=20 IBIS experts,
I have the same problem as Muthu Lakshmi. I don't = really=20 understand how the parser calculated the DC end points from the = IV=20 table... If anyone can kindly show me how it works or know where to = find the=20 information on the internet, it will be greatly=20 appreciated.
 
Thanks in advance,
 
Allen Chao 
 
 
-----Original Message-----
From: = owner-ibis-users@eda.org=20 [mailto:owner-ibis-users@eda.org]On Behalf Of=20 erik.van.der.ven@philips.com
Sent: Monday, July 28, = 2003 4:10=20 AM
To: KMuthulakshmi@scmmicro.co.in;=20 ibis-users@eda.org
Subject: Re: [IBIS-Users] IBIS model=20 Error


Hello = Muthu=20 Lakshmi,

I think = there is=20 something wrong with your IBIS model. The DC end points calculated = from the=20 I-V tables should match the V-T end points (logic levels). If the = error is=20 within 2% you should get a warning, otherwise an error. The = difference you=20 mention is very big.

The DC end=20 points are calculated from the I-V tables (therefore you don't see = the end=20 points of the V-T table). The I-V tables are derived by sweeping a = voltage=20 at the output and measuring the resulting current. By applying a = certain=20 load the voltages from the logic high and low level are calculated = from the=20 pull-up and pull-down curve (and of course they should be near the = waveform=20 end point voltages). The V-T data is used to describe the transition = between=20 the logic levels calculated from the I-V data.

I would suggest that you contact Cypress = to ask if=20 this model is OK.


With kind=20 regards,


Ir. = Erik van der=20 Ven
Room DB1032
Business Line Networking = Infrastructure
Philips=20 Semiconductors BV
Gerstweg 2
6534 AE Nijmegen
The=20 Netherlands
Phone: +31-24-3534334







KMuthulakshmi@scmmicro.co.in=20

Sent by:
owner-ibis-users@eda.org=20

07/28/03 05:32 AM =

       =20
    =    =20 To:        ibis-users@eda.org =
        cc: =    =20    (bcc: Erik van der Ven/NYM/SC/PHILIPS) =
        = Subject:  =20      [IBIS-Users] IBIS model Error=20

      =  =20 Classification:        =20





Hi All,=20

I need IBIS expert's = advice on=20 the following. =

I get the following warning when = checking the=20 IBIS model.

Warning - Model = bp_usb_hs: The [Rising=20 Waveform]
     with [R_fixture]=3D45 Ohms and=20 [V_fixture]=3D0V =
     has = TYP column DC=20 endpoints of  0.00V and  0.38v, but
     an equivalent load applied to the = model's I-V=20 tables yields =
    =  different voltages=20 ( 1.14V and  1.06V),=20
  =    a=20 difference of 99.99% and 64.28%, respectively.

Warning - Model bp_usb_hs: The [Falling Waveform] =
 =20    with [R_fixture]=3D45 Ohms and = [V_fixture]=3D0V
     has TYP column DC endpoints of =  0.00V=20 and  0.38v, but =
     an = equivalent load=20 applied to the model's I-V tables yields
   =20  different voltages ( 1.14V and  1.06V),
     a difference of 99.99% and 64.28%,=20 respectively. =

Warning - Model = bp_usb_hs: The [Rising=20 Waveform]
     with [R_fixture]=3D45 Ohms and=20 [V_fixture_min]=3D0V =
     has = MIN column DC=20 endpoints of  0.00V and  0.37v, but
     an equivalent load applied to the = model's I-V=20 tables yields =
    =  different voltages=20 ( 1.09V and  0.93V),=20
  =    a=20 difference of 99.99% and 60.23%, respectively.

Warning - Model bp_usb_hs: The [Falling Waveform] =
 =20    with [R_fixture]=3D45 Ohms and = [V_fixture_min]=3D0V
     has MIN column DC endpoints of =  0.00V=20 and  0.37v, but =
     an = equivalent load=20 applied to the model's I-V tables yields
   =20  different voltages ( 1.09V and  0.93V),
     a difference of 99.99% and 60.23%,=20 respectively. =

Warning - Model = bp_usb_hs: The [Rising=20 Waveform]
     with [R_fixture]=3D45 Ohms and=20 [V_fixture_max]=3D0V =
     has = MAX column DC=20 endpoints of  0.00V and  0.39v, but
     an equivalent load applied to the = model's I-V=20 tables yields =
    =  different voltages=20 ( 1.77V and  1.83V),=20
  =    a=20 difference of 100.00% and 78.72%, respectively.

Warning - Model bp_usb_hs: The [Falling Waveform] =
 =20    with [R_fixture]=3D45 Ohms and = [V_fixture_max]=3D0V
     has MAX column DC endpoints of =  0.00V=20 and  0.38v, but =
     an = equivalent load=20 applied to the model's I-V tables yields
   =20  different voltages ( 1.77V and  1.83V),
     a difference of 100.00% and 79.26%, = respectively. =

Please tell me the impact of this = warning in=20 simulation. I use XTK for the simulation. Eventhough the rise and = fall VT=20 curve values range from 0V to 0.38V in the IBIS model, I get 1.04V = to 1.14V=20 wave form. How is this related with this. I am using Cypress = "CY7C65640" hub=20 controller's IBIS model.=20

Awaiting your = earliest=20 reply. =

Note: One IBIS parser gives the above = warnings=20 and another parser gives them as Errors.
    =    I=20 do not know whether files can be attached with this mail. If it is = OK, I can=20 send  the IBIS model.=20

thanks & = regards
K.Muthu=20 Lakshmi (KML)

SCM Microsystems (India) Pvt. Ltd.
email:=20 KMuthulakshmi@scmmicro.co.in
Web: www.scmmicro.com=20

- ------_=_NextPart_001_01C36231.87B06F10-- |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 14 Aug 2003 08:27:15 -0600 From: rrwolff@micron.com Subject: [IBIS-Users] [IBIS] Open forum minutes (08/08/03) DATE: 8/14/03 SUBJECT: August 8, 2003 EIA IBIS Open Forum Meeting Minutes VOTING MEMBERS AND 2003 PARTICIPANTS LIST: Ansoft Corporation (Eric Bracken) Apple Computer Kim Helliwell Applied Simulation Technology Fred Balistreri Cadence Design Lynne Green*, Lance Wang Cisco Systems Syed Huq, Michael LaBonte, Val Mandruson, Hung Pham, Todd Westerhoff Fairchild Semiconductor (Graham Connolly) Hitachi ULSI Systems Kazuyoshi Shoji Intel Corporation Stephen Peters, Michael Mirmak*, Arpad Muranyi*, Eric Magnusson LSI Logic Frank Gasparik* Matsushita (Panasonic) Atsuji Ito Mentor Graphics [Bob Ross], Ian Dodd, Guy de Burgh*, John Angulo*, Mike Donnelly, Weston Beal Micron Technology Randy Wolff* Molex Incorporated Gus Panella Motorola (Rick Kingen) National Semiconductor [Milt Schwartz], [Tim Coyle] NEC Electric Corporation (Itsuki Yamada) Philips Semiconductor (D.C. Sessions), Stephanie Goedecke Siemens (& Automotive) AG Eckhard Lenski, Michael Kindij, Burkhard Muller, Katja Koller, Andre Goerisch, Manfred Maurer, Bernard Unger, Amir Motamedi, Hartmut Ibowski, Gerald Bannert Signal Integrity Software Bob Haller, Barry Katz, Doug Burns Sigrity [Raj Raghuram] Synopsys Warren Wong, Edmund Cheng Texas Instruments Thomas Fisher Teraspeed Scott McMorrow, Tom Dagostino, Kevin Simpson, Bob Ross* Time Domain Analysis Systems Dima Smolyansky, Steve Corey Zuken (& Incases) Michael Schaeder, Ralf Bruening OTHER PARTICIPANTS IN 2003: Agilent Technologies Herbert Lage Brocade Frank Yuan, Yongrue Yu Conexant Gary Felker EADS CCR Alix de la Villeguerin EFM Ekkehard Miersch EMC Corporation Brian Arsenault, John Fernandez, Simba Julian Fraunhofer IZM Ege Engin Fujitsu Tadashi Arai GEIA (Chris Denham) Huawei Technologies (Jiang Xiang Zhong) Independent Kelly Green, Luca Giacotto Infineon Tech AG Christian Sporrer Marvell Semiconductor Itzik Peleg NetLogic Microsystems Eric Hsu North East Systems Associates Edward Sayre North Carolina State Univ. Ambrish Varma* Plexus Joseph Socha Politechnico de Torino Igor Stievano Qlogic Larry Barnes Sintecs BV Hans Klos, Bob te Nijenhuis SiQual (Rob Hinz) Sun Microsystems Tim Coyle* Via Technologies (Weber Chuang) Xilinx Susan Wu In the list above, attendees at the meeting are indicated by *. Principal members or other active members who have not attended are in parentheses. Participants who no longer are in the organization are in square brackets. Upcoming Meetings: The bridge numbers for future IBIS teleconferences are as follows: Date Bridge Number Reservation # Passcode August 22, 2003 1-916-356-2663 4 696-0504 All meetings are 8:00 AM to 9:55 AM Pacific Time. Meeting agendas are typically distributed seven days before each Open Forum. Minutes are typically distributed within seven days of the corresponding meeting. When calling into the meeting, provide the reservation number and passcode at the automated prompts. If asked by an operator, please request to join the IBIS Open Forum hosted by Michael Mirmak. NOTE: "AR" = Action Required. - -------------------------------- MINUTES ----------------------------------- INTRODUCTIONS AND MEETING QUORUM Ambrish Varma of North Carolina State University introduced himself. He is a Ph. D. candidate currently involved in research related to IBIS and is developing the S2IBIS3 IBIS model generation software. CALL FOR PATENTS Michael Mirmak called for any patents or pending patents related to the IBIS Version 4.0, 4.1 or ICM specifications. No companies declared a patent. MEMBERSHIP UPDATE AND TREASURER'S REPORT Michael Mirmak reported that membership remains at 25 organizations. A minor problem with one company's GEIA invoice has been resolved and all payments are expected to be complete shortly. REVIEW OF MINUTES AND AR'S The July 18 meeting minutes were approved without objection. MISCELLANY/ANNOUNCEMENTS No new announcements. PRESS AND WEB PAGE UPDATES Bob Ross reported that Syed has completed a first revision of the on-line IBIS roster. Most member companies and organizations have provided updated information. NEW MODELS AVAILABLE, LIBRARY UPDATE No update. OPENS FOR NEW ISSUES Tim Coyle raised some concerns regarding the IBIS web pages. While the information there is useful, the most recent article is from 2000 and none of the information is searchable. More current technical and application notes from specific companies are not available. Tim proposed more active cross-referencing of papers and other company- specific links, if not creating a new domain for the IBIS community. Bob Ross noted that the EIA controls the eigroup.org web site and that access and updating is limited for security reasons. Michael requested that Tim work with webmaster Syed Huq on either updating the current site per Tim's suggestions, or creating a new web domain. INTERNATIONAL/EXTERNAL PROGRESS - - IEC 62014-3 (ICEM) Integrated Circuit Electromagnetic Model Proposal (formerly, IEC 93/67/NP IBIS and EMC Simulation) No update. - - JEDEC Activity Report. No update. DATE/PCB EAST SUMMIT DISCUSSION Lynne Green reported that a draft call for papers has been prepared and is ready for distribution in mid-August. Michael Mirmak described his conversation with Ed Sayre of NESA and reported that NESA has no objection to hosting a summit at NESA on Tuesday, October 21. Confirmation of the details has not yet been received, however. Michael accepted the AR to confirm room and refreshment/lunch arrangements with Kathy Breda of NESA. DESIGNCON SUMMIT DISCUSSION Michael Mirmak reported that he spoke with Brock Silvey of DesignCon 2004. The contract and logistical arrangements for an IBIS Summit at DesignCon 2004 are nearly complete, though some issues regarding refreshments still remain to be resolved. A draft co-sponsorship agreement has been received and is being reviewed by the Open Forum officers. The summit is tentatively scheduled for Monday, Feb. 2, 2004. IBIS VERSION 4.0 PARSER FUNDING AND STATUS Michael Mirmak reported that all invoices for the 4.0 parser have been distributed and confirmed received. Payment from one remaining organization is expected shortly. A beta version of IBISCHK4 has been compiled and is available for download: http://www.vhdl.org/pub/ibis/ibischk4_beta IBIS QUALITY COMMITTEE Bob Ross reported that meetings continue every two weeks. IBIS MODEL REVIEW COMMITTEE Lynne Green reported that no new models have been received by the committee since the last Open Forum meeting. NEW ADMINISTRATIVE ISSUES No new issues. S2IBIS3 Ambrish Varma noted that he has created a publicly-available website for the S2IBIS3 project: http://www.ece.ncsu.edu/erl/ibis/s2ibis3/s2ibis3.htm Feedback on the current beta version is appreciated and encouraged. Ambrish also mentioned that he and Bob Ross have been recently working on debugging support of the [Series MOSFET] keyword. The current version of the software supports, under Java, all the S2IBIS2 version 2.1 commands except [Summarize]. Previously noted issues with print resolution have been addressed and more comprehensive version control will be implemented. BIRD74.5 - EMI Parameters Guy de Burgh introduced the updated text of BIRD74.5. He described six minor changes suggested to him via e-mail since the last review; these include: - - explicitly stating that the "Domain" and "Cpd" subparameters and arguments are not required and may be omitted - - changing the phrase "The keyword takes one argument" under "Heatsink_cap" to "The sub-parameter takes one argument" - - requiring the "domain_name" column for [Pin EMI] to help the parsing of the columns, even if the data is NA - - using the word "maximum" in place of "dominant" for the description of the clock division sub-parameter - - clarifying that [Begin EMI Model] and [End EMI Model] must occur after the [Model] keyword with which they are associated; similar rules would apply to EMI keywords related to [Component] - - a global change of "sub-parameter" to "subparameter" Guy accepted all suggested changes and noted that these do not affect the technical content of the BIRD. Some discussion arose regarding the version in which this BIRD is to be included. Bob Ross suggested that the BIRD become part of the next major release, likely to be IBIS 5.0. The BIRD was approved as amended by a vote of four in favor with one opposed. The text will be posted with changes as BIRD74.6. BIRD82.1 - Clarification of Clamp Tables Bob Ross noted that BIRD82.1 has been posted and distributed for consideration at the next IBIS Open Forum meeting. IBIS VERSION 4.1 DISCUSSION A discussion was held regarding C_comp and [Ramp], and their relationship to the multi-lingual extensions proposed as part of IBIS 4.1. Lynne Green and John Angulo noted that several tools use the [Ramp] keyword as an approximate slew rate, for efficient automated prioritization of SI problems. This data is also useful for setting initial simulation engine timesteps and therefore improve simulation efficiency. C_comp was considered less strongly needed as a required parameter. At present, both keywords are considered optional when multi-lingual extensions are correctly used in an IBIS file. Arpad Muranyi observed that [Ramp] data, being single-ended, may not apply for differential waveforms or may imply behaviors outside of the normal operation of the device being modeled. John Angulo proposed that a BIRD be written to include a new keyword, [Slew Time] or similar, which would enable tools to set initial timesteps without requiring the [Ramp] keyword. Open issues include whether this is a subparameter under the [External Model] and/or [External Circuit] keywords and whether both [Ramp] and [Slew Time] may exist in the same model. John will provide draft BIRD language at a future Open Forum meeting. A draft version of IBIS version 4.1 is available on the web. Both an ASCII text and an Adobe PDF version are available at: http://www.vhdl.org/pub/ibis/wip/ IBIS ICM SPECIFICATION REVIEW Michael Mirmak reintroduced the ICM specification and called for comment. He mentioned that, with most comments and feedback noted, a vote on approval will be scheduled for Sept. 12. IBISCHK4 & BUG TRACKING STATUS Bob Ross reminded the Open Forum that a beta version of IBISCHK4 is available at: http://www.vhdl.org/pub/ibis/ibischk4_beta Bob also reported that all BUGS except 68, 71, 73, 75 and 76 have been addressed and fixed in the beta IBISCHK4 parser code. These are now listed as "Closed" on the IBIS web site. BUG80, which has not been classified, retains its 3.2 behavior in the beta code. John Angulo noted that he has compared the Windows output running the test cases sent by Atul Agarwal against Atul's reference outputs, and found only minor differences. Further work will be completed separately by Bob, John, and Michael Mirmak. BUG80 - [Model Selector] and [Diff Pin] Ordering Classification of BUG80 was postponed until the next Open Forum meeting. NEXT MEETINGS: By acclamation, the next teleconference has been scheduled for August 22, one week earlier than usual, to accommodate the upcoming US holiday. The meeting will be held from 8:00 AM to 10:00 AM Pacific Daylight Time. ============================================================================ NOTES IBIS CHAIR: Michael Mirmak (916) 356-4261, Fax: (916) 377-1046 michael.mirmak@intel.com Senior Analog Engineer, Intel Corporation M/S FM6-45 1900 Prairie City Rd. Folsom, CA 95630 VICE CHAIR: Lynne Green (425) 788-0412, Fax (425) 451-1871 lgreen@cadence.com Senior Modeling Engineer, Cadence Design Systems 320 120th Ave NE, Suite B-103, Bellevue, WA 98005-3016 SECRETARY: Randy Wolff (208) 363-1764, Fax: (208) 368-3475 rrwolff@micron.com Simulation Engineer, Micron Technology, Inc. 8000 S. Federal Way Mail Stop: 1-711 Boise, ID 83707-0006 LIBRARIAN: Roy Leventhal (847) 590-9398 roy.leventhal@ieee.org Consultant, Leventhal Design and Communications 1924 North Burke Drive Arlington Heights, Illinois 60004 WEBMASTER: Syed Huq (408) 525-3399, Fax: (408) 526-5504 shuq@cisco.com Manager, Hardware Engineering, Cisco Systems 170 West Tasman Drive San Jose, CA 95134-1706 POSTMASTER: John Angulo (425) 497-5077, Fax: (425) 881-1008 John_angulo@mentor.com Development Engineer, Mentor Graphics 14715 N.E. 95th Street, Suite 200 Redmond, WA 98052 This meeting was conducted in accordance with the EIA Legal Guides and EIA Manual of Organization and Procedure. The following e-mail addresses are used: majordomo@eda.org In the body, for the IBIS Open Forum Reflector: subscribe ibis In the body, for the IBIS Users' Group Reflector: subscribe ibis-users Help and other commands: help ibis-request@eda.org To join, change, or drop from either the IBIS Open Forum Reflector (ibis@eda.org), the IBIS Users' Group Reflector (ibis-users@eda.org) or both. State your request. ibis-info@eda.org To obtain general information about IBIS, to ask specific questions for individual response, and to inquire about joining the EIA-IBIS Open Forum as a full Member. ibis@eda.org To send a message to the general IBIS Open Forum Reflector. This is used mostly for IBIS Standardization business and future IBIS technical enhancements. Job posting information is not permitted. ibis-users@eda.org To send a message to the IBIS Users' Group Reflector. This is used mostly for IBIS clarification, current modeling issues, and general user concerns. Job posting information is not permitted. ibischk-bug@eda.org To report ibischk2/3 parser bugs. The Bug Report Form Resides on eda.org in /pub/ibis/bugs/ibischk/bugform.txt along with reported bugs. To report s2ibis, s2ibis2 and s2iplt bugs, use the Bug Report Forms which reside under eda.org in /pub/ibis/bugs/s2ibis/bugs2i.txt, /pub/ibis/bugs/s2ibis2/bugs2i2.txt, and /pub/ibis/bugs/s2iplt/bugsplt.txt respectively. Information on IBIS technical contents, IBIS participants, and actual IBIS models are available on the IBIS Home page found by selecting the Electronic Information Group under: http://www.eigroup.org/ibis/ibis.htm Check the pub/ibis directory on eda.org for more information on previous discussions and results. Logins using anonymous FTP are accepted. |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 14 Aug 2003 15:50:37 -0400 From: "Allen Chao" Subject: [IBIS-Users]DC Sweep for the pullup IV curve Hi IBIS experts, I have a question concerning the sweep. I have a 3.3V power supply for my device. However, the logic high voltage is 0.4V and logic low is 0V for my one of my outputs. Do I do the DC Sweep at the output from -3.3V to 6.6V or -0.4V to 0.8V? Thanks a lot for the help, 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 | 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 ------------------------------ Date: Thu, 14 Aug 2003 16:06:36 -0700 From: "Muranyi, Arpad" Subject: RE: [IBIS-Users]DC Sweep for the pullup IV curve Allen, Strictly speaking the spec says the following: | The stated voltage ranges for I-V tables cover the most common, single | supply cases. When multiple supplies are specified, the voltages shall | extend similarly to values that handle practical extremes in reflected wave | simulations. | ... | 2) Voltage Ranges: | Points for each table must span the voltages listed below: | | Table Low Voltage High Voltage | ----------- ----------- ------------ | [Pulldown] GND - POWER POWER + POWER | [Pullup] GND - POWER POWER + POWER | [GND Clamp] GND - POWER GND + POWER | [POWER Clamp] POWER POWER + POWER | [Series Current] GND - POWER GND + POWER | [Series MOSFET] GND GND + POWER | | As described in the [Pulldown Reference] keyword section, the I-V tables | of the [Pullup] and the [POWER Clamp] structures are 'Vcc relative', | using the equation: Vtable = Vcc - Voutput. | | For example, a model with a 5 V power supply voltage should be | characterized between (0 - 5) = -5 V and (5 + 5) = 10 V; and a model with | a 3.3 V power supply should be characterized between (0 - 3.3) = -3.3 V | and (3.3 + 3.3) = 6.6 V for the [Pulldown] table. | | When tabulating output data for ECL type models, the voltage points must | span the range of Vcc to Vcc - 2.2 V. This range applies to both the | [Pullup] and [Pulldown] tables. Note that this range applies ONLY when | characterizing an ECL output. | | 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. So according to the spec you would need to do -3.3 to 6.6 V. However, this was written in the days when most buffers were switching rail to rail, and most transmission lines were unterminated having a full reflection. The spec wanted to make sure that the IV curves would cover the voltage range even if the waveforms double due to reflections. Lately, the picture has changed, and we have more low voltage swing signals and parallel terminated buses. As a consequence buffers do not necessarily switch rail to rail, and reflections are much smaller. Even if we followed the logic of full reflections, your model should be sufficient if it covered - -0.4 to 0.8 V. This would be officially not according to the spec, but as far as I remember, the parser doesn't issue an error message, so you should be OK. I am dealing with similar situations in other cases too, and I am actually considering writing a BIRD on this subject to clarify/change the spec. Arpad Muranyi Intel Corporation =========================================================================== - -----Original Message----- From: Allen Chao [mailto:achao@synopsys.com] Sent: Thursday, August 14, 2003 12:51 PM To: ibis-users@eda.org Subject: [IBIS-Users]DC Sweep for the pullup IV curve Hi IBIS experts, I have a question concerning the sweep. I have a 3.3V power supply for my device. However, the logic high voltage is 0.4V and logic low is 0V for my one of my outputs. Do I do the DC Sweep at the output from -3.3V to 6.6V or -0.4V to 0.8V? Thanks a lot for the help, 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 | 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 |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 14 Aug 2003 16:28:15 -0700 From: "Tom Dagostino" Subject: RE: [IBIS-Users]DC Sweep for the pullup IV curve This is a multi-part message in MIME format. - ------=_NextPart_000_02C3_01C36281.10C3DA80 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit One thing you need to consider. This part (which I have no knowledge of) may only swing from 0 to 0.4 but if it is placed with another part on the same net that can swing further you should cover a wider region in your model. BTW, what logic family is this? Tom Dagostino Teraspeed Consulting Group LLC Teraspeed Consulting Group LLC 2926 SE Yamhill St. Device Modeling Division Portland, OR 97214 13610 SW Harness Lane Beaverton, OR 97008 http://www.teraspeed.com 503-430-1065 tom@teraspeed.com - -----Original Message----- From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On Behalf Of Muranyi, Arpad Sent: Thursday, August 14, 2003 4:07 PM To: ibis-users@eda.org Subject: RE: [IBIS-Users]DC Sweep for the pullup IV curve Allen, Strictly speaking the spec says the following: | The stated voltage ranges for I-V tables cover the most common, single | supply cases. When multiple supplies are specified, the voltages shall | extend similarly to values that handle practical extremes in reflected wave | simulations. | ... | 2) Voltage Ranges: | Points for each table must span the voltages listed below: | | Table Low Voltage High Voltage | ----------- ----------- ------------ | [Pulldown] GND - POWER POWER + POWER | [Pullup] GND - POWER POWER + POWER | [GND Clamp] GND - POWER GND + POWER | [POWER Clamp] POWER POWER + POWER | [Series Current] GND - POWER GND + POWER | [Series MOSFET] GND GND + POWER | | As described in the [Pulldown Reference] keyword section, the I-V tables | of the [Pullup] and the [POWER Clamp] structures are 'Vcc relative', | using the equation: Vtable = Vcc - Voutput. | | For example, a model with a 5 V power supply voltage should be | characterized between (0 - 5) = -5 V and (5 + 5) = 10 V; and a model with | a 3.3 V power supply should be characterized between (0 - 3.3) = -3.3 V | and (3.3 + 3.3) = 6.6 V for the [Pulldown] table. | | When tabulating output data for ECL type models, the voltage points must | span the range of Vcc to Vcc - 2.2 V. This range applies to both the | [Pullup] and [Pulldown] tables. Note that this range applies ONLY when | characterizing an ECL output. | | 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. So according to the spec you would need to do -3.3 to 6.6 V. However, this was written in the days when most buffers were switching rail to rail, and most transmission lines were unterminated having a full reflection. The spec wanted to make sure that the IV curves would cover the voltage range even if the waveforms double due to reflections. Lately, the picture has changed, and we have more low voltage swing signals and parallel terminated buses. As a consequence buffers do not necessarily switch rail to rail, and reflections are much smaller. Even if we followed the logic of full reflections, your model should be sufficient if it covered - -0.4 to 0.8 V. This would be officially not according to the spec, but as far as I remember, the parser doesn't issue an error message, so you should be OK. I am dealing with similar situations in other cases too, and I am actually considering writing a BIRD on this subject to clarify/change the spec. Arpad Muranyi Intel Corporation =========================================================================== - -----Original Message----- From: Allen Chao [mailto:achao@synopsys.com] Sent: Thursday, August 14, 2003 12:51 PM To: ibis-users@eda.org Subject: [IBIS-Users]DC Sweep for the pullup IV curve Hi IBIS experts, I have a question concerning the sweep. I have a 3.3V power supply for my device. However, the logic high voltage is 0.4V and logic low is 0V for my one of my outputs. Do I do the DC Sweep at the output from -3.3V to 6.6V or -0.4V to 0.8V? Thanks a lot for the help, 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 | 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 |------------------------------------------------------------------ |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 - ------=_NextPart_000_02C3_01C36281.10C3DA80 Content-Type: application/ms-tnef; name="winmail.dat" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="winmail.dat" eJ8+IhAXAQaQCAAEAAAAAAABAAEAAQeQBgAIAAAA5AQAAAAAAADoAAEIgAcAGAAAAElQTS5NaWNy b3NvZnQgTWFpbC5Ob3RlADEIAQ2ABAACAAAAAgACAAEGgAMADgAAANMHCAAOABAAHAAAAAQAIAEB A5AGALQPAAAlAAAACwACAAEAAAALACMAAAAAAAMAJgAAAAAACwApAAAAAAADAC4AAAAAAAMANgAA AAAAHgBwAAEAAAAtAAAAW0lCSVMtVXNlcnNdREMgU3dlZXAgZm9yIHRoZSBwdWxsdXAgSVYgY3Vy dmUAAAAAAgFxAAEAAAAWAAAAAcNiu7w0xBovPFMGQM6e+lfjT8CSBAAAAgEdDAEAAAAXAAAAU01U UDpUT01AVEVSQVNQRUVELkNPTQAACwABDgAAAABAAAYOAAAes7tiwwECAQoOAQAAABgAAAAAAAAA Kz1QzPFl0hGYaQAgeBDdK8KAAAADABQOAQAAAAsAHw4BAAAAAgEJEAEAAACiCwAAngsAAF4aAABM WkZ1+nsWmAMACgByY3BnMTI14jIDQ3RleAVBAQMB9/8KgAKkA+QHEwKAD/MAUARWPwhVB7IRJQ5R AwECAGNo4QrAc2V0MgYABsMRJfYzBEYTtzASLBEzCO8J97Y7GB8OMDURIgxgYwBQMwsJAWQzNhZQ C6YgTyBuZSB0aAuAZyA6eQhgIB0ACYAdIG8ghwWgAIEEgS4gIFQdQGcEIAqxBUAodx1AE9AgVEkg E+B2HRBuHjBrKSCAd2wJgGcdEG9mlCkgAMB5IUBubCGwNwPhHWEDUiAWUB4hMC7wNCBidQVABpAj kCOBux8xC2BjHfED8B0wIABwfm8dMBKBH1MCIB0hHRBzXmEHgB3BBUAdMGEFQGOfA5EiJQhwJSMd knNoCGDebB4ABaAgUAXAYSSRBIHyIBggZ2klwQuAHYIFwIUEYmwe0UJUVywkkPcm0hewKeBjInAm QAMQIbBzHyEdMXM/CqIKhAqAVIEioURhZ29zdAuAxm8thQSQYXNwHeIIUC8AgCigLnEdcEcDYHVw 8CBMTEMe4DDxLx8wKyEtJDI5MjYGAEUgllkmQB1AbAMgU3Qe0bE0vERldg3gHRBNKsJ5HVJEaTXQ AJACIC0kUBMXwQtgbmQrYE9SIFo5AcAxI0A0vjEcIDHhFlBTVyBICsAdAAQRekwAcGUtJDi/PH8r IGWPIEEAIAIgOBUwMDgtJIJoAkBwOi8vdz+gdi4OsDFlLgWgIrA0uDWAMDMtNDMwLTnwnDY1CuMK gB4gbUA/7GstKi0kLUTSTwUQKeBubwdABdA6kS4wZUTTLSRGFQNhOiFAdx0Aci1psmIEAC11FBAU AEAJgOxhLgWwHXBbAMADEB4g9jpHP0hFXRzwLSQ9oBPgamwjsE8jsE0IcABwef5pK2AHEAqwCzEt QgZgAjBXRxAfAAhwc0hQeUxhdX5nR+AFQDiQK2AB0EFwIIA0OjA3IFBNLYaLRxBJr2dM1XViagWQ CU1hUkVHEFtJQklIUy1VR/JdRDDgU753CeAwoAIQBcAl8nAooPMKQDCgSVYeQAhwIFBDv+wKQTRQ CfAsLSo0gAUQ51IQIfIxkGFrHVIl8zGQ+ywQJjB5LLIdEAIQNFAg0PUdUTotKnwe8SYRAZAOsP0e AHYG8AGQISFMESEgBCD5U+JJLVTgAZEg8AQgKOT/JfIEYE5xQIEEYD4hAJAdYPcg8FrWMhBwC1Ah sCcQFBDacx7RVyUwA6BtL/ILUP8mEWBCCJAEIArAWNQGkAiQ/zgBJfJcBQQgKHAHQBgAWuX/DsEJ 8B4AAJAsUQrAIfEeIf52B0AKUCyyJuET4DfwYdH+cDFgUhAN4EWBDsEYIAeC+yoxGCBmIPBSECRy IEFft38HcCigJuAp8WDgWtZCJS6rbABa1jIhcFZcFVJck99aZWtxHuA3kAuAdFzUPbD/H+Fdc2Fh TnExgAORY4s2YPsuYB3xYirgINBtxm3ZMRK/b5NzzzCwINBs1zS6SEVA/x/wbOVynETTeBQ0u3gf eS9TRlZy9ltQVHFkSUFdMTS8R05EeYBPcE9X7kU4QDS6fpQrfoR731SCv31Nfh9/L4A/fJGC4kML YP5tgc2C74cphM98dIgDhm+3h/+D/4mfWwZiB5FDCHB/GCACMH1Hh5+Ir45vj3RN4E9TRkVUfUiQ 0pWf+5Ivcl1BBCABAATyccAeAP8qMSXyfLcH8AEQBJAJ8CRg4X1Aa2V5dwWwZXFSAf8p8WNkXTht 2SFQmaiBsjfh15mlisouYHIa0HQIcGJleCdWYywQGCBqkiBQJ49WxW4zR+BYdmVxdWqj90cQbNBv hD1s0KFRejBs4E8jcFRgNJCXnyBGbwJ4/4aBIPArYClAKrMklUFQbND/H0Ag0BKBYDVcBih1ccBt 2dcT0mexBnF6cZN0U5EDoLYoFlB6MDUhcKSQLahyPZ7iKEIQkpCs8znxVjvPntOnim3ZKUAzLk8Q qJ3/qgerP6xJsKGtA7Cjr9qtouewopKQtCU2LjOwVOBT5v98uV1zpY9hBQGRaoMdYaUkj5jQJuAp QFPiRUNMHSD2eTGQKqRzY2uooW6jb+J/bdlwN1yDndKhQh4hpLQy2i4UQFYe1r+UYWIlHiF/BuAk wSUhbdmeW7euYOJOHyUQHRIm4SzSwb1PTkz+WStxCfCqn7LTHVIDkbvS/6UvbgZbURQQW/5v47JB cDL/HQBxoWYRnGJSsbszHtEuIH+7YQSBodEeAANSbdkLYGL/IYA9sDIQaGK90yGhJQGyMv9vkx4h cDbMQVyFMXBgIR/h/Z7icx4wIZFt2R3WskFoMv/CIAbwW7MeI14GJ7A0UVyD/x7Wu0Nv49Kk0KoB cVPXalX/U/IeMGeQKPAekWsFTNUeMPsA0AWhZFhzHjBYtx2Sm3B/KKId1n0AtIQeIbbDHtFI/6jB KQFWxcazaZDjcQUQAkD/YUGZhU3h43FhM16SI2ABIO9IASSQmsEiEnQT0B1SMWD/AxEeIC0k5yKn YWVhXoPXUX8AgCxQBBAp8jZhB5HmM3X/AjAEkCxQRXBbwiAxydPYxf9pBCnxawVbU1kCaZDqcR4D /wDAm0BgIWKhxmWccVTl43H/KJljieeWv7LikSoRnfRpkt1T4W2YwQhgb6JkClAeEn/sGGr2LSQ6 0A6wIfBjZHD/V8HushPgXcFnISEgOAGe4v9TkCAkBGBioVoBqXgiMwCQ/mdFcZ0lnuIKsWRxp9Hq if8jYGDEmLEpQB5So3Ga4uWn/+Ex0qIdACRgRdEFECHwOxT/5oTnFucT6DT0mWKDYXAf4ffpIPuS HsJF8mX4YVnUCYD/4ub5EivyneHrzbyxKmixyf8yEA8wDeCPkA+gI5Uo4wTG9i0jIiLzOMEo4FTS 4SFQ/wiyZHHShN6/WPIrYCNh+sX/XNGykNSiIBBoYhSA4rFUFM8T8hKQfQDHsG4nI+IyEP/CAfKA FZAwcCqRRcRfMR4w5x2R/9Wx509L3bsgECZA/5jRS4AdUiSjZZVfQaCQaqTvKiIlFGCjHhFv6DQV kvrFf6CBDKMeVhYj4+HrZFKwUv+Q8CXD2cEyEFHj1/Nl0WMQ/Hkv95QE6Fjy3btMg0vm/xU16nFo EDHgTJC7oGqjOxT+PSKvI78kzyXfJkREH0Uv/0Y+VnOPwPdQ2BBIxm9BKvDoQHN5LpBwLABg4ECB Dl1M303vTvIxMjo1/jFPf1CPUZpSn1OvVL81Of92cM7EpvC8MD3wvLAVNviT/6eA/cGgQOlyGpH/ QDpwWHe/M2LBQDensKGonbuSbdXF/weg3XD/QMFA4mbcIwWE2bD/dpG9JtnBCnGtdAWE+TI/En+3 BDwWGqC/0zwQusVg4UT/2BAP0OEx3DIzJ+8VutXQYv+0g11R4cO05buhCmJFgwrh3FY/NSrBcOkA a/0y+TDz29jT8GxwVssqeFp7exr/TM9N307ve6imskmSDDDc8/scwZkEL+pgUZcHEGhwALK97lBq m4Fe4CvgMQt8FmP+as1y3DLCId1QZ5AMkMYx517S6FMSxChztFCZhcLQ/GR5ch9Jknc3UZeFwTBS 3ZXlPCwwm/JoAi1TEyAwfmSgsbyx8qHfAP5SLUE+/1m/WsYwlFuvXL9dzVI5Wv//YH9hj2KfY6Zf v2U/Zk9yd/9vAmnE48cJ0Dgz1/IwU22F/zD2yu3O0+wVu6GykObB7+K/pvDpQNLAUjAa4VhvaOQA oHA6Ly93dBAuMQVSLzQwYi8wUi9TA1/9cVUvwVCakP9ACQBzH3Qv/3UxMKN1v3bPd991NnnhleYO RWm0ouH+ATE5OTP/S09/34Dvgf9Pj1CfUa9Sv/9Tz1TfVe9W/1gPWR9en2Qf/2mfaq+Nj2hfj6+Q v2cfY7//k++U/5YPZ/+YL5k/ay9sP/9tT25fb29wf3GPi/97D3S//3l/pr+nz3j/qe+q/3wvfT8F fk19stAAAB4AQhABAAAAQgAAADw4MEMzODg0NzU4OENFRjRFQjk3MjEwRDUzQjc1MkM1RDAyMDMx QTAyQGZtc21zeDQwMi5mbS5pbnRlbC5jb20+AAAAAwAJWQEAAAALAACACCAGAAAAAADAAAAAAAAA RgAAAAADhQAAAAAAAAMAAoAIIAYAAAAAAMAAAAAAAABGAAAAABCFAAAAAAAAAwAFgAggBgAAAAAA wAAAAAAAAEYAAAAAUoUAAHN5AQAeACWACCAGAAAAAADAAAAAAAAARgAAAABUhQAAAQAAAAQAAAA5 LjAAAwAmgAggBgAAAAAAwAAAAAAAAEYAAAAAAYUAAAAAAAALAC+ACCAGAAAAAADAAAAAAAAARgAA AAAOhQAAAAAAAAMAMIAIIAYAAAAAAMAAAAAAAABGAAAAABGFAAAAAAAAAwAygAggBgAAAAAAwAAA AAAAAEYAAAAAGIUAAAAAAAALAAWBCCAGAAAAAADAAAAAAAAARgAAAAAGhQAAAAAAAAIB+A8BAAAA EAAAACs9UMzxZdIRmGkAIHgQ3SsCAfoPAQAAABAAAAArPVDM8WXSEZhpACB4EN0rAgH7DwEAAABZ AAAAAAAAADihuxAF5RAaobsIACsqVsIAAG1zcHN0LmRsbAAAAAAATklUQfm/uAEAqgA32W4AAABD OlxNeSBEb2N1bWVudHNcTWFpbGJveFxvdXRsb29rLnBzdAAAAAADAP4PBQAAAAMADTT9NwAAAgF/ AAEAAAAxAAAAMDAwMDAwMDAyQjNENTBDQ0YxNjVEMjExOTg2OTAwMjA3ODEwREQyQkE0RDQ5MTAx AAAAAAMABhCppt4hAwAHENQQAAADABAQAAAAAAMAERAAAAAAHgAIEAEAAABlAAAAT05FVEhJTkdZ T1VORUVEVE9DT05TSURFUlRISVNQQVJUKFdISUNISUhBVkVOT0tOT1dMRURHRU9GKU1BWU9OTFlT V0lOR0ZST00wVE8wNEJVVElGSVRJU1BMQUNFRFdJVEhBTgAAAADVhA== - ------=_NextPart_000_02C3_01C36281.10C3DA80-- |------------------------------------------------------------------ |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 ------------------------------ Date: Thu, 14 Aug 2003 18:06:25 -0700 From: "Mirmak, Michael" Subject: [IBIS-Users] Agenda, IBIS Open Forum Teleconference 8/22 IBIS Open Forum Meeting Agenda for 8/22/03 Bridge Number Reservation # Passcode 1-916-356-2663 4 696-0504 (International Dial-In: 1-617-801-9781) All meetings are 8:00 AM to 9:55 AM Pacific Time. When calling into the meeting, provide the reservation number and passcode at the automated prompts. If asked by an operator, please request to join the IBIS Open Forum hosted by Michael Mirmak. 8:00 Check-In, Intros, Announcements Mirmak - Intros of New IBIS Participants, Meeting Quorum Mirmak - Call for Patents (IBIS 4.0, 4.1, ICM) Mirmak - Membership Update and Treasurers Report Mirmak - Review of Previous Meeting's Minutes (and ARs) Mirmak August 8 Open Forum Minutes - Miscellany/Announcements All - Press & Web Page Updates Huq, All - New Models Available, Library Update Leventhal - Opens for New Issues All 8:15 Administrative and Project Discussions International/External Progress - IEC 62014-3 (ICEM) Integrated Circuits Electromagnetic Model Proposal (IEC 93/67/NP IBIS and EMC Simulation) Perrin/Peters - JEDEC Committee Report Green PCB East Summit Planning Mirmak DesignCon 2004 Summit Planning Mirmak IBIS Version 4.0 Parser Status Mirmak/Peters IBIS Quality Committee Report Katz/Green IBIS Model Review Committee Green S2IBIS3 Status Varma/Ross Web Page Redesign Discussion All, Huq/Coyle New Administrative Issues All 8:45 Technical Discussion BIRD81.2 - Clarification of Clamp Table Use Green/Haller - Call for vote Clarification BIRD Discussion Mirmak - [Series MOSFET] and Vds - [Series Switch Groups] and function_table_group IBIS Version 4.1 Discussion All - [Ramp]/C_comp BIRD status ICM Specification Second Reading and Discussion Mirmak, All - Call for vote on Sept. 12 IBISCHK4 Status Ross/Angulo - Missing tests for BUGs 68, 71, 73, 75, 76 BUG80 - [Model Selector] and [Diff Pin] Ordering Angulo Driver Schedule Clarification Muranyi - Subparameters ignored in scheduled models? New Technical Issues All 9:50 Wrap Up and Next Meetings Plans Mirmak 9:55 Sign Off |------------------------------------------------------------------ |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 ------------------------------ Date: Fri, 15 Aug 2003 13:32:10 -0400 From: "Ingraham, Andrew" Subject: Re: [IBIS-Users] pullup/pulldown range LVDS This is a multi-part message in MIME format. - ------=_NextPart_000_0076_01C36331.A1096280 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MessageDr. Lynne Green gave this as an example: > Example with 12mA end points in I-V data: > -3.3V -12m | dummy end point > 0.4V -12m | end point of actual operating region > | normal data, up to 96 more lines > 1.6V 12m | end point of actual operating region > 6.6V 12m | dummy end point Personally, and maybe this is just me, I hate it when the endpoints do = not continue the slope of the measured (or simulated) data. In the case above the regions <0.4V and >1.6V have zero slope. Is this = realistic? Probably not. Only part of the reason for the dummy end = points is to satisfy the IBIS Parser. But it is also there to tell the = simulator SOMEthing about how your device really behaves, to get the = simulator back on track when it wanders out into that region. Personally I would prefer that you extrapolate the measured or simulated = data points when picking the dummy end points. If the = measured/simulated data has a positive slope, pick dummy end points that = are on that slope, or close to it. Accuracy is less important when you are that far out of the normal = operating region, as is giving the simulator the right ballpark idea. = If not, you might find that the circuit actually converges at 6V and = then you'd be in trouble. Regards, Andy - ------=_NextPart_000_0076_01C36331.A1096280 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Message
Dr. Lynne Green gave this as an = example:
 
>=20 Example with 12mA end points in I-V data:
>=20 - -3.3V   -12m   | dummy end point
>=20 0.4V    -12m   | end point of actual operating=20 region
>=20             &= nbsp;       =20 | normal data, up to 96 more lines
>=20 1.6V     12m   | end point of actual = operating=20 region
>=20 6.6V     12m   | dummy end = point
 
Personally, and maybe this is just me, I hate = it when the=20 endpoints do not continue the slope of the measured (or simulated)=20 data.
 
In the case above the regions <0.4V and >1.6V = have zero=20 slope.  Is this realistic?  Probably not.  Only part of = the=20 reason for the dummy end points is to satisfy the IBIS Parser.  But = it is=20 also there to tell the simulator SOMEthing about how your device really = behaves,=20 to get the simulator back on track when it wanders out into that=20 region.
 
Personally I would prefer that you extrapolate the = measured or=20 simulated data points when picking the dummy end points.  If the=20 measured/simulated data has a positive slope, pick dummy end points = that=20 are on that slope, or close to it.
 
Accuracy is less important when you are that far out = of the=20 normal operating region, as is giving the simulator the right ballpark=20 idea.  If not, you might find that the circuit actually converges = at 6V and=20 then you'd be in trouble.
 
Regards,
Andy
 
 
  - ------=_NextPart_000_0076_01C36331.A1096280-- |------------------------------------------------------------------ |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 ------------------------------ Date: Fri, 15 Aug 2003 15:35:59 -0400 From: "Mike LaBonte" Subject: RE: [IBIS-Users] pullup/pulldown range LVDS This is a multi-part message in MIME format. - ------=_NextPart_000_0014_01C36342.ECF1B240 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit MessageLet's clarify which curves we are flattening. Each characteristic (high state V/I and low state V/I) is split up into 3 separate curves that must be combined mathematically by the simulator, and the curves can have overlapping voltage ranges. At each voltage point the simulator has to add the current of all 3 curves. But at any given voltage, only 1 or 2 of the curves has data. The simulator has to extrapolate the other curves to see what their current would be at the given voltage, assuming the slope at the end continues on. Alternatively it could assume zero current at all points beyond the ends of the curves. But the right approach depends on what the model generator had in mind. For example, s2ibis2 generates curves that *usually* fall to zero current at the ends, and it "believes" that the current remains at zero beyond the ends. If it produces curves that do not fall to zero at the ends, the IBIS model does not work right when an extrapolating simulator adds the curves together. The end flattening method that Lynne presents is derived from the way of making it so that the same result is achieved in either type of simulator, if the current falls to zero at the end. Given that, the 2 clamp curves should be flattened at the inner ends to zero current and slope, and the pullup and pulldown curves should have some non-zero slope as Andy suggests. Otherwise, current would be constant in the region between the driver voltage swing and VCC, because the clamps contribute nothing there. That is not realistic, and simulators might not do the right thing. Continuing the last known slope is preferable, but if it results in a final current value at -VCC and 2VCC that the parser doesn't like, it seems like reducing the slope wouldn't hurt. After all, the simulation is not expected to go there, and in fact the device is probably not designed to go there. If a V/I point is altered (in a forest :-) and no one simulates it, was it ever altered? Mike LaBonte -----Original Message----- From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On Behalf Of Ingraham, Andrew Sent: Friday, August 15, 2003 1:32 PM To: ibis-users@eda.org Subject: Re: [IBIS-Users] pullup/pulldown range LVDS Dr. Lynne Green gave this as an example: > Example with 12mA end points in I-V data: > -3.3V -12m | dummy end point > 0.4V -12m | end point of actual operating region > | normal data, up to 96 more lines > 1.6V 12m | end point of actual operating region > 6.6V 12m | dummy end point Personally, and maybe this is just me, I hate it when the endpoints do not continue the slope of the measured (or simulated) data. In the case above the regions <0.4V and >1.6V have zero slope. Is this realistic? Probably not. Only part of the reason for the dummy end points is to satisfy the IBIS Parser. But it is also there to tell the simulator SOMEthing about how your device really behaves, to get the simulator back on track when it wanders out into that region. Personally I would prefer that you extrapolate the measured or simulated data points when picking the dummy end points. If the measured/simulated data has a positive slope, pick dummy end points that are on that slope, or close to it. Accuracy is less important when you are that far out of the normal operating region, as is giving the simulator the right ballpark idea. If not, you might find that the circuit actually converges at 6V and then you'd be in trouble. Regards, Andy - ------=_NextPart_000_0014_01C36342.ECF1B240 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Message
Let's=20 clarify which curves we are flattening. Each characteristic (high state = V/I and=20 low state V/I) is split up into 3 separate curves that must be combined=20 mathematically by the simulator, and the curves can have overlapping = voltage=20 ranges. At each voltage point the simulator has to add the current of = all 3=20 curves. But at any given voltage, only 1 or 2 of the curves has data. = The=20 simulator has to extrapolate the other curves to see what their current = would be=20 at the given voltage, assuming the slope at the end continues on. = Alternatively=20 it could assume zero current at all points beyond the ends of the = curves.=20 But the right approach depends on what the = model=20 generator had in mind. For example, s2ibis2 generates curves that = *usually* fall=20 to zero current at the ends, and it "believes" that the current remains = at zero=20 beyond the ends. If it produces curves that do not fall to zero at the = ends, the=20 IBIS model does not work right when an extrapolating simulator adds the = curves=20 together. The end flattening method that Lynne presents is derived = from=20 the way of making it so that the same result is achieved = in either=20 type of simulator, if the current falls to zero at the=20 end.
 
Given=20 that, the 2 clamp curves should be flattened at the inner ends to zero = current=20 and slope, and the pullup and pulldown curves should have some non-zero = slope as=20 Andy suggests. Otherwise, current would be constant in the region = between the=20 driver voltage swing and VCC, because the clamps contribute nothing = there. That=20 is not realistic, and simulators might not do the right thing.=20 Continuing the last known slope is = preferable, but if=20 it results in a final current value at -VCC and 2VCC that the = parser=20 doesn't like, it seems like reducing the slope wouldn't hurt. After all, = the=20 simulation is not expected to go there, and in fact the device is = probably not=20 designed to go there. If a V/I point is altered (in a forest :-) and no = one=20 simulates it, was it ever altered?
 
Mike=20 LaBonte
-----Original Message-----
From: = owner-ibis-users@eda.org=20 [mailto:owner-ibis-users@eda.org]On Behalf Of Ingraham,=20 Andrew
Sent: Friday, August 15, 2003 1:32 PM
To:=20 ibis-users@eda.org
Subject: Re: [IBIS-Users] pullup/pulldown = range=20 LVDS

Dr. Lynne Green gave this as an = example:
 
>=20 Example with 12mA end points in I-V data:
>=20 -3.3V   -12m   | dummy end = point
>=20 0.4V    -12m   | end point of actual = operating=20 region
>=20 =             &= nbsp;       =20 | normal data, up to 96 more lines
>=20 1.6V     12m   | end point of actual = operating=20 region
>=20 6.6V     12m   | dummy end = point
 
Personally, and maybe this is just me, I hate = it when=20 the endpoints do not continue the slope of the measured (or simulated) = data.
 
In the case above the regions <0.4V and = >1.6V have=20 zero slope.  Is this realistic?  Probably not.  Only = part of=20 the reason for the dummy end points is to satisfy the IBIS = Parser.  But=20 it is also there to tell the simulator SOMEthing about how your device = really=20 behaves, to get the simulator back on track when it wanders out = into that=20 region.
 
Personally I would prefer that you extrapolate the = measured=20 or simulated data points when picking the dummy end points.  If = the=20 measured/simulated data has a positive slope, pick dummy end = points that=20 are on that slope, or close to it.
 
Accuracy is less important when you are that far = out of the=20 normal operating region, as is giving the simulator the right ballpark = idea.  If not, you might find that the circuit actually converges = at 6V=20 and then you'd be in trouble.
 
Regards,
Andy
 
 
 
- ------=_NextPart_000_0014_01C36342.ECF1B240-- |------------------------------------------------------------------ |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 ------------------------------ End of ibis-users V1 #28 ************************