RE: [IBIS-Users] Mixed-Signal Analog IO ?

Lars

IBIS is really not meant to model analog signal pins.  It was designed to
model two state logic signal driver and receivers.  None of the SI
simulators I'm aware of will simulate an analog function properly.  Though
some of the simulators will do nice things if you have an IBIS 4.1 model and
are using the -AMS capabilities.  But with IBIS 3.2 and earlier you will not
see analog effects.

With that said it is possible to create models for analog pins that will at
least allow analog nets to be included in crosstalk simulations.  For
example, if you have an input of an op amp.  If no IBIS model is assigned to
it then the net will not show up in a SI simulation.  If you assign a simple
IBIS input model to it then the simulator will recognize this net and you
can perform crosstalk simulations that include this net.

For an analog output you could create a simple output model that has the
same slew rate and strength of the analog function.  This would now allow
the simulator to include that net in a simulation.  Obviously this model
will not simulate like the real part but will give you crosstalk into other
nets and show any transmission line effects that might be interesting.

For a voltage output like the output of a band gap reference make a voltage
divider using a terminator model whose values give you the Thevenin
equivalent for the output impedance of the part you are trying to model.

Hope this helps.

Tom Dagostino
Teraspeed(R) Labs
13610 SW Harness Lane
Beaverton, OR 97008
503-430-1065
tom@teraspeed.com
www.teraspeed.com

Teraspeed Consulting Group LLC
121 North River Drive
Narragansett, RI 02882
401-284-1827

-----Original Message-----
From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On
Behalf Of Lars Snith
Sent: Tuesday, August 16, 2005 11:51 PM
To: Muranyi, Arpad
Cc: ibis-users@eda.org
Subject: Re: [IBIS-Users] Mixed-Signal Analog IO ?


Hello Arpad,

Sorry for my vague description. I will try to do better below.
It is clear that every buffer type have clamp tables, and that I can
model analog IO using *-AMS langauges and External Circuit declarations.
But since my tools fall back on s2ibis2 which conforms to IBIS v2.1
specification which (correct me if I'm wrong), does not include these
features, it would be an additional manual procedure to add it.
Hence my question how you did model analog IO (like for example the
output of a bandgap reference with sink & source capability) in "the
good old days" with only I,O,I/O,VDD,GND and NA to choose from.
If this is not possible/advisable I will of course have to take measures.

As I hope is clear from the above, I have no intention to add clamps tp
power pins (although ESD protection calls for snapbacks), that would not
be a good idea having multiple supplies.


Best regards,  Lars Snith


Muranyi, Arpad wrote:

> Lars,
>
> I think you are asking the question in the right
> forum.  However, I don't understand your question.
> To include a clamp, you don't have to make it an
> input, every buffer type can have clamp tables.
> Especially if you do it using the *-AMS languages,
> you can do it any whichever way you want.  So I
> am not sure what you are asking really, or how
> to answer the question.  Please explain it better.
>
> On the other hand, I wouldn't put clamp tables on
> power pins, it doesn't make any sense...  Clamps
> are supposed to clamp overshoot on signal pins.
> What would they do on power pins?
>
> Arpad
> ==================================================
>
>
> -----Original Message-----
> From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org] On Behalf
Of Lars Snith
> Sent: Monday, August 15, 2005 11:22 PM
> To: ibis-users@eda.org
> Subject: [IBIS-Users] Mixed-Signal Analog IO ?
>
> Hello,
>
> I hope I am adressing the right forum with this. If not, please excuse
> my intrusion.
>
> I have been making a number of IBIS models for microcontrollers of
> medium complexity, successfully as it looks.
> As I will soon take on more complex products, with purely analog parts,
> I have been considering the "correct" way of modeling a truly analog IO
> (if at all).
> Clearly, to include the power clamps I should treat it as an input. On
> the other hand this might not be the correct procedure, maybe it would
> be better to treat it as a power/ground pin? It would certainly mean
> less work. A third option is to ignore it altogether (NA).
>
> Any comment/feedback on this issue would be most welcome.
>
>
> Best regards,  Lars Snith

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


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