Forwarding the original email I sent to IBIS-Users
for those who didn't get it...
Thanks,
Arpad
====================================================
-----Original Message-----
From: Muranyi, Arpad
Sent: Wednesday, August 26, 2015 7:30 PM
To: IBIS-users (ibis-users@xxxxxxxxxxxxx)
Subject: Question about [C Comp Corner]
Hello everyone,
I ran across an IBIS file recently which makes use of the [C Comp Corner]
keyword. As I was looking into the details, I noticed that this model had
a C_comp_pullup and C_comp_pulldown subparameter under this keyword, but
only a C_comp subparameter under the [Model] keyword.
This made me wonder, so I opened up the IBIS specification trying to answer
my questions, and also contacted the model maker to find out what their
intent was. The model maker and I seem to have a different interpretation
of the IBIS specification, and I am writing this email to find out what
the real intent was for [C Comp Corner] in the specification.
It doesn't seem to be explicitly stated in the specification, but it seems
to me that the two keywords, [Model] and [C Comp Corner] has their own
independent set of subparameters, even though the names of these subparameter
are identical:
C_comp, C_comp_pullup, C_comp_pulldown, C_comp_power_clamp, C_comp_gnd_clamp
The way I understand the specification is that if someone writes a [Model]
in which [C Comp Corner] is present, the above listed subparameters would
have to appear twice (according to the associated rules between C_comp and
the C_comp_*** variants). The only difference between these two sets of
subparameters will be the order of the values in the typ/min/max positions.
But this does not imply that the two keywords has to have an identical set
of these subparameters.
Now, taking this one step further, if [Model] has C_comp, and [C Comp Corner]
has C_comp_pullup and C_comp_pulldown, and there are two EDA tools, one that
supports [C Comp Corner] and another that doesn't support it, this is what
I would expect to happen:
#1) the tool that supports it will use the C_comp_pullup and C_comp_pulldown
"split capacitors" between the signal terminal of the buffer and its power and
its ground supply terminals.
#2) the tool that does NOT support it will use the C_comp value as a single
capacitor between the signal terminal of the buffer and the much debated
ideal node 0 ground.
I would need to add, that as far as I understand, the tool that does NOT
support [C Comp Corner] will NOT use the C_comp_pullup and C_comp_pulldown
values in this particular example, because these subparameters are placed
after [C Comp Corner] which puts an end to the subparameters under [Model],
i.e. these subparameters do NOT belong to [Model] (even though they are
spelled the same way).
I would like to get some feedback on the above to verify whether my
understanding
of the specification is correct or not. Once we have settled that, I may
suggest
to look at the wording of this section in the specification and possibly propose
that we clarify it so that model makers would not get an incorrect impression on
how they supposed to write their models correctly.
Thanks,
Arpad
====================================================================================
---------------------------------------------------------------------
IBIS Macro website : http://www.eda.org/pub/ibis/macromodel_wip/
IBIS Macro reflector: //www.freelists.org/list/ibis-macro
To unsubscribe send an email:
To: ibis-macro-request@xxxxxxxxxxxxx
Subject: unsubscribe
