******************************************************************************
******************************************************************************
BIRD ID#: 97.2
ISSUE TITLE: Gate Modulation Effect
REQUESTER: Arpad Muranyi, Intel Corp.
DATE SUBMITTED: March 4, 2005
DATE REVISED: April 22, 2005, May 25, 2005
DATE ACCEPTED BY IBIS OPEN FORUM: Rejected November 30, 2007
******************************************************************************
******************************************************************************
STATEMENT OF THE ISSUE:
The IV and VT tables of IBIS models are extracted at fixed, ideal supply
voltage conditions. However, due to SSO noise, the actual drive strength
(and consequently the edge rate) may vary during transients depending on the
instantaneous value of the supply voltage. This phenomena is usually called
the "gate modulation effect", which is not described by IBIS models. To
achieve more accurate switching waveforms, and power and ground supply
currents it would be desirable to include these effects in IBIS models.
******************************************************************************
STATEMENT OF THE RESOLVED SPECIFICATIONS:
Add two subparameters, named "Power_coeff" and "n" for the IV tables to
describe how much the instantaneous power supply variations should scale the
IV tables.
These subparameters relate the IV tables in [Model]s to the effective IV
tables used during the simulations when power droop and/or ground bounce
effects are also simulated. The effective IV table can be derived from the
IV tables in the IBIS models using a scaling coefficient k as shown in the
following equation:
IV_effective(Vinst) = k * IV_model(Vnom)
where the scaling coefficient k is calculated using the following formula:
k = [ 1 + Power_coeff * (VinstVnom)/Vnom ]**n
where k is a scaling coefficient (used internally by the simulator's
algorithms) that scales the IV table at each iteration of the simulation,
Power_coeff is the subparameter defined with the IV tables providing a ratio
between the percent change in the supply voltage across the buffer model and
its IV curve, Vinst is the actual, instantaneous power supply voltage
between the pullup and pulldown reference nodes or between the power clamp
and pulldown reference nodes, or between the pullup reference and ground
clamp reference nodes of the buffer model at each iteration of the
simulation, Vnom is the corresponding nominal value of the power supply
voltage at which the IV tables were generated for the [Model] ([Voltage
Range], or [Pullup Reference][Pulldown Reference] for the pullup and
pulldown IV tables, [POWER Clamp Reference][Pulldown Reference] for the
power clamp IV tables, and [Pullup Reference][GND Clamp Reference] for the
ground clamp tables), and n is a parameter that reflects the mobility
degradation at high electric fields (related to the electrical properties of
the gate length). Typically the value of n will be between 1 and 2 which
correspond to short channel (1) and long channel (2) effects respectively.
With n=1, if Power_coeff is zero, the IV tables will remain unchanged,
regardless of what the supply voltage variations are. If Power_coeff = 1,
the IV curve will be scaled by the same percentage as the change in the
actual power supply voltage. For example, if a 5 volt nominal supply drops
to 4.5 volts during the transients of a simulation (which is a 10% change),
an IV curve having a 100 mA saturation current will be scaled down to 90
mA, given a Power_coeff of 1.
Original IV table section:
=============================================================================
 Keywords: [Pulldown], [Pullup], [GND Clamp], [POWER Clamp]
 Required: Yes, if they exist in the model
 Description: The data points under these keywords define the IV tables of
 the pulldown and pullup structures of an output buffer and the
 IV tables of the clamping diodes connected to the GND and the
 POWER pins, respectively. Currents are considered positive
 when their direction is into the component.
 Usage Rules: In each of these sections, the first column contains the
 voltage value, and the three remaining columns hold the
 typical, minimum, and maximum current values. The four
 entries, Voltage, I(typ), I(min), and I(max) must be placed on
 a single line and must be separated by at least one white
 space.

 All four columns are required under these keywords. However,
 data is only required in the typical column. If minimum
 and/or maximum current values are not available, the reserved
 word "NA" must be used. "NA" can be used for currents in the
 typical column, but numeric values MUST be specified for the
 first and last voltage points on any IV table. Each IV
 table must have at least 2, but not more than 100, rows.

 Other Notes: The IV table of the [Pullup] and the [POWER Clamp] structures
 are 'Vcc relative', meaning that the voltage values are
 referenced to the Vcc pin. (Note: Under these keywords, all
...
...
...

[Pulldown]
 Voltage I(typ) I(min) I(max)

5.0V 40.0m 34.0m 45.0m
4.0V 39.0m 33.0m 43.0m
 .
 .
0.0V 0.0m 0.0m 0.0m
 .
 .
5.0V 40.0m 34.0m 45.0m
10.0V 45.0m 40.0m 49.0m

...
...
...
=============================================================================
Modified IV table section:
=============================================================================
 Keywords: [Pulldown], [Pullup], [GND Clamp], [POWER Clamp]
 Required: Yes, if they exist in the model
***SubParam: Power_coeff, n
 Description: The data points under these keywords define the IV tables of
 the pulldown and pullup structures of an output buffer and the
 IV tables of the clamping diodes connected to the GND and the
 POWER pins, respectively. Currents are considered positive
 when their direction is into the component.
**
*** The Power_coeff and n subparameters describe how a given
** instantaneous power supply variation at the nodes of the
** buffer model's supply terminals will effect the IV tables.
*** The relationship between the effective IV tables and the
*** IV tables found in the [Model] is established though a
*** scaling coefficient k.
***
*** IV_effective(Vinst) = k * IV_model(Vnom)
***
** The scaling coefficient is calculated based on the following
** formula.
**
*** k = [ 1 + Power_coeff * (VinstVnom)/Vnom ]**n
**
** where k is a scaling coefficient (used internally by the
** simulator's algorithms) that scales the IV table at each
** iteration of the simulation, Power_coeff is the subparameter
** defined with the IV tables providing a ratio between the
** percent change in the supply voltage across the buffer model
** and its IV curve, Vinst is the actual, instantaneous power
** supply voltage between the pullup and pulldown reference
** nodes or between the power clamp and pulldown reference
** nodes, or between the pullup reference and ground clamp
** reference nodes of the buffer model at each iteration of the
** simulation, and Vnom is the corresponding nominal value of
** the power supply voltage at which the IV tables were
** generated for the [Model] ([Voltage Range], or [Pullup
** Reference][Pulldown Reference] for the pullup and pulldown
** IV tables, [POWER Clamp Reference][Pulldown Reference] for
** the power clamp IV tables, and [Pullup Reference][GND
*** Clamp Reference] for the ground clamp tables), and n is a
*** parameter that reflects the mobility degradation at high
*** electric fields (related to the electrical properties of the
*** gate length).
**
*** With n=1, if Power_coeff = 0, the IV tables will remain
** unchanged regardless of what the supply voltage variations
** are. If Power_coeff = 1, the IV curve will be scaled by
** the same percentage as the percent change in the actual
** power supply voltage. For example, if a 5 volt nominal
** supply drops to 4.5 volts during the transients of a
** simulation (which is a 10% change), an IV curve having a
** 100 mA saturation current will be scaled down to 90 mA,
** given a Power_coeff of 1.
**
 Usage Rules: In each of these sections, the first column contains the
 voltage value, and the three remaining columns hold the
 typical, minimum, and maximum current values. The four
 entries, Voltage, I(typ), I(min), and I(max) must be placed on
 a single line and must be separated by at least one white
 space.

 All four columns are required under these keywords. However,
 data is only required in the typical column. If minimum
 and/or maximum current values are not available, the reserved
 word "NA" must be used. "NA" can be used for currents in the
 typical column, but numeric values MUST be specified for the
 first and last voltage points on any IV table. Each IV
 table must have at least 2, but not more than 100, rows.

**
** Subparameter Usage Rules:
**
*** The Power_coeff and n subparameters are optional. However,
*** if the Power_coeff subparameter is present, it must be
*** placed immediately after the IV curve keyword(s) and its
*** three arguments must follow the subparameter on the same
*** line separated by at least one white space. Similarly,
*** if the n subparameter is present, it must be placed
*** immediately after the Power_coeff subparameter and its three
*** arguments must follow the subparameter on the same line
*** separated by at least one white space. If the n
*** subparameter is present, the Power_coeff subparameter must
*** also be present.
***
*** An "NA" value may be entered in the min and max places only.
*** An "NA" in the min and/or max places will default to the
*** value found in the typical entry. A numerical entry may be a
*** positive, zero, or negative number.
***
*** If the Power_coeff subparameter is not present, the power
*** supply variations will not have an effect on the IV tables.
*** This is also equivalent to having zero entries in it. If
*** the n subparameter is not present, the power supply
*** variations will effect the IV tables linearly only. This
*** is also equivalent to having entries with the value of 1.0
*** in it.
**
** The Power_coeff subparameter's 1st (typical) argument is
** associated with the 1st (typical) column of the IV table,
** the 2nd (minimum) argument is associated with the 2nd
** (minimum) column of the IV table, and the 3rd (maximum)
** argument is associated with the 3rd (maximum) column of the
** IV table.
**
**
 Other Notes: The IV table of the [Pullup] and the [POWER Clamp] structures
 are 'Vcc relative', meaning that the voltage values are
 referenced to the Vcc pin. (Note: Under these keywords, all
...
...
...

[Pulldown]
** typ min max
**Power_coeff 0.5 0.8 0.3  Optional parameter
***n 1.1 1.0 1.2  Optional parameter

 Voltage I(typ) I(min) I(max)

5.0V 40.0m 34.0m 45.0m
4.0V 39.0m 33.0m 43.0m
 .
 .
0.0V 0.0m 0.0m 0.0m
 .
 .
5.0V 40.0m 34.0m 45.0m
10.0V 45.0m 40.0m 49.0m

...
...
...
=============================================================================
******************************************************************************
ANALYSIS PATH/DATA THAT LED TO SPECIFICATION
This proposal is based on IV curve data obtained from actual lab and
simulation measurements. In the context of BIRD95, which attempts to refine
the power and ground bounce simulation capabilities of IBIS simulators, this
feature becomes very important, because the nonideal ground and power
supplies introduce significant deviations in the buffer's IV curves, i.e.
drive strength. Omitting these IV curve variations reduces the accuracy
of the SSO and power delivery simulations when the buffers are surrounded by
non ideal power networks.
*****************************************************************************
ANY OTHER BACKGROUND INFORMATION:
This version of the BIRD incorporates the feedback received in the 4/1/2005
IBIS Open Forum and the 5/5/2005 IBIS Futures/Cookbook teleconferences, the
feedback received from Bob Ross in a private teleconference held on
4/22/2005, and the feedback received from Katja Koller during these weeks
via email.
In BIRD97.1 the equation was corrected to fix an editorial mistake and to
reflect the original intent of the BIRD, in addition to some other editorial
changes in the syntax, and the usage rules the of typ., min., max entries.
For clarity, the orignal version of the BIRD was moved to the end of this
section.
On Katja Koller's request BIRD97.2 includes an additional subparameter n to
describe the nonlinear effects of this relationship.
When the model maker wants to describe a simple linear relationship (i.e.
n=1), he/she can obtain the value for Power_coeff by generating a series
of IV tables at different supply voltages while keeping the remaining
simulation or lab measurement conditions unchanged. The ratio between the
power supply voltage and saturation current variations can be determined
with simple algebraic calculations and entered as the argument for the
Power_coeff subparameter under the IV table keywords.
If the model maker wants to include the nonlinear effects of this
relationship, he/she will need to extract 2 additional points on the IV
table (a total of 3 points) to obtain the Power_coeff and n parameters.
(V1,I1) @ vcc1
(V2,I2) @ vcc2
(V3,I3) @ vcc3
For example: vcc1=3.3V, vcc2=3.0V, vcc3=3.6V
Content of BIRD97:
=============================================================================
 Keywords: [Pulldown], [Pullup], [GND Clamp], [POWER Clamp]
 Required: Yes, if they exist in the model
* SubParams: Power_coeff_typ, Power_coeff_min, Power_coeff_max
 Description: The data points under these keywords define the IV tables of
 the pulldown and pullup structures of an output buffer and the
 IV tables of the clamping diodes connected to the GND and the
 POWER pins, respectively. Currents are considered positive
 when their direction is into the component.
* The Power_coeff_*, subparameters describe how much the
* instantaneous power supply variations should scale the IV
* tables. The scaling coefficient is calculated based on the
* following formula.
*
* k = Power_coeff * (Vinst/Vnom)
*
* where k is the scaling coefficient that scales the IV table
* at each iteration of the simulation, Power_coeff is the
* subparameter provided with the IV tables to relate how much
* the the IV tables should be effected by the power supply
* variations, Vinst is the instantaneous power supply voltage
* across the buffer at each iteration of the simulation, and
* Vnom is the value of the ideal power supply voltage at which
* the IV tables were generated for the [Model].
*
* If Power_coeff is zero, the IV tables will not change,
* regardless of what the supply voltage variations are. If
* Power_coeff = 1, the IV curve will be scaled by the same
* percentage as the change in the actual power supply voltage.
*
*
* Subparameter Usage Rules:
*
* The Power_coeff_* subparameters are optional. If they
* are not present, the power supply variations will not have
* an effect on the IV tables. This is equivalent to a zero
* entry for the subparameter value.
*
* The Power_coeff_typ subparameter is applied to the 1st
* (typical) column of the IV table, the Power_coeff_min
* subparameter is applied to the 2nd (minimum) column, and the
* Power_coeff_max subparameter is applied to the 3rd column
* (maximum) of the IV table.
*
* Numerical arguments are separated from their associated
* subparameter by an equals sign (=); white space around the
* equals sign is optional. The numerical entries may be a
* positive, zero, or negative number. "NA" is not allowed.
*
*
 Other Notes: The IV table of the [Pullup] and the [POWER Clamp] structures
 are 'Vcc relative', meaning that the voltage values are
 referenced to the Vcc pin. (Note: Under these keywords, all
...
...
...

[Pulldown]
*Power_coeff_typ = 0.5
*Power_coeff_min = 0.8
*Power_coeff_max = 0.3

 Voltage I(typ) I(min) I(max)

5.0V 40.0m 34.0m 45.0m
4.0V 39.0m 33.0m 43.0m
 .
 .
0.0V 0.0m 0.0m 0.0m
 .
 .
5.0V 40.0m 34.0m 45.0m
10.0V 45.0m 40.0m 49.0m

...
...
...
=============================================================================
Rejected by the IBIS Open Forum on November 30, 2007.
******************************************************************************