[SI-LIST] IBIS Model Parasitic

Dear Si-List Members
I wish all of you wonderful holidays and a very happy 2004!

Having worked with IBIS models package and pin parastics, I have noted:

1. Many models include package parasitics  (e.g. R_pkg, C_pkg, L_pkg) for the 
typ, min and 
   max corners without specifying the individual Pin parasitic  (e.g. R_pin, 
C_pin, L_pin).  Certain cases
  necessitate comparing different packages to select the optimum device (please 
see EXAMPLE 1).

2. Some models furnish both the Package and Pin parasitics (illustrated by 
EXAMPLE 2) allowing
 more accurate simulations.

3. It is frequently desirable to change the lumped parasitics into a distributed
format (EXAMPLE 3) towards achieving higher accuracy.

4. Package parasitics may be also available as a .pkg file; which are readily
converted into TOPSPEC (using:  >IBIS2XTK -pkg  ) for XTK modeling .

5. The uncoupled single net package models (  lumped or distributed ) 
   offer simplicity and simulation efficiency (as compared to coupled models), 
but can prove
   unsuitable in certain cases involving fast switching signal.


EXAMPLE 1.
      Let us consider a high-speed PCB design which offers a device choice  
in either TSOP or BGA package; with model parasitics as: 

For TSOP package:

[Package]  
|                         typ             min         max
R_pkg               0.089           0.061       0.094
L_pkg               5.543nH         4.030nH     6.350nH
C_pkg               0.833pF         0.339pF     1.040pF
|

 
For FBGA  package:

[Package]  
|                        typ         min         max
R_pkg               68m         48m         86m
L_pkg               2.29nH      1.34nH      3.16nH
C_pkg               505fF       387fF       571fF
|

In this case, it can prove beneficial to evaluate/simulate with each of 
these package types to decide which one is the optimum selection for the design.


EXAMPLE 2.
      The attachment below shows section of an IBIS model which includes
both Package and individual pin parasitics:


|***************************************************************************
[Component]     LV00A_D
[Manufacturer]  Texas Instruments, Inc.
[Package]
|                 typ        min         max
R_pkg           0.031       0.028       0.034
L_pkg           3.109nH     2.462nH     3.889nH
C_pkg           0.473pF     0.363pF     0.628pF
|
|**************************************************************************
|
[Pin]     signal_name     model_name         R_pin     L_pin      C_pin
|
1         1A              LV00A_IN           0.033    3.854nH    0.622pF
2         1B              LV00A_IN           0.032    3.188nH    0.473pF
3         1Y              LV00A_OUT          0.028    2.462nH    0.377pF
4         2A              LV00A_IN           0.031    2.596nH    0.363pF
5         2B              LV00A_IN           0.028    2.521nH    0.372pF
6         2Y              LV00A_OUT          0.034    3.250nH    0.476pF
7         GND             GND                0.034    3.889nH    0.628pF

Such models which present pin to pin parasitic variations can allow higher 
accuracy simulations as compared to those which provide only R_pkg, L_pkg and 
C_pkg.


EXAMPLE 3
     The lumped format R_pkg, L_pkg, C_Pkg can be converted into a T line model
with characteristic impedance ( Zo) and delay ( TD ) via:

Zo = SQRT( L_pkg/C_pkg)
TD = SQRT(L_pkg * C_pkg)

For instance, when nominal R_pkg=0.192 Ohms, L_pkg = 3.6 nH, C_pkg = 1.59 pF

Then, for SPICE simulations it is possible to create the following T line 
representation:

     Zo=SQRT(L_pkg/C_PKG) = 47.58 Ohms
     TD=SQRT(L_pkg * C_pkg) = 7.57E-11 Sec.

Also, for XTK modeling it is feasible to utilize the pin TOPSPEC feature
(which is preferable over lumped values),
T: PKG_TYP  with PKG_TYP defined by: 

 TOPSPEC PKG_TYP
 NEXT 1.0 Z:47.58 S:0.91     
 SERIES P1 1 TYPE R_PKG
 ENDTOPSEPC
 LOADSPEC R_PKG
 REFF: 0.192

To generate above TOPSPEC required these formulas: 
Z =SQRT(L_pkg/C_PKG)
TD=SQRT(L_pkg * C_pkg), and
S=TD/L  (in unit of nS/ft) , with the package trace length L set to 1.0 inches.
(In this example, translation of lumped parasitics to the equivalent distribued 
representations was shown only for the typ case. Of course, conversion (from 
lumped to 
distributed) for the min and max corner values are also required for the Slow 
and Fast 
corner simulations)


The above cases discussed have involved uncoupled package parasitics. As 
mentioned
earlier, certain fast edge rate applications can dictate use of the more complex
coupled models.

Kind Regards,

Abe Riazi
ServerWorks

 

------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List technical documents are available at:
                http://www.si-list.org

List archives are viewable at:     
                http://www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu

Other related posts: