Bird 13.2

All Ibis folk -

This is a second try at publishing BIRD 13 to try to clarify some conditions of
measurement. I have incorporated most comments that I received, I think.

On behalf of the network administration folks here at TI, I apollogize for the
previous attempt not getting out. Our e-mail facility has _again_ entered the
definitely unreliable region. Thus I have moved most of my Ibis work to my
alternate Internet connection at Neosoft. Replies to my neosoft address are
less likely to get lost in the shuffle. Thanks for your patience.

++++++++++++++++++++++++++++++++ SNIP SNIP ++++++++++++++++++++++++++++++++++++

                 Buffer Issue Resolution Document (BIRD)
BIRD ID#: 13.2
ISSUE TITLE: Clarify Some Conditions of Measurements
REQUESTOR: Bob Ward Texas Instruments

DATE SUBMITTED: 22 APR 94, 13 MAY 94
DATE ACCEPTED BY IBIS OPEN FORUM:

*******************************************************************************
*******************************************************************************

STATEMENT OF THE ISSUE:

Certain statements are made in the Version 1.1 standard that need clarification
for the sake of newcomers to the Ibis community regarding the conditions under
which measured data is taken. These changes bring the standard more into line
with the discussions on the forum and the cookbook, which is taken to reflect
the intention and not just the letter of the specification.
  
*******************************************************************************

STATEMENT OF THE RESOLVED SPECIFICATIONS:
       
The majority of the change is to the NOTES TO DATA DERIVATION METHOD section of
the spec.

In paragraph numbered 1)

Old text is:
| V/I Curves for CMOS devices:
| typ = nominal voltage, 50 degrees C, typical process
| min = low voltage tol, 100 degrees C, typical process, minus "X%"
| max = hi voltage tol, 0 degrees C, typical process, plus "X%"
|
| V/I curves for bipolar devices:
| typ = nominal voltage, 50 degrees C, typical process
| min = low voltage tol, 0 degrees C, typical process, minus "X%"
| max = hi voltage tol, 100 degrees C, typical process, plus "X%"

Proposed text is:
| V/I Curves for CMOS devices:
| typ = nominal voltage, nominal temperature deg C, typical process
| min = low voltage tol, max temperature deg C, typical process, minus"X%"
| max = hi voltage tol, min teperature deg C, typical process, plus "X%"
|
| V/I curves for bipolar devices:
| typ = nominal voltage, nominal temperature deg C, typical process
| min = low voltage tol, min temperature deg C, typical process, minus "X%"
| max = hi voltage tol, max teperature deg C, typical process, plus "X%"
|
| where nominal, min, and max temperature are specified by the manufacturer of
| the part. The preferred range is 50C nom, 0C min and 100C max temperatures.

Add after the end of paragraph numbered 2) and before 3):

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 fullrange. This must not be left for the simulator

Add in paragraph numbered 3) after step 3 and before step 4 the following note:

There may be devices which will not drive a load of only 50 ohms into any useful
level of dynamics. In these cases use the manufacturers suggested ( non-
reactive ) load and add the load sub parameter to the [Ramp] specification.

Under the heading Ramp times for CMOS devices, make the same temerature
specifications as above.

Add a note after step 7. in the same section that during the IV measurements
the driving waveform should be of a rise/fall time fast enough to avoid thermal
feedback which is probably not realistic but slow enough that package parasitics
do not distort the data. The specific choice of sweep time is left to the
modelling engineer.

Add specification of the sub parameter 'load' to the [Ramp] keyword.
Under Keyword: [Ramp]
change Sub-Params: dV/dt_r, dV/dt_f, load

Add text to the Usage Rules:
The load sub-parameter is optional if the preferred 50 ohm load is used. It is
required if a non-standard load is used.

Add to the example as follows:
[Ramp]
| variable typ min max
dV/dt_r 4.2/1.8n 3.5/2.5n 5.0/1.1n
dV/dt_f 2.5/1.5n 2.0/2.3n 3.0/0.8n
load 300ohms

Add to the discussion of voltage ranges [ item numbered 2) in the spec ] :

The voltage range is determined by the typical conditions, so that if nominal
Vcc is 5 volts, all voltage extremes run from -5 to +10 volts. There is no
derating for the min and max conditions. So in this example, if min called for
Vcc of 4.5 volts, the range is still -5 to +10, not -4.5 to +9. Similarly for
the max range.

Add after the [Volatge range] keyword discussion:

|==============================================================================|
| Keyword: [Temperature range] |
| Required: Yes, if other than the preferred 0, 50, 100 degree C range |
| Description: Used to define the temperature range over which the model is |
| to operate. |
| Usage Rules: Actual temperatures (not percentages) are to be presented in |
| the usual typ, min, max format. "NA" is not allowed. |
| Other Notes: [Temperature range] also describes the temperature range over |
| which the various V/I curves and ramp rates were derived. Note|
| that these are die temperatures, not ambient temperatures. |
|------------------------------------------------------------------------------|
| variable typ min max
[Temperature range] 27.0C -50C 130.0C

*******************************************************************************

ANALYSIS PATH/DATA THAT LED TO SPECIFICATION:

The spec should have flexibility enough to handle models of parts manufactured
to MIL spec and automotive spec as well as parts for special purposes which are
perhaps more sensitive than even consumer or commercial spec allows. Thus the
relaxation , or tightening as the case may be, of the temperature and voltage
range mandates. The added keyword and sub-parameter are to allow the simulator
useable specification of the relaxed or tightened ranges for the relevant
measurements.

The ramp rate is still mandated to be determined between 20 and 80 % of actual
swing to promote the linearity of the measured portion of the edge. The load
is mandated to be non-reactive so as to preserve the inherent dynamics of the
driver, and not introduce false dynamics due to the load.

Backward compatibility is addressed by making the new specifications optional
if the preferred votage and temperature ranges and load resistance are used.

*******************************************************************************

ANY OTHER BACKGROUND INFORMATION:

  N/A

*******************************************************************************

++++++++++++++++++++++++++++++++++++ SNIP ++++++++++++++++++++++++++++++++++++++

=============================================================================

     __ /
    / \ / / Bob Ward
   /__ / / / / / /
  / \ _ /_ / / / _ __ _ / INET: bward@dadhb1.ti.com -or-
 (____ / (_)_ /__) (__(__/ (_(_/ (_(_/ bward@neosoft.com
                                                      713+274-4146 Voice
                                                      713+274-3911 Fax
=============================================================================
                                   ___
                                  (o o)
-------------------------------ooO-(_)-Ooo-----------------------------------

        Here I sit in endless joy, 'cause I was here before Kilroy!!!
Received on Fri May 13 12:15:44 1994