ix-1

MINITAB: A Statistical Tool

Appendix I: Minitab: A statistical Tool 158

1.1 Introduction
Here it describes two main software tools which are used for this study. First
is MINITAB a statistical tool can be used to set up and generate the plan of design of
experiments using orthogonal array suitable for the given problems. Another
important tool is Microsim's PSpice, used to perform various types of analysis in
respective experimental studies for finding output responses without using transfer
functions. MINITAB [3.11] facilitates to design both static and dynamic response
experiments. We have used it for the research studies to handle required type of
orthogonal arrays for design of experiments for different circuits and applications.
1.2 DOE using MINITAB
Dr. Genichi Taguchi is regarded as the foremost proponent of robust parameter
design, which is an engineering method for product or process design that focuses
on minimizing variation and/or sensitivity to noise. When used properly, Taguchi
designs provide a powerful and efficient method for designing products that operate
consistently and optimally over a variety of conditions.
In robust parameter design, the primary goal is to find factor settings that
minimize response variation, while adjusting (or keeping) the process on target. After
you determine which factors affect variation, you can try to find settings for
controllable factors that will either reduce the variation, make the product insensitive
to changes in uncontrollable (noise) factors, or both. A process designed with this
goal will produce more consistent output. A product designed with this goal will
deliver more consistent performance regardless of the environment in which it is
used.
Engineering knowledge should guide the selection of factors and responses
[3]. Noise factors for the outer array should also be carefully selected and may
require preliminary experimentation. The noise levels selected should reflect the
range of conditions under which the response variable should remain robust.
Robust parameter design uses Taguchi designs (orthogonal arrays), which
allow you to analyze many factors with few runs. Taguchi designs are balanced, that
is, no factor is weighted more or less in an experiment, thus allowing factors to be
analyzed independently of each other.
MINITAB provides both static and dynamic response experiments.

Appendix I: Minitab: A statistical Tool 159

 1. In a static response experiment, the quality characteristic of interest has a
fixed level.
2. In a dynamic response experiment, the quality characteristic operates over a
range of values and the goal is to improve the relationship between an input
signal and an output response. We can create a dynamic response
experiment by adding a signal factor to a design
The goal of robust experimentation is to find an optimal combination of control factor
settings that achieve robustness against (insensitivity to) noise factors. MINITAB
calculates response tables and generates main effects and interaction plots for:
• signal-to-noise ratios (S/N ratios, which provide a measure of robustness) vs.
the control factors
• means (static design) or slopes (dynamic design) vs. the control factors
• standard deviations vs. the control factors n the natural log of the standard
deviations vs. the control factors
Use these tables and plots to determine what factors and interactions are important
and evaluate how they affect responses. To get a complete understanding of factor
effects it is advisable to evaluate S/N ratios, means (static design), slopes (dynamic
design), and standard deviations. Make sure that you choose an S/N ratio that is
appropriate for the type of data you have and your goal for optimizing the response.
1.2.1 Taguchi design experiments in MINITAB
Performing a Taguchi design experiment may consist of the following steps:
1 Before we begin using MINITAB, we need to complete all pre-experimental
planning. For example, you need to choose control factors for the inner array and
noise factors for the outer array. Control factors are factors you can control to
optimize the process. Noise factors are factors that can influence the performance of
a system but are not under control during the intended use of the product. Note that
while you cannot control noise factors during the process or product use, you need
to be able to control noise factors for experimentation purposes.
2 Use Create Taguchi Design to generate a Taguchi design (orthogonal array) or,
use Define Custom Taguchi Design to create a design from data that you already
have in the worksheet. Define Custom Taguchi Design allows us to specify which
columns are our factors and signal factors. We can then easily analyze the design
and generate plots.

Appendix I: Minitab: A statistical Tool 160

3 After we create the design, you may use Modify Design to rename the factors,
change the factor levels, add a signal factor to a static design, ignore an existing
signal factor (treat the design as static), and add new levels to an existing signal
factor.
4 After we create the design, we4may use Display Design to change the units (coded
or un-coded) in which MINITAB expresses the factors in the worksheet.
5 Perform the experiment and collect the response data. Then, enter the data in the
MINITAB worksheet.
6 Use Analyze Taguchi Design to analyze the experimental data.
7 Use Predict Results to predict S/N ratios and response characteristics for selected
new factor settings.
> Choosing a Taguchi Design
Before one use MINITAB, we need to determine which Taguchi design is most
appropriate for your experiment. In Taguchi designs, responses are measured at
selected combinations of the control factor levels. Each combination of control factor
levels is called a run and each measure an observation. The Taguchi design
provides the specifications for each experimental test run.
A Taguchi design, also known as an orthogonal array, is a fractional factorial
matrix that ensures a balanced comparison of levels of any factor. In a Taguchi
design analysis, each factor can be evaluated independently of all other factors.
When choosing a design we need to
• identify the number of control factors that are of interest
• identify the number of levels for each factor
• determine the number of runs you can perform
• determine the impact of other considerations (such as cost, time, or facility
availability) on your choice of design
> Creating Taguchi Designs
A Taguchi design, or an orthogonal array, is a method of designing experiments that
usually requires only a fraction of the full factorial combinations. In a Taguchi design,
the array is orthogonal, which means the design is balanced so that factor levels are
weighted equally. Because of this, an orthogonal array is one in which each factor
can be evaluated independently of all the other factors. In robust parameter design,
you first choose control factors and their levels and choose an orthogonal array

Appendix I: Minitab: A statistical Tool 161

appropriate for these control factors. The control factors comprise the inner array. At
the same time, you determine a set of noise factors, along with an experimental
design for this set of factors. The noise factors comprise the outer array. The
experiment is carried out by running the complete set of noise factor settings at each
combination of control factor settings (at each run). The response data from each
run of the noise factors in the outer array are usually aligned in a row, next to the
factors settings for that run of the control factors in the inner array.
Each column in the orthogonal array represents a specific factor with two or
more levels. Each row represents a run; the cell values indicate the factor settings
for the run. By default, MINITAB's orthogonal array designs use the integers 1, 2, 3...
to represent factor levels. If you enter factor levels, the integers 1, 2, 3, ..., will be the
coded levels for the design. The following table displays the L8 (27) Taguchi design
(orthogonal array). L8 means 8 runs. 27 means 7 factors with 2 levels each. If the
full factorial design were used, it would have 27 = 128 runs. The L8 (27) array
requires only 8 runs—a fraction of the full factorial design. This array is orthogonal;
factor levels are weighted equally across the entire design. The table columns
represent the control factors, the table rows represent the runs (combination of
factor levels), and each table cell represents the factor level for that run.

I 8 (27) T.K|»chi D f s i i j n

A \i C D F f G
1 I I I I I I I

2 I I \ 2 2 2 2

i I 2 2 I \ 2 2

,\ I 2 2 2 2 I I

•> 2 I 2 I 2 I 2

/, 2 I 2 2 I 2 I

/ 2 2 I I 2 2 I

K 2 2 I 2 I I

In this stated example, levels 1 and 2 occur 4 times in each factor in the array. If you
compare the levels in factor A with the levels in factor B, you will see that B1 and B2

Appendix I: Minitab: A statistical Tool 162

each occur 2 times in conjunction with A1 and 2 times in conjunction with A2. Each
pair of factors is balanced in this manner, allowing factors to be evaluated
independently.

1.2.2 Creating STATIC/ DYNAMIC Taguchi designs: Orthogonal array designs

focus primarily on main effects. Some of the arrays offered in MINITAB's catalog
permit a few selected interactions to be studied. You can also add a signal factor to
the Taguchi design in order to create a dynamic response experiment. A dynamic
response experiment is used to improve the functional relationship between an input
signal and an output response. You can also add a signal factor to the Taguchi
design in order to create a dynamic response experiment. A dynamic response
experiment is used to improve the functional relationship between an input signal
and an output response.
• To create a Taguchi design

1 Choose Stat-> DOE -> Taguchi -^ Create Taguchi Design.

Taguchi Design 1x l

Type of Design
f* i2-Level Design;
<~ 3-Level Design
f* 4-Level Design
c~ 5-Level Design
c M i x e d Level Design
(2 to 31 factor sj
(2 to 13 factor
(2 to 5 factors
;
(2 to 6 factors
(2 to 26 factor B]

D i s p l a y Available Designs...
Number of facte rs:
^3
Designs...

Help OK Cancel

2 If you want to see a summary of the Taguchi designs available, click Display
Available Designs. Click OK.

Appendix I: Minitab: A statistical Tool 163

 Taguchi Design - Available Oatign

Available Taguchi Designs (with Number of Factors)

Single level designs vlixed Z-4 level desi gns

Desicns 2 level 3 level 4 level 5 level | Designs 2 level 4 level 1
14 2-3 L8 1-4 1
18 2-7 116 2-12 1
J_9 2-4 LI 6 1-9 2
L12 2-11 116 1-6 3
LI 6 2-15 116 1-3 4
1.16 2-5 132 1 2-9
1.25 2-6
Mixed 2-8 level designs
127 2-13
L32 2-31

Mixed 2-3 level designs

f LIS
Designs |2 l e v e l |3 l e v e l 1
ti^i n
Mixed 3-6 level designs
Desicns 2 level 3 level|
L18
136
1
1-11
1-7
2-12
[ Designs |3 l e v e l 76 l e v e l 1
L18 ]l-6 li

136 1-3 13
L54 1 3-25

Help OK

3 Under Type of Design, choose a design.

4. From Number of factors, choose a number. The choices available will

vary depending on what design you have chosen.

5. Click Designs.
Taguchi Design - Design

Runs 2 * * Columns

LS 2 ** 2
LI 2 2 ** I
Lib 2 ** 2
L3 2 2 ** C

r Add a sjgnal factor for dynamic characteristics

Help OK Cancel

6 In the Designs box, highlight the design you want to create. If you like, use
the option described under Design sub dialog box below.

7 Click OK even if you do not change any options. This selects the design and
brings you back to the main dialog box.

Appendix I: Minitab: A statistical Tool

8 If you like, click Factors or Options to use any of the options listed below,
then click OK in each dialog box to create your design

Adding a signal factor for a dynamic response experiment

You can add a signal factor to a Taguchi design to create a dynamic response
experiment. A dynamic response experiment is used to analyze and improve the
functional relationship between an input signal and an output response. Generally,
you would use a signal factor when the quality characteristic operates over a range
of values depending on some input to the system [3] [5]. Robustness requires that
there is minimal variation in this relationship due to noise.
The signal factor values are repeated for every run of the Taguchi design
(orthogonal array). Thus, the total number of runs (rows in the worksheet) will be the
number of rows in the orthogonal array times the number of levels in the signal
variable. For example, adding a signal factor with 2 levels to an L4 (23) design,
which has 4 runs, creates a design with 8 total runs; adding a signal factor with 3
levels creates a design with 12 total runs.

:r itic :\*-\ .in •:k;icin

[.'•. I U I I T I K L-.nomic clesicin
• H o :K|llill faciei- •'.KIIkllKKtM i Si .iiiil factor
,vi("h 2 - . - I - - J •-vitVi ', le-.el',i
A E A E Signal >. E Si anal
1 1 fact./I' factor
1 2 1 1 1 1 1
i 1 7 1 :
2 i
2 I i •"* i 1 3
2 1
: i i
: i i-

2 ? i ^ —
1 -'1
i :
1 3
I l

'Alien you odd ci nciru 1 factor '.vhil- creating a i e i v l"a .luchi design, the a m order rt-ill be
different f r o m the order that results from acl> In-. : | o ; icinal factor uiinci Modify Desicin—:ee
Aoji'\t o '\T'\v .v."'•:•.' '•;• J O -:'ii';'\'> ;.\x".v >A" : 0 ' on pa. j e 24-1 ^ The order of the IOVVS
d o e ; n o t affect t h e TV> :iuchi a no k si

• To add a signal factor for a dynamic response experiment

1 In the Create Taguchi Design dialog box, click Design.

Appendix I: Minitab: A statistical Tool 165

 ^ H TmKtmieMn - [>«4m B
E„ a .. ? .. 7c'i'.i*i,J The designs that
display depend on
'•'• : :
. •" the number of
Victors and levels
In your design.

' *ilil • '«^iitil "cell llK ll/Hb>.. ; (l: ^•omlcrMltM

•IrAv | :.K J _. .,

2 Check Add a signal factor for dynamic characteristics. Click OK.

3 In the Create Taguchi Design dialog box, click Factors.
If you like, in the signal factor table under Name, click in the first row and type the
name of the signal factor.
5 Under Level Values, click in the signal factor row and enter numeric values. You
must enter at least two distinct values. Click OK.
• Storing the design
If you want to analyze a design or see whether or not selected interactions can be
estimated from the design, you must store it in the worksheet. By default, MINITAB
stores the design. If you want to see the properties of various designs before
selecting the design you want to store, uncheck Store design in worksheet in the
Options sub dialog box.

•> Estimating selected interactions

Taguchi designs are primarily intended to study main effects of factors.

Occasionally, you may want to study some of the two-way interactions. Some of the
Taguchi designs (orthogonal arrays) allow the study of a limited number of two-way
interactions. This usually requires that you leave some columns out of the array by
not assigning factors to them. Some of the array columns are confounded with
interactions between other array columns. Confounding means that the factor effect
is blended with the interaction effect, thus they cannot be evaluated separately. You
can ask MINITAB to automatically assign factors to array columns in a way that
avoids confounding—see To select interactions on page 24-11. Or, if you know
exactly what design you want and know the columns of the full array that correspond
to the design, Interaction tables show confounded columns, which can help you to
assign factors to array columns. For interaction tables of MINITAB's catalog of

Appendix I: Minitab: A statistical Tool 166

Taguchi designs (orthogonal arrays), see Help. The interaction table for the L8 (27)
array is shown below.

2 ?, 4 *\ 6 7
7
1 3 2 N 4 o

7 *)
2 1 6 4

3 7 6 5 4

4 1 2

5 3 2

6 1

The columns and rows represent the column numbers of the Taguchi design
(orthogonal array). Each table cell contains the interactions confounded for the two
columns of the orthogonal array. For example, the entry in cell ( 1 , 2) is 3. This
means that the interaction between columns 1 and 2 is confounded with column 3.
Thus, if you assigned factors A, B, and C to columns 1, 2, and 3, you could not study
the AB interaction independently of factor C. If you suspect that there is a substantial
interaction between A and B, you should not assign any factors to column 3.
Similarly, the column 1 and 3 interaction is confounded with column 2, and the
column 2 and 3 interaction is confounded with column 1.
• To select interactions
1 In the Create Taguchi Design dialog box, click Factors.

11**1*1 VftlUMK
_ . JS'-S-.K^^U
11 ' n \~
I-' U
Thl; option only
j'/jiitabk'.vhvii you
hove J dynamic ---.. -ir.. I..I

d~:ign. - •
|1*%**J Vol**"*
_ !
laai
n't | r w n «-.. i

2 Under Assign Factors, choose To allow estimation of selected Interactions

and

Appendix I: Minitab: A statistical Tool 167

 __^^_ ^^_, ^ — ,
'£-•«*«.-*«* I f i n i t e :

> J
>> 1
J
,..;.. J

FkUtarsc
ft. A

•<*•* i r UK
—
€MMH

3 Move the interactions that you want to include in the design from Available
Terms to Selected Terms using the arrow buttons
1. to move the interactions one at a time, highlight an interaction, then click or

2. to move all of the interactions, click on or

You can also move an interaction by double-clicking it.
4. Click OK.This option only available when you have a dynamic design.
• To assign factors to columns of the array
Assigning factors to columns of the array does not change how the design is
displayed in the worksheet. For example, if you assigned factor A to column 3 of the
array and factor B to column 2 of the array, factor A would still appear in column 1 in
the worksheet and factor B would still appear in column 2 in the worksheet.
1 In the Create Taguchi Design dialog box, click Factors.

MKtgnftGMr*

l i e . ufti l l m i j f j

-4-+
|p*etr<T | K » B » |Lrt**i TkloMi

1
This option Is only
available when
you have ,i -^ Sit**! f * i«t

dynamic design IBSFHS

i
1
M if SK GOWtl

2 Under Assign Factors, choose To columns of the array as specified below.

3 In the factor table, click under Column in the cell that corresponds to the factor
that you want to assign. From the drop-down list, choose the array column to which
you want to assign the factor. Then, use the Z key to move down the table and
assign the factors to the remaining array columns. Click OK.

Appendix I: Minitab: A statistical Tool 168

 Assigning factors to columns of the array does not change how the design is
displayed in the worksheet. For example, if you assigned factor A to column 3 of the
array and factor B to column 2 of the array, factor A would still appear in column 1 in
the worksheet and factor B would still appear in column 2 in the worksheet. This
option is only available when you have a dynamic design.
• Setting factor levels
By default, Minitab sets the levels of a factor to the integers 1, 2, 3 You may
change these to other numbers, such as the actual values of the factor level, or to
text levels. One useful technique for customizing Taguchi designs (orthogonal
arrays) is the use of "dummy treatments." You can create a dummy treatment in
MINITAB by repeating levels for the same factor, as long as there are at least two
distinct levels.

L° [lA) ."iiT.\y L° i3 4 ) .'inv»y •: d u m m y ;

Run A B C D Run A 6 C D

Dummy

• Creating dummy treatments

One useful technique for customizing Taguchi designs (orthogonal arrays) is the use
of "dummy treatments." You can create a dummy treatment in MINITAB by repeating
levels for a factor, as long as there are at least two distinct levels. For example, if
you wanted to use an L9 (34) array, which has four three-level factors, but had one
factor with only two levels, you could use a dummy treatment to accommodate this.
Here, the L9 (34) array is shown, both without and with a dummy treatment. In the
dummy example, the factor levels for factor A are 1 2 1, where 1 is the repeated

Appendix I: Minitab: A statistical Tool 169

level for the dummy treatment. In the L9 (34) orthogonal array with dummy treatment
above, factor A has repeated level 1, in place of level 3. This results in an L9 (34)
array with one factor at 2 levels and three factors at 3 levels. The array is still
orthogonal, although it is not balanced. When choosing which factor level to use as
the dummy treatment, you may want to consider the amount of information about the
factor level and the availability of experimental resources. For example, if you know
more about level 1 than level 2, you may want to choose level 2 as your dummy
treatment. Similarly, if level 2 is more expensive than level 1, requiring more
resources or time to test, you may want to choose level 1 as your dummy treatment.

• Summary of Available Taguchi Designs

• Single-level designs
The table below summarizes the single-level Taguchi designs available. The number
following the "L" indicates the number of runs in the design. For example, the L4 (23)
design has four runs. The numbers in the table indicate the minimum and maximum
number of available factors for each design. For example, an L8 (27) design can
have from two to seven factors with two levels each; an L16 (45) design can have
from two to five factors with four levels each.
Mixed 2-3 level designs: The table below summarizes the available Taguchi
designs for mixed designs in which factors have 2 or 3 levels. The number in the
table cells indicates the minimum and maximum number of factors available for each
level. For example, an L18 (21 37) design can have one factor with two levels and
from one to seven factors with three levels.
r ; u : t i b - - l .;f K'\ >.•!•

1 4 •..'")

Lr. • . ' ' . .

L •-'•>.-'; --i

I \1 I . : 1 1 : - I I

L'...2'-: --I?

L \<. l4\.

Uf> i V :•

1
L^l.r : -'I

Appendix I: Minitab: A statistical Tool 170

Mixed 2-4 level designs
The table below summarizes the available Taguchi designs for mixed designs in
which factors have 2 or 4 levels. The number in the table cells indicates the
minimum and maximum number of factors available for each level. For example, an
L8 (24 41) design can have from one to four factors with two levels and one factor
with four levels.

N u m b e ' of level-"

Desi :tns 2 >

L I * ,2' V) I I-7

II -;I2,
L : :6 2 l-l I 2-12

L3i> . 2 J 3IJ) I-:' I .:,

L54 >l 323} I r*~ £..'

Mixed 2-8 level designs

The table below show the available Taguchi design for mixed designs in which
factors have 2 and 8 levels. The number in the table cells indicates the minimum and
maximum number of factors available for each level.

Nunil IHH of levels

Designs 2 4

L8i244') 1-4 1

LI6«2,24') 2-\2 1

LI6<2942) l-° "i

LI6-;2°4J) 1-6 ?•

L16 .:23 44> 1-?. 4

L12<2] A9) 1 /.'"i

An L16 (28 81) design can have from one to eight factors with two levels and one
factor with eight levels.

Appendix I: Minitab: A statistical Tool 171

 Number of levels
Design 2 8
LI6{288') I-.'?. 1

Mixed 3-6 level designs

The table below shows the available Taguchi design for mixed designs in which
factors have 3 and 6 levels. The number in the table cells indicate the minimum and
maximum number of factors available for each level. An L18 (36 61) design can
have from one to six factors with three levels and one factor with six levels.

Number of levels
Design I level 6 level
L 18(3*6') 1-6 1

• Defining Custom Taguchi Designs

Use Define Custom Taguchi Design to create a design from data you already have
in the worksheet. For example, you may have a design you:
• created using MINITAB session commands
• entered directly in the Data window
• imported from a data file
• created as another design type in MINITAB
• created with earlier releases of MINITAB
You can also use Define Custom Taguchi Design to redefine a design that you
created with Create Taguchi Design and then modified directly in the worksheet.
Define Custom Taguchi Design allows you to specify which columns contain your
factors and to include a signal factor.

After you define our design.

To define a custom Taguchi design
1 Choose Stat => DOE => Taguchi=> Define Custom Taguchi Design.
2 In Factors, enter the columns that contain the factor levels.

Appendix I: Minitab: A statistical Tool 172

 EE3E=XE3EE:
fWWKB

•• Ma uiijtal ImAif
f Sptclfr fcv oaiimir I

a*.

3.lf you have a signal factor, choose Specify by column and enter the column that
contains the signal factor levels. Click OK.
Modifying Designs
After creating a Taguchi design and storing it in the worksheet, you can use Modify,
Design to make the following modifications:
• rename the factors and change the factor levels for the control factors in the
inner array
• add a signal factor to a static design.
• ignore the signal factor (treat the design as static)
• add new levels to the signal factor in an existing dynamic design—see Adding
new levels to the signal factor.
• By default, MINITAB will replace the current design with the modified design.
To store the modified design in a new worksheet, check Put modified design
in a new worksheet in the Modify Design dialog box.
Renaming factors and changing factor levels
> To rename factors or change factor levels
1. Choose Stat => DOE ==> Modify Design.

Appendix I: Minitab: A statistical Tool 173

 I Choose Stat >• DOE >• Modify Design

Static Design Dynamic Design

W^P^W^PHWB|
EJ
Wiiilrh:«l^*
;
Wmtlity#ii»ir,Hi!."tJ»*in('i»r» • U t i l i t y «•.(-**": •«» "tiM'i :ir»*y
Ail*I M l p « laUut S**lHlV >l'l»ll>ll ' * U » 1

!"" i^S JS.sntillr.fr ?i*t;ijj* * * fl D t w v » r t S * ^ r l

i.-i, I Spc-ir* I
""', 1 «-™^;!!xJ " ! : ™ . .

2 Choose Modify factors in inner array Click Specify

,ir ->—rrr

3 Under Name, click in the first row and type the name of the first factor. Then, use
the LL key to move down the column and enter the remaining factor names.
4 Under Level Values, click in the first row and type the levels of the first factor.
Then, use the LL key to move down the column and enter the remaining levels. Click
OK in each dialog box.
Adding a signal factor to an existing static design
When you add a signal factor to an existing static design, MINITAB adds a new
signal factor column after the factor columns and appends new rows (replicates) to
the end of the existing worksheet. For example, if you add a signal factor with 2
levels to an existing L4 (23) array, 4 rows (1 replicate of 4 runs) are added to the
worksheet; if you add a signal factor with 3 levels, 8 rows (2 replicates of 4 runs) are
added to the worksheet. A replicate is the complete set of runs from the initial
design.

Appendix I: Minitab: A statistical Tool 174

 Static des •:in Dy mmii design Dynamic design
(No signal factor) (Added siq nal factor (Added signal factor
with 2 evels) with 3 levels)
A B •J. B Stonal r>. E Stonal
1 1 factor factor
1 2 1 1 1 1 1 1
2 1 1 2 1 1 2 1
2 2 1 1 2 1 1
2 2 1 2 2 1

1 1 2 1 1 2
1 7
1 2 2
2 1 2 2 1 2
2 2 2 2 2 2

1 1 3
I 2 3
2 1 3
2 2 3

To add a signal factor

1 Choose Stat -> DOE -» Modify Design.
2 Choose Add signal factor. Click Specify.

|bseel V-5l*e*
i
r." -
Hrlp J !!
k
m CSVCPI

3 If you like, in the signal factor table under Name, click in the first row and type the
name of the signal factor.
4 Under Level Values, enter the levels of the signal factor. You must enter at least
two distinct values. Click OK.
Ignoring the signal factor
You can choose to ignore the signal factor in a dynamic design and thus treat the
design as static.
> To ignore the signal factor
1 Choose Stat a DOE a Modify Design.
2 Choose Modify signal factor. Click Specify.
i II i mm^ammmmmim |
"mill .<• '

i i ^ U t S '.LJlfe r^CKf &t** f, |tf»-ly!lW»«J

HH> | [ BC ] CUWW |

3 Select Ignore signal factor (treat as non-dynamic). Click OK in each dialog box.
Adding new levels to the signal factor

Appendix I: Minitab: A statistical Tool 175

When you add signal factor levels to an existing dynamic design, new rows
(replicates) are appended to the end of the existing worksheet. For example, if you
add 3 new signal factor levels to an existing L8 (27) design, 24 rows (3 replicates of
8 rows each) are added to the worksheet.
1 Choose Stat => DOE=> Modify Design.
2 Choose Modify signal factor. Click Specify.

' §p»rc sign* ta&w Hnr#t * * (wMynawirt

(• &iri nmw Irvwl* I * ilgtrtl foidtar?

Displaying Designs
After you create the design, you can use Display Design to change the way the
design points are stored in the worksheet. You can display the factor levels in coded
or uncoded form. If you assigned factor levels in Factors subdialog box, the uncoded
(actual) factor levels are initially displayed in the worksheet. If you did not assign
factor levels (used the default factor levels, which are 1, 2, 3, ...), the coded and
uncoded units are the same.
To change the units for the factors
1 Choose Stat => DOE=> Display Design.

• iw i- - .|tiji.--- • |IH ssr,intewotfct*«Tt

l*illota«Utk»s:

- cofetiium
C &a>M ntoi

M*» j r ___.j tUKCl j

2 Choose Coded units or Uncoded units. Click OK.

Collecting and Entering Data
After you create your design, you need to perform the experiment and collect the
response (measurement) data. To print a data collection form, follow the instructions
below. After you collect the response data, enter the data in any worksheet column
not used for the design.

Appendix I: Minitab: A statistical Tool 176

Printing a data collection form
You can generate a data collection form in two ways. You can simply print the Data
window contents, or you can use a macro. A macro can generate a "nicer" data
collection form—see Help for more information. Although printing the Data window
will not produce the prettiest form, it is the easiest method. Just follow these steps:
1 When you create your experimental design, MINITAB stores the factor
settings in the worksheet. These columns constitute the basis of your data collection
form. If you did not name factors or specify factor levels when you created the
design and you want names or levels to appear on the form.
2 In the worksheet, name the columns in which you will enter the measurement data
obtained when you perform your experiment.
3 Choose File a Print Worksheet. Make sure Print Grid Lines is checked, then
click OK.
3.2.3 Analyzing S/D designs
Analyzing Taguchi Designs
To use Analyze Taguchi Design, you must
• create and store the design using Create Taguchi Design or create a design from
data already in the worksheet using Define Custom Taguchi Design
• enter the response data in the worksheet.
Using Analyze Taguchi Design, you can
• generate main effects and interaction plots of the S/N ratios, means (static
design), slopes (dynamic design), and standard deviations vs. the control factors
• display response tables for S/N ratios, means (static design), slopes (dynamic
design), and standard deviations
The response tables and main effects and interaction plots can help you determine
which factors affect variation and process location.
A.2.4 ANOVA
A.3 Two-step optimization
Two-step optimization, an important part of robust parameter design, involves first
reducing variation and then adjusting the mean on target. Use two-step optimization
when you are using either Nominal is Best signal-to-noise ratio. First, try to identify
which factors have the greatest effect on variation and choose levels of these factors

Appendix I: Minitab: A statistical Tool 177

that minimize variation. Then, once you have reduced variation, the remaining
factors are possible candidates for adjusting the mean on target (scaling factors).
Initial process performance
• h i g h variation
• process n o t o n target

lowsr Target Upper

S t e p I: M i n i m i z e variation
• f i n d f o t t o r settings t h a t m i n i m i z e t h
effects of noise o n t h e response
• variation minimized
• process n o t o n target 1 1
J \.1 —
Lower T.IK|*I Up pel

Step 2 Adjust mean on target

• find factor settings th.it adjust t h e
m e a n o n t.iraet
• variation minimized
• process o n target
• robust design Lov.'«r T*-ii-:|~t Lfppei

A scaling factor is a factor in which the mean and standard deviation are
proportional. You can identify scaling factors by examining the response tables for
each control factor. A scaling factor has a significant effect on the mean with a
relatively small effect on signal-to-noise ratio. This indicates that the mean and
standard deviation scale together. Thus, you can use the scaling factor to adjust the
mean on target but not affect the S/N ratio.
Use main effects plots to help you visualize the relative value of the effects of
different factors. Recall, the inner array represents the control factors. There are two
noise conditions in the outer array (Noise 1 and Noise 2). There are two
responses—one for each noise condition—in the outer array for each run in the
inner array. You can have 1 response column if you are using the Larger is Better or
Smaller is Better signal-to-noise ratio and you are not going to analyze or store the
standard deviation.
If you have a design and response data in your worksheet that was
• created using Minitab session commands,
• entered directly in the Data window,
• imported from a data file,
• created using as another design type in Minitab,

Appendix I: Minitab: A statistical Tool 178

n or created with earlier releases in Minitab,
you can use Analyze Taguchi Design, which will prompt you to define your design.
> To fit a model to the data
1 Choose Stat -^ DOE -^ Taguchi -^ Analyze Taguchi Design.
2 In Response data are in, enter the columns that contain the measurement data.
3 If you like, use any of the options listed below, then click OK.

pi B**l»w* data** fee

I PBHUW.,. I £nrm„, |
Help | | flK | Cmri j

Options
Graphs sub dialog box
n for static designs, display main effects plots and selected interaction plots for the
signal-to-noise (S/N) ratios, the process means, and/or the process standard
deviations for dynamic designs, display main effects plots and selected interaction
plots for the S/N ratios, the slopes, and/or the process standard deviations. Also,
display scatter plots with fitted lines.
n display interaction plots for selected interactions. Display the interaction plots in a
matrix on a single graph or to display each interaction plot on a separate.
Tables sub-dialog box
• for static designs, display response tables for signal-to-noise ratios, the means,
and the standard deviations.
• for dynamic designs, display response tables for signal-to-noise ratios, the slopes,
and the standard deviations
Options sub dialog box
• for static designs, choose the signal-to-noise (S/N) ratio that is consistent with
your goal and data.
• for dynamic designs, enter a response reference value and a signal reference
value for the fitted line or choose to fit the line with no fixed reference point.
• use natural logs in graphs and tables for standard deviations

Appendix I: Minitab: A statistical Tool 179

Storage sub dialog box
• for static designs, store the
- S/N ratios
- means
- standard deviations
- coefficients of variation
- natural log of the standard deviations
• for dynamic designs, store the
- S/N ratios
- slopes
- intercepts
- standard deviations (square root of MSE)
- natural log of the standard deviations
Displaying main effects and interaction plots
You can display main effects and selected interaction plots for signal-to-noise
(S/N) ratios, means (static designs), slopes (dynamic designs), and/or standard
deviations.
> To display main effects and interaction plots
1 In the Analyze Taguchi Design dialog box, click Graphs.

St.'itiu Design Dynamic Design

• M R * M O M M D S M - BU0N
i.r •.-,(,- , : - . ,it - t i i l:< - t ' • ' I ,t . 1. ; - 1 . U- l u \ . 1 i ; --. >•) - - .. . - I " I- • • I •..- •• ' I r - - l i f t - -ir; t--

• ;i.r-»- to | i i - t ' imtn-. ! •' :t$,--- I.. N r - . i^.t :

! - • • ! « . : -

-|.I>V li l - i - n ill:--* n r - . l r . l i i l ra.: ...|.-v , l . r , . - ... | . . — I... . . I . 1..I -r,*.- . .......

.(•I.J.- ' '"•-' ; I-. .- " Mil i • .

M|<

2 Under Generate plots of main effects and selected interactions for check
Signal-to-noise ratios, Means (for static design) or Slopes (for dynamic
design), and/or Standard deviations. Click OK.
Selecting terms for the interaction plots

Appendix I: Minitab: A statistical Tool 180

You can choose which interactions to plot. You can also choose whether to display
the interaction plots in a matrix on a single graph or to display each interaction plot
separately on its own page,
h To select which interactions to plot
1 In the Analyze Taguchi Design dialog box, click Graphs.
2 In the Graphs subdialog box, click Interactions.

U t * *ltt :(«..:* S-lB .d IftffM lit' Kill t l |:•;•*•!3

ktni*f**i rVlst:
# Qfeplfty Mew*** |»IIM M a n t * h H » sS*§fe MI I,»II

I'll IX >

3 Move the interactions that you want to include in the plot from Available Terms to
Selected Terms using the arrow buttons.
• to move the interactions one at a time, highlight an interaction, then click or
D|
• to move all of the interactions, click on —J - - ' J . You can also move an
interaction by double-clicking it. The available terms in the Interactions subdialog
box list the interactions available to plot. The second factor in the term (B in AB) is
used as the horizontal scale for the plot. Thus, you can view the AB interaction
both ways by selecting both AB and BA.
Displaying response tables
You can display response tables for signal-to-noise (S/N) ratios, means (static
designs), slopes (dynamic designs), and/or standard deviations.
h To display response tables
1 In the Analyze Taguchi Design dialog box, click Tables.

St.itic Design Dynamic Design

P^ignlliNiiHi*)
P Mtmii. U biuret

>id|i sm u*ta , Help | OK CMKtl

Appendix I: Minitab: A statistical Tool 181

2 Under Display response tables for check Signal-to-noise ratios, Means (for
static design) or Slopes (for dynamic design), and/or Standard deviations. Click
OK.
Analyzing static designs
If you have a static design (no signal factor), you can choose signal-to-noise (S/N)
ratios depending on the goals of your design. S/N ratios differ, therefore you should
use your engineering knowledge, understanding of the process, and experience to
choose the appropriate S/N ratio [3].
Choose... U s e w h e n t h e g o a l is t o . . . A n d y o u r data are...

Larger is better M a x i m i z e the response Positive

S/N=.10f1og(I(l/Y2)/n»

N o m i n a l is best Target the response a n d y o u Positive, zero, or negative

w a n t t o base the S/N ratio on
S/N-ICKIogfs2))
standard deviations only

N o m i n a l is best (default; Target the response a n d y o u N o n - n e g a t i v e w i t h an

w a n t t o base the S/N ratio o n "absolute z e r o " i n w h i c h t h e
S/N=10(log((72)./s2»
means a n d standard deviations staridaid deviation is zero
w h e n t l v m e a n is zero

Smaller is better Minimize ' l i e response N o n - n e g a t i v e w i t h a target

value of zero
S N=-|rVlocj(IY2.'n))

Not £ The Nominal is Best idef 0 LI It 1S/N ratio is flood for anahzing or ide -itifymciscalinci factors,
which .ire factors in win : h t l e mean and standard deviation vaiy >roportionaHy Scaling
factors con be used to .1 :l|ust the mean on target w i t h o u t aftectin- 1 S/N ratios

To select a signal-to-noise ratio

1 In the Analyze Taguchi Design dialog box, click Options

F-mmrwrmw
SigpLlI H) ItoSsn Rl*« funwihi
f~ l H * » K « r n n -•m.»ffiiMi|i|Y*'?p|
r MiMiMiiibett •tttiHhra
»IrtrnSmi M !>•« io-i«»(rtB»i~«)f»"?i
r SmathrKhrHK -vnttumtfim
r Vhf M i l ib araplw »td hrtte-s liw 5ta»*8«4 <irvt»i<w*f»

dpi, I ' Q* 1 ClMeJ J

2 Under Signal-to-Noise Ratio, choose the S/N ratio that best fits the goals of the
design. Choose from one of the following:
• Larger is better
• Nominal is best
• Nominal is best
• Smaller is better
3 Click OK.
Analyzing dynamic designs

Appendix I: Minitab: A statistical Tool 182

Dynamic response experiments are used to improve the functional relationship
between input signal and output response, in other words, to optimize tunability. The
output response should be directly proportional to the input signal. The ideal
functional relationship between input signal and output response is a line through the
origin. In some cases, you may wish to choose a reference point, other than the
origin, through which the line should pass. For example, your results may be
generated far from zero, by specifying a reference point in the range of results you
can enhance the sensitivity of the analysis. Or, you can choose to fit the line with no
fixed reference point. In this case, the intercept will be fitted to the data.
To specify a reference point for the response
1 In the Analyze Taguchi Design dialog box, click Options.

tff - ••• . ;> • ' • • - *? i s - . . i

& * " * lit l i - - , In , . | l i ., i . I . M . i r in><a*

i:> • <• i !• r - i i u t v * l U £ •{

• < * • f<
r f nftn«BwMh no feci* roliwm ptIM

P V w |nft$ l» graphs m4 Itble* tnummAvi dc»M»ns

HH,i j QK | CwrfrtJ

2 In Response reference value, enter a numeric value corresponding to the desired

output (response) value.
3 In Signal reference value, enter a signal factor level corresponding to the
response reference value. Click OK.
To fit a line with no fixed reference point
1 In the Analyze Taguchi Design dialog box, click Options.
2 Select Fit lines with no fixed reference point. Click OK.
Interpreting the results

The response tables show the average of the selected characteristic for each level
of the factors. The response tables include ranks based on Delta statistics, which
compare the relative magnitude of effects. The Delta statistic is the highest average
for each factor minus the lowest average for each factor. Ranks are assigned based
on Delta values; rank 1 is assigned to the highest Delta value, rank 2 to the second
highest Delta value, and so on. The main effects plot provide a graph of the
averages in the response table.

Appendix I: Minitab: A statistical Tool 183

 Look at the response tables and main effects plots for the signal-to-noise (S/N)
ratios to see which factors have the greatest effect on S/N ratio, which in this
example is nominal-is-best.
Predicting Results
Use Predict Results after you have run a Taguchi experiment and examined the
response tables and main effects plots to determine which factor settings should
achieve a robust product design. Predict Results allows you to predict S/N ratios and
response characteristics for selected factor settings.
For example, you might choose the best settings for the factors that have the
greatest effect on the S/N, and then wish to predict the S/N and mean response for
several combinations of other factors. Predict Results would provide the expected
responses for those settings. You should choose the results that come closest to the
desired mean without significantly reducing the S/N ratio. You should then perform a
follow-up experiment using the selected levels, to determine how well the prediction
matches the observed result. If there are minimal interactions among the factors or if
the interactions have been correctly accounted for by the predictions, the observed
results should be close to the prediction, and you will have succeeded in producing a
robust product. On the other hand, if there is substantial disagreement between the
prediction and the observed results, then there may be unaccounted for interactions
or unforeseen noise effects. This would indicate that further investigation is
necessary.You can specify the terms in the model used to predict results. For
example, you may decide not to include a factor in the prediction because the
response table and main effects plot indicate that the factor does not have much of
an effect on the response. You can also decide whether or not to include selected
interactions in the model. Interactions included in the model will affect the predicted
results.

Data
In order to predict results, you need to have
• created and stored the design using Create Taguchi Design or created a
design from data already in the worksheet with Define Custom Taguchi Design and

• analyzed it using Analyzing Taguchi Designs

Appendix I: Minitab: A statistical Tool 184

To predict results

1 Choose Stat => DOE => Taguchi=> Predict Results.

Static Design Dynamic Design

*V r iM •..1 1
/- V :.l|
P siput hi &*r*r mi.. P % • • ! to HMie i.lk.
(7 Slwtor* ilnvMttn
17 N>IU»ll<l«.«<ull>MM*:vlili<a P Hum il In* * a i i M M tcvlMia

F Sto*op«n4KMv4liip*i»MDifcffcfBt P SfcJ— y w i K M valtw tew»ifcifcfBl

' - • • Ltwlf*. leraw...

I
He* | ' W COKCI Mrl. flK CMCCI
1

2 Choose to predict one or more of the following:

• mean (static design) or slope (dynamic design)
• signal-to-noise ratio
• standard deviations
• natural log of standard deviation
3 Click Terms.
4 Move the factors that you do not want to include in the model from Selected
Terms to Available Terms using the arrow buttons, then click OK.

to move the terms one at a time, highlight a term, then click or

to move all of the term, click on or
You can also move a term by double-clicking it.

5 Click Levels.

i.Trriiir'xinr.i
. t.l 1/ Ii.i'-. It.

I 4.*III!?4 m i l '

^ E ' I I I M I * ' ipitcMytnf w«* Iscli.r Umfa

i Selprl *-irl»Mrt ishveith* WHfttft**!
SfilKI kVtlli M i l l t< ttit

I M|

6 Do one of the following

To specify factor levels that are already stored in a worksheet column
- Choose Select variables stored in worksheet.

Appendix I: Minitab: A statistical Tool 185

- Under Levels, click in the first row and enter column containing the new levels of
the first factor. Then, use the Z key to move down the column and enter the
remaining factor level columns. Click OK.
To select levels from a list of the existing factor levels
- Choose Select levels from a list.
- Under Levels, click in the first row and choose the factor level from the drop-down
list. Then, use the Z key to move down the column and choose the
remaining factor levels. Click OK.
Options
Predict results dialog box
Store the predicted values in the worksheet (default)
Terms sub-dialog box
Choose terms to include in the prediction model
Levels sub-dialog box
Enter the new factor levels in coded or uncoded units, We will now predict results for
the selected experiments.

References
[3.1] G.S. Peace (1993). Taguchi Methods. Addison-Wesley Publishing Company.
[3.2] J.H. Lochner and J.E. Matar (1990). Designing for Quality. ASQC Quality
Press.
[3.3] W. Y. Fowlkes and C M . Creveling (1995). Engineering Methods for Robust
Product Design. Addison-Wesley Publishing Company.
[3.4] S.H. Park (1996). Robust Design and Analysis for Quality Engineering.
Chapman & Hall.
[3.5] M.S. Phadke (1989). Quality Engineering Using Robust Design. Prentice-Hall.

Appendix I: Minitab: A statistical Tool 186

 Appendix-2
Circuit Simulation using PSpice

Appendix-ll: Circuit Simulation using PSpice

2.1. Introduction:

The practical way to check an electric circuit performance is either build

hardware or conduct virtual experiments. Cadence OrCAD / PSpice (PC version
of Simulation Program with Integrated circuit Emphasis) is an industry standard
for circuit simulation and used in the present study. It is a general purpose circuit
simulation progrm, versatile, highly reliable and accurate and widely used in both
industries and universities for the product or process design. Again first approach
is expensive and time consuming, however the later is approach is preferred as it
allows the designer to quickly test a circuit over a variety of operating conditions
including temperature, parameter ageing effects and power supply variations,
etc., for obtaining highly accurate, fast and low cost circuit simulation and design.

I. Opening PSpice

II. Drawing the circuit

A. Getting the Parts

B. Placing the Parts
C. Connecting the Circuit
D. Changing the Name of the Part
E. Changing the Value of the Part
F. Making Sure You Have a GND
G. Voltage and Current Bubbles
H. Saving
I. Printing

HI. Probe

A. Before you do the Probe

B. To Start the Probe
C. Graphing
D. Adding/Deleting Traces
E. Doing Math
F. Labelling

Appendix-ll: Circuit Simulation using PSpice

 G. Finding Points
H. Saving
I. Printing

IV. Analysis Menu

A. AC Sweep
B. DC Sweep
C. Bias Point Detail
D. Parametric
E. Sensitivity
F. Temperature
G. Digital Setup
H. Transient

Note 1: There are other options, but these are the most useful ones.
Note 2: The Monte Carlo analysis was left out due to it's incredible depth. In
general PSpice will take a statistical look at a component while randomly
altering other values in the circuit, see references for pages that introduce this
method.

V. Types of Sources

A. Voltage Sources
i. VDC
ii. VAC
'Hi. VSIN
iv. VPULSE
v. PWL

B. Current Sources

VI. Digital Simulation

VII. References and Links

I. Opening PSpice:

189
Appendix-ll: Circuit Simulation using PSpice
 • Find PSpice on the C-Drive. Open Schematics or you can go to PSpice

A_D and then click on the schematic iconUi].

II. Drawing the circuit:

A. Getting the Parts:

• The first thing that you have to do is get some or all of the parts you need.
• This can be done by

o Clicking on the 'get new parts' button I 1 ** , or

o Pressing "Control+G", or
o Going to "Draw" and selecting "Get New Part..."
• Once this box is open, select a part that you want in your circuit. This can
be done by typing in the name or scrolling down the list until you find it.

P a r i Browser Basic
Part Name:

Description:
resistor

Close
R_var
RAMSKxl break
RAM8Kx3break Place
R break
readme Place & Close
ROM32KX8break
S break
SIN Help
SOFTLIM
SORT
STIM1
STIM16 Libraries..
STIM4
STIM8
SUM Advanced > >
Sw tClose

Full List

Some common parts are:

o r - resistor
o C - capacitor
o L - inductor
o d - diode

190
Appendix-ll: Circuit Simulation using PSpice
 o GND_ANALOG or GND_EARTH - this is very important, you
MUST have a ground in your circuit
o VAC and VDC
• Upon selecting your parts, click on the place button then click where you
want it placed (somewhere on the white page with the blue dots). Don't
worry about putting it in exactly the right place, it can always be moved
later.
• Once you have all the parts you think you need, close that box. You can
always open it again later if you need more or different parts.

B. Placing the Parts:

• You should have most of the parts that you need at this point.
• Now, all you do is put them in the places that make the most sense
(usually a rectangle works well for simple circuits). Just select the part and
drag it where you want it.
• To rotate parts so that they will fit in you circuit nicely, click on the part and
press "Ctrl+R" (or Edit "Rotate"). To flip them, press "Ctrl+F" (or Edit
"Flip").
• If you have any parts left over, just select them and press "Delete".

C. Connecting the Circuit:

• Now that your parts are arranged well, you'll have to attach them with
wires.
• Go up to the tool bar and

o select "Draw Wire" _^Jor

o "Ctrl+W" or
o go to "Draw" and select "Wire".
• With the pencil looking pointer, click on one end of a part, when you move
your mouse around, you should see dotted lines appear. Attach the other
end of your wire to the next part in the circuit.
• Repeat this until your circuit is completely wired.

191
Appendix-ll: Circuit Simulation using PSpice
 • If you want to make a node (to make a wire go more then one place), click
somewhere on the wire and then click to the part (or the other wire). Or
you can go from the part to the wire.
• To get rid of the pencil, right click.
• If you end up with extra dots near your parts, you probably have an extra
wire, select this short wire (it will turn red), then press "Delete".
• If the wire doesn't go the way you want (it doesn't look the way you want),
you can make extra bends in it by clicking in different places on the way
(each click will form a corner).

D. Changing the Name of the Part:

• You probably don't want to keep the names C1, C2 etc., especially if you
didn't put the parts in the most logical order. To change the name, double
click on the present name (C1, or R1 or whatever your part is), then a box
will pop up (Edit Reference Designator). In the top window, you can type
in the name you want the part to have.

Edit Reference Designator

Package
Reference Designate-! : IH
:'f
Gate: .:*
,zJ
Package T vpe:
(Footprint)
|RC05
zl

OK Cancel

• Please note that if you double click on the part or its value, a different box
will appear.

E. Changing the Value of the Part:

• If you only want to change the value of the part (if you don't want all your
resistors to be 1K ohms), you can double click on the present value and a

192
Appendix-ll: Circuit Simulation using PSpice
 box called "Set Attribute Value" will appear. Type in the new value and
press OK. Use u for micro as in uF = microFarad.

Set Attribute Value

VALUE

Cancel

• If you double click on the part itself, you can select VALUE and change it
in this box.

VI PartName: vdc X
Name Value
REFDES = VI SaveAttr ]

" REFDES=V1
TEMPLATE =V@REFDES %+ %• @DC
DC=0v'
SIMULATIONONLY=
PAR T= vdc
MODEL=
PKGREF=V1

W Include Non-changeable Attributes OK

l ^ Include System-defined Attributes Cancel

F. Making Sure You Have a GND:

• This is very important. You cannot do any simulation on the circuit if you
don't have a ground. If you aren't sure where to put it, place it near the
negative side of your voltage source.

G. Voltage and Current Bubbles:

• These are important if you want to measure the voltage at a point or the
current going through that point.
• To add voltage or current bubbles, go to the right side of the top tool bar

and select "Voltage/Level Marker" (Ctrl+M) JLJor "Current Marker" L£EJ. TO

193
Appendix-ll: Circuit Simulation using PSpice
 get either of these, go to "Markers" and either "Voltage/Level Marker" or
"Current Marker".

H. Saving:

I Hi
• To save the circuit, click on the save button U^Jon the tool bar (or any

other way you normally save files).

I. Printing:
• To print, you must first use your mouse to make a rectangle around your
circuit, this is the area of the page that will be printed. Then select print as
usual. (You can select E J ).

III. Probe:

A. Before you do the Probe:

• You have to have your circuit properly drawn and saved.

• There must not be any floating parts on your page (i.e. unattached
devices).
• You should make sure that all parts have the values that you want.
• There are no extra wires.
• It is very important that you have a ground on your circuit.
• Make sure that you have done the Analysis Setup and that only the things
you want are enabled.

B. To Start the Probe:

• Click on the Simulate button on the tool bar lM±U(or Analysis, Simulate, or
F11).
• It will check to make sure you don't have any errors. If you do have errors,
correct them.
• Then a new window will pop up. Here is where you can do your graphs.

C. Graphing:

194
Appendix-ll: Circuit Simulation using PSpice
 If you don't have any errors, you should get a window with a black
background to pop up (even with errors, it will be OrCAD PSpice A/D
Demo).
If you did have errors, in the bottom, left hand side, it will say what your
errors were (these may be difficult to understand, so go To "View - Output
File").

D. Adding/Deleting Traces:

• PSpice will automatically put some traces in. You will probably want to
change them.

• Go to Trace - Add Trace or ±non the toolbar. Then select all the traces
you want.
• To delete traces, select them on the bottom of the graph and push Delete.

E. Doing Math:

• In Add Traces, there are functions that can be performed, these will
add/subtract (or whatever you chose) the lines together.

I A d d Traces
Simulation Output Variables Functions or Macros
^^H
Analog Operators and Functions _^J
r """^^^^•_^_ W Analog u — n
V(0)
V[C1:1)
o
V(C1:2)
r +
—
V(C2 II
W Voltages
VIC2 2) /
V[R1:1) -
I Currents
V[R1:2) .At:SI l
V|R2:1) ARCTAN()
V|R2:2)
r ATANII
V(IJ2:+) AVG()
l ^ Alias flames
V(U2:-) AVGX(.)
V(U2:0UT) IMS | |
W Subcircuit Nodes
v'(U2:V+) DO
VIU2:V-) DB[)
VIV3:+) ENVMAXj,)
Vfv'3:-) ENVMIN(.)
vjva+) EXPO
V(V5:-) GO
VfVS:+) IMG()
V|VS:-) LOGO
V(Vin) L0G10O
107 variables listed
V(Vout) MO
V(X_U2.x_lm741.6) 2 MAX()
I Full List
i
Trace Expression: \j OK Cancel He)p I

195
Appendix-ll: Circuit Simulation using PSpice
 • Select the first output then either on your keyboard or on the right side,
click the function that you wish to perform.
• There are many functions here that may or may not be useful. If you want
to know how to use them, you can use PSpice's Help Menu.
• It is interesting to note that you can plot the phase of a value by using
IP(xx), where xx is the name of the source you wish to see the phase for.

F. Labelling: Click on Text Label I-*uon top tool bar.

• Type in what you want to write.

. Click OK
• You can move this around by single clicking and dragging.

G. Finding Points:

• There are Cursor buttons that allow you to find the maximum or minimum
or just a point on the line. These are located on the toolbar (to the right).
• Select which curve you want to look at and then select "Toggle Cursor"

• Then you can find the max, min, the slope, or the relative max or min ( l ^ \ l
is find relative max).

H. Saving:

• To save your probe you need to go into the tools menu and click display,
this will open up a menu which will allow you to name the probe file and
choose where to save it. You can also open previously saved plots from
here as well.

I. Printing:

• Select Print in Edit or on the toolbar UEJ.

• Print as usual.

IV. Analysis Menu

196
Appendix-ll: Circuit Simulation using PSpice
 Analysis Setup x|
E nabled Enabled

r AC Sweep... Options... Close

r Load Bias Point... r Parametric...

[^ ^aveBias Point... r Sensitivity...

r DC Sweep... I
r Temperature...

r Monte Carlo/Worst Case... r Transfer Function...

w Bias Point Detail r Jransient...

Digital Setup...

To open the analysis menu click on the J=Ubutton.

A. AC Sweep

• The AC sweep allows you to plot magnitude versus frequency for different
inputs in your circuit. In the AC sweep menu you have the choice of three
types of analysis:
• Linear,
• Octave and
• Decade.
• These three choices describe the X-axis scaling which will be produced in
probe. For example, if you choose decade then a sample of your X-axis
might be 10Hz, 1kHz, 100kHz, 10MHz, etc.... Therefore if you want to see
how your circuit reacts over a very large range of frequencies choose the
decade option.
• You now have to specify at how many points you want PSpice to calculate
frequencies, and what the start and end frequency will be. That is, over
what range of frequencies do you want to simulate your circuit.
• In the AC sweep you also have the option of Noise enable in which
PSpice will simulate noise for you either on the output or the input of the
circuit. These noise calculations are performed at each frequency step
and can be plotted in probe.
• The two types of noise are:
o V(ONOISE) for noise on the outputs and

l<)7
Appendix-ll: Circuit Simulation using PSpice
 o V(INOISE) for noise on the input source.
• To use input noise you need to tell PSpice where you consider the 'input'
in your circuit to be, for example, if your voltage source is labelled 'V1'.
• Finally you need to specify in what interval you want the noise to be
calculated (note: the default interval for spice is zero, i.e.: no noise will be
calculated).

B. DC Sweep

• The DC sweep allows you to do various different sweeps of your circuit to

see how it responds to various conditions.
• For all the possible sweeps,
o voltage,
o current,
o temperature, and
o parameter and global

you need to specify a start value, an end value, and the number of
points you wish to calculate.

• For example you can sweep your circuit over a voltage range from 0 to 12
volts. The main two sweeps that will be most important to us at this stage
are the voltage sweep and the current sweep. For these two, you need to
indicate to PSpice what component you wish to sweep, for example V1 or
V2.
• Another excellent feature of the DC sweep in PSpice, is the ability to do a
nested sweep.
• A nested sweep allows you to run two simultaneous sweeps to see how
changes in two different DC sources will affect your circuit.
• Once you've filled in the main sweep menu, click on the nested sweep
button and choose the second type of source to sweep and name it, also
specifying the start and end values. (Note: In some versions of PSpice you
need to click on enable nested sweep). Again you can choose Linear,
Octave or Decade, but also you can indicate your own list of values,
example: 1V 10V 20V. DO NOT separate the values with commas.

198
Appendix-ll: Circuit Simulation using PSpice
C. Bias Point Detail

• This is a simple, but incredibly useful sweep. It will not launch Probe and
so give you nothing to plot. But by clicking on enable bias current
display or enable bias voltage display, this will indicate the voltage and
current at certain points within the circuit.

D. Parametric

• Parametric analysis allows you to run another type of analysis (transient,

sweeps) while using a range of component values using the global
parameter setting. The best way to demonstrate this is with an example,
we will use a resistor, but any other standard part would work just as well
(capacitor, inductor).
• First, double-click the value label of the resistor that is to be varied. This
will open a "Set Attribute Value" dialog box. Enter the name {RVAL}
(including the curly braces) in place of the component value. This
indicates to PSpice that the value of the resistor is a global parameter
called RVAL. In order to define the RVAL parameter in is necessary to
place a global parameter list somewhere on the schematic page. To do
this, choose "Get New Part" from the menu and select the part named
param.
• Place the box anywhere on the schematic page. Now double-click on the
word PARAMETERS in the box title to bring up the parameter dialog box.
Set the NAME1= value to RVAL (no curly braces) and the VALUE1= value
to the nominal resistance value. This nominal value is required, but it is
only used if the DC bias point detail is computed. Otherwise, the value is
ignored by PSpice.
• Finally, go to the "Analysis Setup" menu and enable "Parametric" analysis.
Open the Parametric setup dialog box and enter the sweep parameters:
Name: RVAL Swept variable type: Global Parameter. Make sure the other
analysis type(s) are selected in the analysis setup menu (transient,
sweeps). PSpice will now automatically perform the simulation over and
over, using a new value for RVAL during each run.

199
Appendix-ll: Circuit Simulation using PSpice
E. Sensitivity

• Sensitivity causes a DC sensitivity analysis to be performed in which one

or more output variables may be specified.
• Device sensitivities are provided for the following device types only:
o resistors,
o independent voltage and current sources,
o voltage and current-controlled switches,
o diodes, and
o bipolar transistors.
• You would use the sensitivity setting for discovering the maximum range
of circuit performance and the causes of extreme operation. These
techniques are used to identify effective changes to improve the quality of
circuit operation (for example, which components need to have tight
tolerance and which can be lower quality and less expensive).
• This isn't as important for us in the lab, but some day when you are
constructing real circuits that need to function under various conditions
this will be useful.

F. Temperature

• The temperature option allows you to specify a temperature, or a list of

temperatures (do not include commas between temperature values) for
which PSpice will simulate your circuit.
• For a list of temperatures that simulation is done for each specified
temperature.

3: Displaying the simulation Results

PSpice has a user-friendly interface to show the results of the simulations. Once the
simulation is finished a Probe window will open.

200
Appendix-ll: Circuit Simulation using PSpice
 1. From the TRACE menu select ADD TRACE and select the voltages and
current you like to display. In our case we'll add V(out) and V(in). Click OK.
2. You can also add traces using the "Voltage Markers" in the schematic. From
the PSPICE menu select MARKERS/VOLTAGE LEVELS. Place the makers on
the Out and In node. When done, right click and select End Mode.

Add T faces
Simulation Output Variables Functions or Macros

KC1) V Analog
1(11]
l(R1)
l(R2)
r
1(V1) V Voltages
V(0]
V(C1:1) V Currents
(51
V(C1:2) ABS|)
V(I1:+) V Power ARCTANI.)
'.,'(11 ATAN()
AVG()
(OUT)
r • AVGK( . )
VIR11] v Alias Names C0S()
V(R1:2) DO
V(R2:1) DB()
V(R2:2) ENVMAX(. ]
VtV1:+) ENVMINI .)
V(V1:-J EXP()
V1(C1) G()
V1(I1) IMG{)
V1(R1) LOG( |
V1(R2) LOG10I I
Vlfvl] 34 variables listed M( J
V2(C1) MAX()
Full List

If ace Expression: JV(0UT)V|IN) OK | Cancel Help

Add Traces window

201
Appendix-ll: Circuit Simulation using PSpice
G. Digital Setup

• This paragraph will only indicate the features of the digital setup
analysis menu, see below for a more complete description.
• In addition to letting you simulate analog circuits, PSpice provides a n i
of digital parts that can be used in a homogeneous digital circuit
heterogeneous analog/digital combination. The digital analysis option a
you to specify the timing of your circuit, by running the gates at
minimum, maximum and typical values. A superb feature allows you to
the worst case timing of your circuit to see how it will operate under tht
extreme conditions. You also have the option of setting the value of any
flops you have in your circuit to predefined states which is good to Simula
any startup conditions for finite state machines that you are simulating.

H. Transient: The transient analysis is probably the most importan

analysis you can run in PSpice, and it computes various values of your
circuit over time. Two very important parameters in the transient
analysis are:

o print step
o final time.
• The ratio of final time: print step determines how many calculations
PSpice must make to plot a wave form. PSpice always defaults the start
time to zero seconds and going until it reaches the user defined final time
It is incredibly important that you think about what print step you should
use before running the simulation, if you make the print step too small the
probe screen will be cluttered with unnecessary points making it hard to
read, and taking extreme amounts of time for PSpice to calculate
However, at the opposite side of that coin is the problem that if you set the
print step too high you might miss important phenomenon that are
occurring over very short periods of time in the circuit. Therefore play with
step time to see what works best for your circuit.
• You can set a step ceiling which will limit the size of each interval, thus
increasing calculation speed. Another handy feature is the Fourier

202
AnnfinHiv-ll- Cirmit Simulation usinn PSnire
 analysis, which allows you to specify your fundamental frequency and the
number of harmonics you wish to see on the plot. PSpice defaults to the
9th harmonic unless you specify otherwise, but this still will allow you to
decompose a square wave to see it's components with sufficient detail.

V. Types of Sources

A. Voltage Sources

i. VDC

o This is your basic direct current voltage source that

simulates a simple battery and allows you to specify
the voltage value.

ii. VAC

o A few things to note about the alternating current

source, first PSpice takes it to be a sine source, so if
you want to simulate a cosine wave you need to add
(or subtract) a 90° phase shift. There are three values
which PSpice will allow you to alter, these being:
• ACMAG which is the RMS value of the
voltage.
• DC which is the DC offset voltage
• ACPHASE which is the phase angle of the
voltage
o Note that the phase angle if left unspecified will be set
by default to 0°

iii. VSIN

203
Appendix-ll: Circuit Simulation using PSpice
 .» . •
uaUi.»>ii.iMliafriii.M
Name V :lll.C

r, Save Mb
'F these values Ch sngc Display
AC = 10v
VOFF ilv Offset Voltage ",.--' y"
VAMPI 14.14V
1 HI i) :'nril Frequency m H i
r i i ii
•
1
I'HAlil II

V Include Ngn t.h^i geaUe Attributes OK

r" lni IHIIK System i eflned Attributes Cancel

3 The SIN type of source is actually a damped sine with time

delay, phase shift and a DC offset. If you want to run a transient
analysis you need to use the VSIN see how AC will effect your
circuit over time. Do not use this type of source for a phasor or
frequency sweep analysis, VAC would be appropriate for that.
• DC the DC component of the sine wave
• AC the AC value of the sine wave
• Voff is the DC offset value. It should be set to
zero if you need a pure sinusoid.
• Vamplitude is the undamped amplitude of the
sinusoid; i.e., the peak value measured from
zero if there were no DC offset value.
• FREQ is the frequency in Hz of the sinusoid.
• TD is the time delay in seconds. Set this to
zero for the normal sinusoid.
DF is the damping factor. Also set this to zero
for the normal sinusoid.
• PHASE is the phase advance in degrees. Set
this to 90 if you need a cosine wave form.
o Note that the normal usage of this source type is to set Voff,
TD and DF to zero as this will give you a 'nice' sine wave.

iv. VPULSE

o The VPULSE is often used for a transient simulation of a circuit

where we want to make it act like a square wave source. It should

204
Appetflix-ll: Circuit Simulation using PSpice
 never be used in a frequency response study because PSpice
assumes it is in the time domain, and therefore your probe plot
will give you inaccurate results.
• DC the DC component of the wave.
• AC the AC component of the wave.
• V1 is the value when the pulse is not "on." So
for a square wave, the value when the wave is
'low'. This can be zero or negative as required.
For a pulsed current source, the units would be
"amps" instead of "volts."
> V2 is the value when the pulse is fully turned
'on'. This can also be zero or negative.
(Obviously, V1 and V2 should not be equal.)
Again, the units would be "amps" if this were a
current pulse.
• TD is the time delay. The default units are
seconds. The time delay may be zero, but not
negative.
• TR is the rise time of the pulse. PSpice allows
this value to be zero, but zero rise time may
cause convergence problems in some transient
analysis simulations. The default units are
seconds.
» TF is the fall time in seconds of the pulse.
• TW is the pulse width. This is the time in
seconds that the pulse is fully on.
• PER is the period and is the total time in
seconds of the pulse.
o This is a very important source for us because we do a lot of
work on with the square wave on the wave generator to see
how various components and circuits respond to it.

v. PWL (Piece-Wise Linear)

205
Appendix-ll: Circuit Simulation using PSpice
 o The PWL source is a Piece Wise Linear function that you can use
to create a wave form consisting of straight line segments drawn by
linear interpolation between points that you define. Since you can
use as many points as you want, you can create a very complex
wave form This source type can be a voltage source or a current
source.
o The syntax for this source type is flexible and has several optional
parameters. The required parameters are two-dimensional points
consisting of a time value and a voltage (or current) value. There
can be many of these data pairs, but the time values must be in
ascending order, and the intervals between time values need not
be regular.
o The two optional parameters are "DC" and "AC." The use of an AC
parameter with this source is not very well documented and
because this source is intended for use with a transient analysis
any AC value would be ignored.

B. Current Sources

• For any of the previous discussed voltage sources, there exist the exact
source except that they produce current. There is one thing that should be
mentioned; current sources in PSpice get a little confusing. For those
current sources whose circuit symbol has an arrow, you have to point the
arrow in the direction of conventionally flowing current. This applies to all
current sources, including AC and DC. Therefore placing the current
source in the circuit backwards with seemingly incorrect polarities will give
the correct results.
• An interesting little feature under the markers menu is the ability to add
markers to your circuit so you can see where the current and voltage have
imaginary values in the circuit, and the phase of your source at any point
in the circuit.

VI. Digital Simulation

206
Appendix-ll: Circuit Simulation using PSpice
 • PSpice can simulate digital circuits and Probe can output a timing diagram
showing the relationship between all the signals propagating in the circuit.
The following will be a brief introduction to digital analysis using PSpice,
you should consult the online PSpice manual if you are unsure about any
of the following properties.
• The evaluation version of PSpice provides many of the common digital
parts that we use in the lab exercises these include, but are not limited to,
o counters,
o multiplexers,
o decoders,
o flip-flops,
o latches,
o all common gates,
o buffers, .
o adders and a lot more.
. To begin, create a digital circuit the same way you would create an analog
one by getting and placing the parts (see above for details on creating a
circuit).
. Now instead of an analog voltage source you want to place digital stimuli,
these are located in the parts menu as HI and LO.
. Draw wires connecting the parts like usual, but whenever you have an
output of a gate that would be an output to your circuit you need to
terminate it with a block.
• You can add these blocks to your circuit. On the tool bar there is a button
labelled Add Block; connect this block to any output signal of your circuit.
. Then run a bias point analysis and click on the button to see the enable
voltage bias display, to see the logic level on each of your lines.
. You can run a transient analysis at this point and plot the signals, but they
will just be straight lines, showing the logic level of each signal. However a
digital circuit that remains in a constant state all the time is of little
interesting. Eventually, you are going to encounter a circuit that needs to
be stimulated with a clock pulse.

207
Appendix-ll: Circuit Simulation using PSpice
 To add a time varying 'clock pulse' to your circuit, enter the parts menu
and get a part labelled STIM1 and place that as the input to a device.
Double click on the STIM1 to open up a dialogue box that looks like this:

DSTM1 PartName: STIM1

Name Value
COMMAND2 = 1m 1 Save Attr
TIMESTEP=1m
Change Display
COMMAND1=0s 0
C0MMAND2=1m1
C0MMAND3=2m 0 Delete
COMMAND4=3m 1
C0MMAND5=4m 0
COMMAND6=5m 1
COMMAND7=6m 0

[x Include Non-changeable Attributes OK

|x Include System-defined Attributes Cancel

and fill it in with the Time Step that you want your 'clock' to have.
This Time Step value is the your 'clock' pulse. There are various
ways you can fill out the command prompts, one way is to do it as
shown in the picture, however here are some simple commands
that will allow you to do a range of simulations:

o REPEAT FOREVER, to simulate a real

clock
o REPEAT <n> TIMES
o END REPEAT
o « t i m e » INCR « v a l u e »
o « t i m e » DECR « v a l u e »
o for «value» you can use,
0,1,R(rising), F(falling), X(unknown), or
Z(high impedance)
There are other useful commands that are documented well in the online
manual. If you are interesting in seeing the voltage levels attached to
these gates to see if you need to buffer any signals or to see if you are
exceeding the fan out and/or fan in of any gate, you need to add dropping
resistors to the inputs of gates, and a load resistor the outputs of the gates

208
Appendix-ll: Circuit Simulation using PSpice
 and simulate again. When you are running the transient analysis, don't
forget to change your print step and the final time so you will actually be
able to see the results.
• If your digital circuit contains a counter, or decoder or some other device
with many outputs that are all going to a single destination, instead of
drawing many individual signal lines from each source to the proper
destination you can use a bus. In the draw menu at the top of the screen
is an option for drawing a bus which is a thick wire able to carry many
separate signals to a single place. The order in which the signals go on
the bus is the order in which they come off at the destination.

VII. References and Links

1. This page is not very in-depth, but it covers a lot of good concepts.

http://www.home.clc.cc.il.us/home/eng516/pspice.htm#introintro

2. PSpice help page: http://www.coe.uncc.edu/project mosaic/PC/PSpice/

3. For the more obscure questions you might have go right to the source at
Oread© http://www.orcad.com/
4. In the PC221 text, at the end of each chapter, there is information on how
to use PSpice.
5. Electronic Devices and Circuit Theory ,7th Edition Boylestad and
Nashelsky, Prentice Hall.

Note:AII pictures and screen shots came from OrCAD/ PSpice evaluation
version 9.2

209
Appendix-ll: Circuit Simulation using PSpice