Power Electronics

Prof. B.G Fernandes

Department Of Electrical Engineering
Indian Institute of Technology, Bombay
Lecture No - 13
Let me recapitulate whatever that I did in my last class. In one of the configurations of half
control bridge, we found that as alpha tends to pi, the incoming thyristor fails to turn on and
therefore the outgoing thyristor fails to turn off. As a result, the output voltage wave form if I
see, from 0 to pi we have a freewheeling period and from pi to 2 pi output voltage is same as the
input voltage. So, there is source supplies power only in 1 of the half cycles and in the remaining
half cycle, there is a complete freewheeling part provided, the load current is continuous.
So, output voltage is same as that of a half wave rectifier. In a very first lecture on converters, if
you see the wave form, they are both the same. Hence the name, half waving effect. Then second
point what we discussed was the expression for the average value of the output voltage; is given
by Vm by pi into 1 plus cos alpha. Remember, this expression is valid only if the load current is
continuous. Remember, load current should be continuous, only then this expression is valid
because we got this expression by integrating the output voltage.
The output voltage is Vm sin omega t from pi to alpha to pi and pi to pi plus alpha, it is 0, is a
freewheeling period and from pi plus alpha to 2 pi, again it is Vm sin omega t. So, only when you
integrate this wave form, you will get the average value is Vm by pi into 1 plus cos alpha. So,
only if the current is continuous, this is the output, output voltage is given by V0 Vm by pi into 1
plus cos alpha. Remember, invariably students make a mistake while solving a problem. The
moment they see a half controlled bridge, they immediately, automatically, they write average
value of output voltage is Vm by pi into 1 plus cos alpha. Even though current is discontinuous,
do not make this mistake.
Now, all this time we have discussing about controlled rectification. So, I have a question to ask.
Why do we require the variable voltage DC supply? Why, why do we require variable voltage
DC supply? Now, let me take a diversion, in the sense, I will go back to machines. In my
introductory lecture I told you that in power electronics, to appreciate power electronics, you
need to have a really a good back ground, a good knowledge of machines. Little bit of everything
is there in power electronics, remember. So to appreciate power electronics, a reasonably good
knowledge of machines is essential. Let me review DC machines.

(Refer Sllide Time: 4:44)

A separaately excited DC machin

ne, very popu
ular equationn, Vin is equual to Eb, is tthe back emff plus
Iara, Vin is equal to Eb plus Iara, Eb is the back
k emf which proportionaal to flux andd speed. Thiss flux
is proporrtional to thee field current. So therefore, Eb is ggiven by k innto If into om
mega. Thereefore,
omega iss given by Vin Vin, the in
nput voltagee to the armaature; Input voltage to tthe armaturee, it is
not the in
nput voltagee to the moto
or or someth
hing like thatt. In the sensse, if I conneect the resisttor in
series, so
o that is not Vin. Vin is a voltage
v
appllied to the arrmature term
minals A, AA
A, remember.
Vin divid
ded by k into
o If minus Iara divided by
y k If. Your tteacher wouuld taught maachines, he m
might
have told
d you that fo
or all the prractical purp
poses a separrately exciteed DC machhine is a connstant
speed maachine.
uction in speeed with load
d is not mucch. This term
m is generallly small beccause
In other words, redu
w
comess in the path of the main power flow
w should be vvery small. If ra is
armature resistance which
high, i square
s
ra, i square r lossses are high
h, efficiencyy comes doown. So, arm
mature resisstance
should be as small as
a possible. Moreover, armature
a
reaaction armatture reactionn what it doees? It
tries to reduce
r
the air
a gap flux. So thereforre, armaturee reaction trries to keep the speed oof the
motor alm
most constan
nt.
So thereffore, for all practical
p
purrposes, separrately excitedd DC machiine is a consttant speed m
motor.
Applicatiion requires the speed variation in sp
peed over a wide range.. What should I do? I haave to
change th
he no load sp
peed. This teerm is indep
pendent of thhe armature current. So, I have to chhange
the no load speed.
So we fo
ound that om
mega, the no load speed is proportioonal to the innput voltagee to the armaature,
Vin and inversely
i
prroportional to
t the flux or
o If. So, if I have to ddecrease the speed below
w the
rated, I will
w apply a reduced vo
oltage to thee armature. S
So, speed iss directly prroportional tto the
input volltage to the armature.
a
So
o, for any speeed from 0 too the rated, Vin should vvary from 0 tto the
rated voltage of the machine
m
and
d I will keep the flux connstant.
2

If I want to increase the speed above the base or above the rated, I will keep voltage applied to
the armature at its rated value because a general, a good engineering practice is; one should not
apply a voltage higher than the rated to the motor, a general engineering practice, a good
engineering practice. So, I will try to reduce the field flux. So, my speed, no load speed
increases.
So therefore, from 0 to base speed, applied voltage to the armature terminals should be varied
and from the speed above the base, decrease the field flux or decrease the field current. So, both
these regions or if I have to vary the applied voltage to the armature or if I have to vary If, I need
to have a variable DC supply. You may say that I will connect a rheostat in series. That is what
we do in our machines lab. It is basically, highly inefficient way of handling power.
Having done a course on power electronics, do not use a resistor in series of any circuit. It is
crime. Having done a course on power electronics, your solution should not contain the use of,
resistor comes in a main part of the power flow, just not accepted. That is what I said, one of the
ways, may be our teacher, machine teacher, may the one know who taught machines, might have
told that connect a resistor in series with the armature and connect resistance in series with field,
vary them to vary the speed. That is fine in course 1 or that is fine in machines course.
Definitely, this solution is not acceptable once you do a course in power electronics. So, coming
to power electronics again, both of them require a variable voltage DC power supply. So, I can
have a semi control or half controlled bridge here. I can have a half, I can, let us see, as the
power rating increases what modification do I need to make in the power circuit. One of the
options is I can have, both the cases I can have a semi controlled bridge. May say, why not use
half wave controlled rectification using 1 thyristor?
What did we found yesterday? I solved one problem, a very good problem. The moment I use
only 1 thyristor between the source and the load or in other words, in half wave controlled
rectification, we are drawing a DC current from the source. In other words, average value of the
current drawn from the source is finite. Source does not supply any power beyond pi if I use
freewheeling diode, you observed that.
So, there is a DC component coming from the source plus the harmonics. Harmonics are there
even in both the cases, whereas, in this case if I use a semi controlled or half controlled bridge,
DC component is 0. There is a pulse or source supplies power from alpha to pi in the positive
half and pi plus alpha to 2pi in the negative half. So, average value of the source current is 0. So,
DC component is 0.
So, half wave rectification could be a solution for a small motor, a small, a fractional hp DC
motor, use of half wave rectification may be justified because small hp motor cost is very low.
So, might controller also, the so called power processing equipment should also be very low. I
cannot have a very expensive equipment to vary the speed of a small motor.
So, varying speed of a small motor, half wave rectification could be a solution because current
that I am drawing from the DC, the average value the current I am drawing from the source is
3

very smaall. I told you

y
yesterdaay that I will
w solve a very intereesting problem. See, a very
interestin
ng problem, lot of thingss are there, trry to undersstand. It is noot very simpple but lookss very
simple.
3:38)
(Refer Sllide Time: 13

ooks invariabbly, you knoow, you willl get cheatedd. So,

My sinceere request to you is do not go by lo
the circu
uit is very sim
mple. May be,
b the perfo
formance maay not be ass simple as tthat, remembber. I
have a simple
s
transformer, inpu
ut winding, primary is connected to the sourrce, secondaary is
connected to or secon
ndary is supp
plying a resiistive load thhrough a dioode.
The mom
ment I conneect a diode in
n the circuitt, circuit is nno longer linnear, remembber. Linear iin the
sense, thee circuit doees not or in other
o
words, the law of ssuper positioon does not hhold good foor this
circuit. Circuit
C
is non linear. Inp
put voltage is sinusoidaal, therefore I mean I am
m there is a very
small win
nding resistaance and maay be a very small leakagge flux also.. So, if V1 iss sinusoidal, I can
safely assume that E1 is also sinu
usoidal.
Since, wiinding resisttance and leaakage flux are
a very smaall, I can safe
fely assume tthat V1 and E1 are
sinusoidaal. V2 is also
o sinusoidal. Output volltage, voltagge across R is voltage aacross R is a half
wave recctified.

(Refer Sllide Time: 15

5:35)

The seco
ondary voltaage, output voltage
v
acrosss R. Load iis purely ressistive, curreent wave form is
also halff wave rectifiier. So, curreent flows in the load, rem
member, currrent flows inn the load frrom 0
to pi. Thee expression
n is given by
y secondary sinusoidal
s
vooltage divideed by r that is if I say soome x
into sin omega
o
t, so expression
e
for
fo i is x into sin omega t divided by r and from ppi to 2 pi, cuurrent
is 0.
So in oth
her words, seecondary doees not supply
y any powerr to the load from pi to 2 pi. So, if I write
the 4ier series
s
or 4ier spectrum for
f the curreent, it has a D
DC componnent, I discusssed yesterdday. It
has averaage value is finite. So, itt has a DC component.
c
T
Then I have a fundamenntal componeent, x
sin omeg
ga t or the frequency
f
off the first teerm is same as that of tthe input annd it has a hhigher
omega teerms.
Now queestion is how
w does the source
s
curren
nt look like?? What are tthe principlees? What aree the,
our mach
hine teacherr what has he has told
d us 1 thinng that is inn a transforrmer for thee AC
componeents, there has
h to be an ampere turn
n balance. IIf there is ann AC currennt flowing iin the
secondarry, there has to be an eq
quivalent AC
C current in tthe primary.. Neglectingg the magnettizing
current, ampere
a
times N1 I1 shoulld be equal to N2 I2, maggnetizing currrent is negleected.
Secondarry current has
h a sinusoiidal, sinusoiidal componnent has a D
DC componeent. I will rrepeat
secondarry current haas a DC com
mponent. Thee flux in thee core has a sinusoidal ccomponent w
which
is producced by magn
netizing curreent which in
nduces sinus oidal voltagge in the prim
mary as well as in
the secon
ndary and itt can have a DC compo
onent, becauuse DC com
mponent in tthe core doees not
induce an
ny voltage ev
ven in the prrimary or seccondary.
I cannot have DC component in E1 or the secondary volttage E2. In oother words, you cannot have
mponent eith
her in E1 or E2. To have a DC compponent, Whyy am I sayingg this? To hhave a
a DC com
DC comp
ponent in E1 or E2, flux is
i a ramp beecause E is d phi by dt. To have a cconstant E, pphi is
should bee a ramp. In other wordss, if flux is a ramp then thhere is no stteady state, iis that okay??
5

So remem
mber, the flu
ux in the corre has a DC component
c
w
which does nnot induce aany voltage iin the
primary as well as in
n the second
dary. You caannot have a DC compoonent either in E1 or E2. Why
am I sayiing that? To
o have a DC component in E1 or E2, flux shouldd be a ramp, should varyy with
t. So, if flux
f
varies with
w t only then d phi by dt
d is a constaant. Since, fl
flux is varyinng with t, so there
is no stud
dy state. So, we have to eliminate that.
(Refer Sllide Time: 20
0:20)

The mag
gnetizing the sinusoidal flux
f
produceed by the maagnetizing cuurrent drawnn from the soource
is sinusoiidal. For eveery high freq
quency comp
ponent whichh is there in the secondaary, an equivvalent
current should
s
be there in the primary, remember. For every high frequuency sinussoidal
componeent which iss present in
n the secon
ndary currennt, there shoould be an equivalent high
frequency
y current in the primary. You canno
ot have a DC
C current in tthe primary, why am I saaying
this?
Assume that there is a DC com
mponent or if there is a DC compponent, see, there is a small
resistance, winding resistance here.
h
See, I will moduulate an inpuut voltage, a small winnding
resistance and E1. I said
s V1 is sin
nusoidal. I said E1 cannoot have a DC
C componennt, rememberr. So,
f
because if I have the DC com
mponent, iff i average iis finite, thee DC
if I apply KVL, it fails
componeent into r, an
n average DC
C voltage is required
r
herre. There is a voltage droop across r w
which
is a DC. Input is AC which is sin
nusoidal, in this
t voltage iis also sinusoidal, so KC
CL fails.
So, you cannot any circuit cann
not violate th
he physical llaws. Whethher it is pow
wer electroniics or
member. Soo far if you ssee, the entirre, in my couurse I
any subjeect cannot violate the baasic laws, rem
have used KVL and may be, maaximum firstt order diffeerential equaation, R L annd E and inpput is
sinusoidaal, only this much
m
nothin
ng else.
So, KVL
L has to hold
d good. So, I will repeat if there is a DC componnent in the soource, I havee told
you that winding haas a small reesistance. i into
i
r is goiing to be a D
DC voltage,, input voltaage is
6

sinusoidaal. Induced EMF

E
is agaiin sinusoidall. You cannoot have a DC
C componennt there. So, KCL
fails.
Therefore, your inpu
ut current hass no DC com
mponent. It hhas all the coomponents oof AC compoonent
nt which is th
here in the seecondary, th
he equivalentt and the funndamental coomponent.
of curren
(Refer Sllide Time: 23
3:52)

nt look like?? Secondaryy current, thee primary cuurrent shoulld not

So, how does the primary curren
have a DC
D componeent. Rest alll the compo
onents shoulld be there in the prim
mary. So, priimary
current lo
ooks like thiis. It is like, I do not kno
ow how manny of you haave seen the waveform iin the
oscilloscope, if I keeep a waveforrm which is having
h
a DC
C bias and keeep it in AC
C mode, how
w does
it look liike? The DC
C bias goes, I get a waaveform sym
mmetrical along x axis. If I keep inn DC
mode, au
utomatically
y I get a DC
C bias. You go back annd check if your if youu not observved a
waveform
m on the osccilloscope haaving a DC bias.
b
You givee a sinusoid,, using a sig
gnal generato
or give a DC
C bias, keepp it in ac moode. The mooment
you keep
p it in ac mode,
m
the DC
C bias disap
ppears, vani shes. You gget a wavefo
form symmeetrical
about thee x axis and
d if you keep
p it in DC mode;
m
whateever the wav
aveform be, the DC biass will
come on the screen. So, same herre, this waveeform, actuaally is the waaveform whiich you can see it
in the osccilloscope with
w the DC bias.
b
So, rem
move the DC bias I will gget this waveeform. So, soource
current lo
ooks like thiis. Source do
oes not supp
ply a DC com
mponent. Buut then, I toldd you that fllux in
the core can
c have a DC
D componeent. So what??

(Refer Sllide Time: 25

5:46)

So, flux in the core has a DC component, phi DC annd a sinusoidal componnent. This fllux is
produced
d by the mag
gnetizing current. So, flu
ux in the corre is phi DC
C plus phi fuundamental. So, I
have a waveform,
w
a sinusoidal
s
waveform
w
hav
ving a DC shhift or DC bbias. What is the effect? W
What
is the con
nsequence? What
W will haappen if therre is a flux, a DC flux inn the core?
Again, our
o machinee teacher miight have to
old was thaat or told yoou that of ccourse, the same
machine teacher who
o thought me
m has also told
t
me thatt invariably, we operate the machinnes or
any magn
netic circuit at the knee point. Knee point is a point whereinn the saturation just begiins to
start. So,, all the mag
gnetic circuitts when I ap
pply the rateed voltage, itt produces a rated flux, rated
AC flux,, I am talkin
ng about AC
C now, AC flux. So, thee operating point is thee knee point, it is
assumed..
But then,, in this casee because of half wave reectification, there is a DC flux. So, m
my excitatioon has
increased
d in 1 cycle or
o in 1 cyclee there is an increase andd in another cycle there is a decreasee. So,
what will happen to the bh loop?? If I see, I have
h
a non ssymmetrical bh loop. Ass the value oof DC
flux increeases, I may
y have a flux
x in the core or the value of the DC fflux may be equal to the peak
value of this
t flux.
So, it maay so happen
n that it may
y not even come
c
down tto negative and it may sso happen thhat in
that case, flux or coree may saturaate. What haappens if the core saturattes? They haave been toldd that
if the corre saturates, core loss, your
y
current is going to be peaky. T
The moment current beccomes
peaky, th
he harmonicss, the effect of harmoniccs again, I w
will discuss ssometime laater. So thereefore,
if there iss a DC flux in the core, in some parrt of the cyclle, core mayy get saturateed and if thee core
saturates, a source cu
urrent drawn
n from the so
ource may beecome peakyy and the mooment it beccomes
peaky, it is not a sinu
usoidal now, no longer a sinusoidal. Source cannnot become ppeaky.
As harm
monics and the
t effect off harmonicss, we will ddiscuss someetime later. Of course, core
saturates, core loss also
a increasee, core heatin
ng, what nott? So, theree is a probleem with or tthis is
8

the prob
blem with half
h
wave rectification.
r
. We are ddrawing a D
DC componnent. If thiss DC
componeent flows in the transform
mer core, transformer ccore may gett saturated; ccore loss, cuurrent
becoming
g peaky and
d harmonics. So that is th
he reason, hhalf wave recctification iss only used ffor or
used for low power application.
a
When I am
a saying lo
ow power, cu
urrent drawn
n from the soource is alsoo very small.. So, the effeect of
saturation
n is not mucch. High pow
wer you cann
not use halff wave rectiffication. So, I hope all of you
understoo
od the effecct of drawin
ng a DC currrent from thhe source. M
Magnetics m
may get satuurated
bottom liine.
(Refer Sllide Time: 30
0:23)

Let me solve
s
anotheer problem. A half contrrolled bridgee, 2 thyristoors in 1 leg, load is RE. The
moment I am saying when load is
i RE, first th
hing that shoould come too your mind is load currrent is
discontin
nuous. First thing that sh
hould come to mind is RE is, loadd is invariablly discontinnuous.
The prob
blem says determine the average valu
ue of load cuurrent for alppha is equal to 0.
Nothing has been saiid whether the
t current iss discontinuuous or curreent is continuuous. By loooking
ould be ablee to tell that current
c
is diiscontinuouss, like a docttor. When hee sees
at the cirrcuit, you sho
a face off a patient, hee can undersstand whetheer the patiennt is having aany illness oor not. That is the
way wheen you are doing,
d
having
g done half controlled bbridge or so far having ddone a courrse on
power electronics, th
he moment I see the circcuit, I shouldd be able to tell or I shoould make a wild
guess thaat whether th
he current is continuous or not.
o not need to
o or you do not need to mention thaat current is discontinuouus. By seeinng the
Here I do
circuit, in
nput is sinu
usoidal, outp
put is R and
d E. E alwayys opposes, R voltage aacross the R and
current flowing
f
theree are in phaase. So, current is disconntinuous. Seee the wavefform. We have a
sinusoid,, this EG. So,
S SCR getts forward biased
b
only when instaantaneous vaalue of the input
voltage is
i higher thaan 100 volts, again I am
m neglectingg the device drop. It hass been mentiioned

that neglect the device drop. If I consider device drop, invariably we consider 2 volts for
thyristor. May be, 1.5 volts for a diode, again, depends on the current rating.
Instantaneous value of the input voltage should be slightly higher than 100. So, having said that
thyristor gets forward biased only at 0.1 pi or sin inverse 100 divided by 230 into root 2 is 18
degrees is 0.1 pi
So, beyond 0.1 pi if you want, you can trigger and thyristor starts conducting. But in the problem
it is mentioned that thyristor is triggered at 60 degrees. So, 0 to 60 degrees, no current is flowing
in the load. So, to draw the load voltage waveform, remember I might have told you many times
that you need to know the type of load.
If the current is continuous, you do not need to know the type of load because the load voltage is
either the input voltage when the thyristors in the powering mode when the source supplies
power to the load, the load voltage is same as the input voltage and in the freewheeling mode,
load voltage is 0. So, you have either the input voltage or 0, whereas, if the load is if the load
current is discontinuous, you need to know the type of load. Here it is RE. So, load voltage is E,
the magnitude of E.
So, 0 to 60 degrees output voltage is 100 volts. At 60 degrees, you have triggered the bridge. So,
T1 and D2 starts conducting in the positive half. The equivalent circuit is T1, load, D2 back to
source. So, input voltage is same as the output voltage or output voltage is same as the input
voltage if I neglect the device drops. Otherwise, it is less by a very small value, may be of the
order of 4 to 5 volts. So, the input is 230 minus 4, does not matter, still 230 for an engineer.
Where does it stop conducting? Load is RE, I did tell you that the difference Vi minus E is across
R. So, voltage across R is nothing but i into R. So, V minus E is 0. So therefore, i should be 0. So
at this, at pi minus alpha min, this is the instant. Wherein, E is equal to the input voltage Vi,
current becomes 0. So, current become 0 at 0.91 pi.
I repeat you could have triggered the thyristor at 0.1. If I had diode here, T D1 and D4 instead of
T1 and T4, D1 would have started conducting at 0.1 pi and would have stopped conducting at 0.9
pi because they get forward biased at that instant. We have here thyristors and they are triggered
at 60 degrees. It is forward biased, it starts conducting but then at 0.9 pi, instantaneous value of
the input voltage is same as the value of E. So, Vi minus E is 0. So, Vi minus E is nothing but i
into R. So, i has to be 0. So, current becomes 0 at 0.9 pi.
So, thyristor turns off, thyristor T1 and D2 turn off at 0.9 pi and from 0.9 pi till you trigger the
next pair that is only at pi plus alpha that is at 240 degrees, current flowing in the load is 0. No,
current flows. So, load voltage is the value of E. So, that is what. Be careful, half controlled
bridge, alpha is 60 degrees. Mechanically do not write, Vm by pi into 1 plus cos alpha. Invariably
you will make this mistake, remember do not ever do this.
First you find out, first you draw the voltage waveform V0 and then you find out the average
value. You make, you can make a mistake while integrating. I told you, this is a course on power
electronics. You cannot make a mistake while drawing the output voltage waveform. You can,
may be you can make a mistake while simplifying, may be. I may not deduct, as a teacher I may
10

not deduct any marks. But then, if you makee a mistake w

while drawinng the waveforms, definnitely,
I will ded
duct marks. It
I just gives to a feeling that you havve not underrstood powerr electronics.
So, outpu
ut voltage iss from 0 to pi
p by 3. It iss E from pi bby 3 to 0.9 pi. It is Vi ssin omega t input
voltage is same as in
nput voltage is same as output
o
voltagge and from
m 0.9 pi till yyou trigger thhat is
at 240 deegrees, it is E. Now why
y there is a jump
j
here, ffrom E to strraight away this value? Why
there is a jump? It iss E all of a sudden.
s
It is because jusst prior to 600 degrees, w
what is the ooutput
voltage? It is E. If I neglect th
he turn on time of the devices at 66, just imm
mediately afteer 60
degrees, what is the output
o
voltaage? Same ass the input vvoltage that iis 230 into rroot 2 into siin 60,
b 3.
a sin pi by
So just before
b
60 deg
grees, it is E or 100 voltts. Immediattely after 60 degrees, it iis 230 into rroot 2
into sin 60. So, inpu
ut the outpu
ut voltage in
nstantaneoussly jumps from E to thee instantaneously
value of the input voltage at omeega t is equall to alpha. So, there is a jump. So thhat is why there is
a jump here. Again, I am neglectting the devices, the turnn on time herre. So we neeed to find ouut the
average value
v
of thiss wave form. So, if I kno
ow the averaage value off this wave fform that is equal
to i averaage into R pllus E, KVL has
h to hold good.
g
Isnt itt?
(Refer Sllide Time: 40
0:42)

mega t plus 00.9 pi

So, averrage value is 0 pi by 3 is E d omega t dt plus pi bby 3 to 0.9 ppi Vm sign om
to pi, it is
i E dt. It is found to bee 193.45 vollts. Do not gget bogged oon with thiss value. So, if the
wave form is right, any
a eleventh standard boy can write tthis equationn and he cann solve. I am
m not I
have not in the sensee, I did solvee this, but th
hen it could bbe, it is 93.445. Even if itt is wrong ddo not
say that sir it is wrong, does nott matter. Ass long as youu understandd the subjecct, as long ass you
have draw
wn the wavee form, fare enough,
e
morre than enouugh.
So, averaage value off i now, i aveerage divideed is equal too V averagee, average vaalue of the ooutput
voltage minus
m
E div
vided by R. So, 193.45
5 is the aveerage value of the output voltage, sorry
average value
v
of the output voltage not averrage line. Avverage valuee output volltage, sorry aabout
11

that minu
us 100 divid
ded by 10 is 9.35 amps. What the paart B says? W
What is the new value oof the
average current
c
flow
wing through
h the load if a large L iss connected in series wiith the load?? The
moment the problem
m says that a large value of inductor is connectedd in series, thhe first thingg that
should co
ome to your mind is currrent is contin
nuous.
You do not
n need to mention
m
thatt current is, specifically you do not need to menntion that cuurrent
is continu
uous. You ju
ust you havee to tell that a large induuctor is connnected in series with the load.
First thin
ng that comes to your miind is currentt is continuoous.
If the currrent is conttinuous, life is very simp
ple man. In the sense, I just need too know that vvalue
of alpha, from alpha to omega t is equal to pi
p is output is Vm sin om
mega t and ffrom pi to pii plus
alpha output voltagee is 0 and ag
gain is Vm sin
s omega t.. I am not ggoing to, I aam not draw
wn the
output vo
oltage wave form.
(Refer Slide
S
Time: 43:17)
4

It is samee as the wav

ve form whicch we have drawn for caase 1 or case 2; 0 to alppha, output w
which
is at 0, alpha
a
to pi itt is Vm sin omega
o
t and
d from pi to pi plus alphha, again 0 and from pii plus
alpha to 2 pi it is Vm sin omegaa t. So, average value iss Vm by pi iinto 1 plus ccos alpha, soo 155
s
as averrage value oof the outputt voltage minnus E divideed by
volts. So, average vaalue of I is same
the current value.

12

(Refer Sllide Time: 44

4:06)

I will sollve, I will geet another prroblem; halff controlled bbridge, RE lload, life is bbit difficult w
when
it becom
mes just a RE
E load. Life is very simp
ple while thhe current is continuous.. Trigger anggle is
maintained at alpha is equal to 110 degrees. Current bbecomes 0, 50 degrees beyond the 0
crossing. I will repeaat see, curreent becomes 0 at 50 deggrees beyondd the 0 crosssing. aree load
a applied voltage
v
wav
ve form and also determ
mine the valuue of E. E iss not known, E is
current and
not know
wn. Again saame, where does one trigger the SC
CR or if I aapply gate puulse, from w
where
onwards it will start triggering?
t
m should bee equal to siin inverse E divided by Vm. E is noot known, deefinitely it shhould
Alpha min
by less th
han, which value?
v
less than?
t
E shou
uld be definnitely less thhan 90 degreees. If it is, ssorry,
alpha min
n should deffinitely, shou
uld be less th
han 90 degreees. I cannott have alphaa min greaterr than
90. If thaat alpha min
n is greater th
han 90 it im
mplies that, w
what does it implies? Maximum vallue of
sin inverse, it implies that the vaalue of E is same
s
as the peak value Vin. So, therre is no curreent at
all, you cannot
c
triggeer the thyristtor at all. Thy
yristor is triggger at 110 ddegrees.
The prob
blem says th
hat current becomes
b
0 beyond 50 deegrees. In thhe sense, yoou have triggger at
110, currrent starts flo
owing and it becomes 0 beyond 50 degrees beyyond this pooint. So, at stteady
state or may
m be at th
he end of thee first cycle may
m be, currrent becomees 0 at 230 ddegrees or pii plus
pi by 18 pi, this is th
he load curreent. So, simiilar wave foorm will be tthere. So it sstarts at 11 bby 18
that is at 110 degreess, increases and
a becomess 0 at 18 deggrees.
But how does the ou
utput voltagee wave form
m look like? F
From 110 too 180, 110 too 180 degreees, Vi
is equal to V0 or V0,0 output volltage is sam
me as the inpput voltage.. So in disccontinuous m
mode,
remembeer, in discon
ntinuous mod
de, as long as
a thyristor cconduct, inpput voltage iis equal to ooutput
voltage in
n the positiv
ve half. I meean, it may so happen thhat I told youu that currennt may becom
me, if
the current were to become
b
0 beffore pi, from
m 110 to the instant wherre the currennt becomes 00, the
load voltage is Vi.

13

Fortunately, here current becomes 0 beyond pi. So, the load voltage is V0 is same as Vi from 110
to pi. 110 degrees to pi, it is Vm sin omega t, from pi to 230 degrees or pi plus pi by 18 pi,
freewheeling diodes are conducting now. Output voltage is 0, theoretically. Otherwise you may
have around, may be 3 volts or so 0. At 230 degrees, current lies down, no current is flowing
now. Vo is same as E.
If I would have RL load, if I had RL load, output voltage is also 0. But, here we have RL, we
have a RE load. So, at 230 degrees current becomes 0. From 230 to 290 that is 180 plus 110 or
till you trigger the next pair of thyristor that is sorry next thyristor here T 4, output voltage is
same as E.
So, if you see here, current starts increasing from 11, from 110 degrees because you have
triggered the thyristor. Current starts increasing, becomes 0 at 230 degrees, load current. So,
from alpha to pi, source supplies power T1 T3 because current is flowing at that time. Output
voltage is same as Vi. 0 from pi 2 pi by 18 pi that is 230 degrees, freewheeling diodes are
conducting here, T 2, T1 turns off because we have the common cathode configuration out of
which 1 of them is a diode.
The moment B potential becomes higher than A, D3 starts conducting which turns off T1. That
half waving effect is not there here irrespective of alpha because immediately at in the negative
half omega t is equal to pi plus, D3 starts conducting, this turns off which is not the case if I put,
replace D3 by thyristor there and T4 by a diode. If I interchange these 2 then we may have a half
waving effect, I told you, because T1 continuous to conduct till T3 is triggered. Just to remind
you, nothing to do with this problem, may be. So, this is the wave form.
So, from pi plus pi by 18 pi to pi plus 11 by 18 pi, it is E. Then again Vm sin omega t. So, I need to
find out the average value of this wave form. See, how does the source current look like? Source
current is same as the load current in the powering mode. When is the powering mode? When T1
and D2 are conducting, it is a powering mode because source comes in series with the load and
similarly in the negative half when D3 and D4 are conducting, source comes in series with the
load. So that is the powering mode.
So, source current is same as the load current. At omega t is equal to pi plus, T1 turns off. I have
assumed source inductance to be 0. So, immediately T1 turns off, so there is a step jump. from
immediately, current becomes 0, source current becomes 0. Instantaneous, this is all most
instantaneous because source current is source inductance is 0, immediately turns off but then the
difference between load current, now the load current starts flowing through D1 sorry, D2 and D3.
Same thing happens in the negative half.
So problem is very simple. I need know, I need to find out the average value. It is given by this
equation, 1 over pi from 110 ten.

14

(Refer Sllide Time: 53

3:12)

In the sen
nse, 110 deg
grees to pi, itt is Vm sin om
mega t. From
m pi to 230 ddegrees, freeewheeling peeriod;
d omega t, 0 d omeg
ga by dt. Fro
om the end of
o the freew
wheeling modde to till youu trigger thee next
pair that is pi plus 110 degrees, itt is E d omeg
ga by dt.
So, this is
i 1 complette, this is hallf a cycle. So
o, you substtitute the vallue for Vm iss 230 into rooot 2,
integrated. So, this iss the averagee value. I go
ot an averagee value in terrm of E. Now, I averagee is V
average minus
m
E div
vided by R. So, value off I average iis 1.8 amperres, R is 6 oohms, it is ggiven.
These aree the 2 param
meters are giiven. So, E is 85.9 volts.. Again, a veery educativee problem it is.
All most all the conccept that aree that which I emphasizeed are there.. Again, I diid not blindly use
the outpu
ut voltage, av
verage valuee of the outp
put voltage is Vm by pi 1 plus cos allpha, alpha is 110
degrees, no, I did nott use that. So
o, RE load, discontinuou
d
us, I drew the wave form
m then I founnd out
the averaage value and
d determined
d my currentt.
So remem
mber, very important
i
co
oncept, that expression
e
iis valid onlyy if the curreent is continnuous.
Otherwisse, it is not. The momen
nt you get a problem, thhink for a whhile; spend ssome time w
with a
problem, think for a while, draw
w the wave form. If youu draw the w
wave form, I am tellingg you
90% of th
he problem is
i done. You
u are through
h, I am tellinng you.
Thank yo
ou.

15