00:01

We continue our coverage of TSC-TCR static power generator in this lecture. TSC-TCR static power
generator contains many thyristors switched capacitor units in combination with one thyristor controlled
reactor unit.

00:24

we are discussing the internal controller for this TSC TCR system in this lecture.

00:36

This internal controller, the function of the internal controller is this internal controller will be presented
with a demanded quantity, which is usually the capacitive reactive current that the unit must take from
the bus where it is connected. This demand is

00:56

usually presented as a DC voltage with a certain proportionality between the actual current and the DC
voltage you are using to represent the actual required current. With this input, the internal controller
should ensure that the TSC-TCR unit takes exactly that much of current from the grid. That is the function
of the internal controller.

01:23

TSC-TCR unit internal controller is very much more complicated than the FC-TCR internal controller. In
fact, the TCR internal controller which we studied under FC-TCR is a unit which is needed here also. But it
is just one unit. There are many other

01:50

units required for internal control of TSC-TCR. One main unit required in the internal controller of TSC-
TCR is the switch on switch off controller for a particular TSC unit. We call it the on off controller for a
TSC unit.

02:13
this controller block has to be there for each and every TSE unit involved in the TSE TCR system that
means you need this controller one for each TSE unit this controller will switch on the TSE unit paying
attention to the initial voltage across the capacitor and paying attention to the correct instant of
switching on

02:42

So the function of the first controller we are discussing today is the name of the controller is the on-off
controller for a TSE unit. The function is after receiving an on-off command signal which it receives here.
So the TSE unit whose controller is given here is receiving an on-off signal here.

03:08

the signal is normally low indicating off condition. When that signal goes from low to high, that is a
command for the TSC system to be switched on. And when it goes from high to low, it's a command to
switch off that TSC unit.

03:27

this ON-OFF control signal will be generated from another controller which is a part of our internal
controller of TSC-TCR. As I mentioned already the internal controller of TSC-TCR contains many blocks
one of which is the ON-OFF controller. The input for the ON-OFF controller that is this signal

03:51

will be generated by another controller which is part of the internal controller. We will take it up after we
complete this. In the last lecture, I pointed out a strategy for switching on the TSE unit.

04:10

Of course each TSE unit will be using a current limiting reactor in series which is usually 4% of the
capacitive impedance. But even with this current limiting reactor transient free switching is possible only
if two conditions are met and those two conditions are the switching must take place.

04:36
at the peak of the sine wave, 90 degree position or 270 degree position. And when switch-on signal is
given at that position, the voltage across the capacitor must be exactly equal to n squared by n squared
minus 1 into Vm, where n is the ratio decided by the capacitor and inductor value.

05:03

practically you will not have this correct voltage in the capacitor and therefore practically completely
transient free switching is not possible. If that is not possible what is the next best? The answer to that
question was given in the last lecture. The next best is

05:23

the following. There are two situations to be covered in the next best solution. The two situations are
what is the initial voltage across the capacitor inside the TSC that you are going to switch on. If that
voltage, we had assigned a certain polarity for that voltage.

05:48

This is the capacitor voltage polarity. With this polarity, if that voltage is a positive quantity, then see
whether that positive quantity is more than the peak of the sine wave or less. If the capacitor voltage is
positive and it is more than the amplitude of the input voltage, then the correct switching angle is 90
degree.

06:17

and we must switch on at 90 degree. If the voltage across the capacitor is negative and its magnitude is
more than Vm, then the correct switching position is 270 degree. That is maximum of the negative half
cycle. If the voltage across the capacitor is positive and it is less than the peak value of the sinusoidal
voltage, less than Vm, then the correct switching or optimum switching instant is

06:46

the instant at which the sinusoidal supply voltage becomes equal to the capacitor voltage. But if the
capacitor voltage is less than Vm, then there are two points at which it can become equal to a sine wave.
We choose the first point. Therefore, the angular position of that point will be between 0 to 90 degree.
So we take the position which is between 0 to 90 degree.
07:13

So the switching on will take place at an angle less than 90 degree and more than 0 degree.

07:21

If the voltage across the capacitor is negative and its magnitude is less than Vm, then we will have to
switch on when the supply voltage becomes equal to the capacitor voltage for the first time. That will
happen at an angular position between 180 degree and 270 degree. So it will be before 270, after 180.
We will have to switch on at that point. So these are the four different cases and solutions for our...

07:50

or the algorithm for four different cases. Now this switching on algorithm is implemented in simple
analog and digital hardware in this diagram. I am not claiming that this is how industry is doing it. I am
just proposing a feasible possible controller circuit which can function as per this algorithm. There are
different ways of implementing this control algorithm.

08:19

this is just one of them you can implement it in a microcontroller you can implement it in a computer i
guess that is how it's going to be done in practice but let's see how it can be done using simple analog
and digital electronics so this is the proposed controller block diagram or schematic diagram for
implementing this switching on algorithm for a TSC

08:48

The signal that you have here is the grid voltage. I mean the power system voltage. But being an
electronic circuit, obviously, you can't give that high voltage to this point. So when I write VS, I mean a
voltage proportional to the actual VS. You will, of course, use potential transformer, step down, and then
use a potential divider to step down and bring it to electronic level.

09:13

So what I mean is that this signal is a signal proportional to the supply voltage. It is a representative of
the supply voltage. This signal is the capacitor voltage. Again the same comments. Capacitor is holding
kilo volts. You can't directly give that capacitor voltage to electronics. So some voltage sensing unit and
stepping down unit is used to reduce the capacitor voltage to a scaled down quantity which is
proportional to the actual capacitor voltage and that is given here.

09:43

The scaling factors are in.

09:48

the supply voltage is integrated with a negative sign, negative integrator. That negative is not really
needed, you can do this with positive integration, but then when you make an integrator using an op-
amp, it usually produces a negative sign. So I assume that it is a negative integrator. So if this is Vm sine
omega t, then this voltage waveform will be minus Vm cos omega t, (because of integration turns sine to
cos).

10:18

V m cos omega t and I will describe the operation briefly. Please work out the details and draw the
waveforms and.

10:33

convince yourself. Now this cosine waveform is passed through a zero crossing detector which produces
a high during this half cycle. So the rising edge of this waveform will coincide with the first peak of the
sine wave here and the falling edge will coincide with the negative maximum peak. So you will have a
rising edge corresponding to this point, 90

11:03

and falling edge which will correspond to 270 degree.

11:09

There are two mono-stable multivibrators here which will produce a very narrow pulse output, maybe
around 5 microseconds or something. But this is a positive edge triggered one and this is negative edge
triggered one. Therefore, this mono-stable produces a narrow pulse at this edge, that is at 90 degree
position. And this mono-stable produces a pulse at 270 degree position.

11:37

you want to do switching we are actually this upper portion of the circuit handles the 90 degree
switching condition. So you want to do the switching at 90 degree or 270 degree depending upon the
polarity of the capacitor voltage.

11:55

Now the capacitor voltage is given to a zero crossing detector again and the output of that zero crossing
detector is ANDed with the output of a zero crossing detector working here. So this signal here is going
to be high during positive half cycle. If the second signal is also high then this AND gate will produce
output. That will happen if Vc is positive.

12:23

So if VCE is positive, this AND gate will produce output. If VCE is negative, this AND gate will produce
output during the negative half cycle of the sine wave because of bubbling here, right? So because of
bubbling here or inversion here, this output will happen only during the negative half cycle and that too
only if VCE is a negative number. So this goes high.

12:53

during positive half cycle if Vc is greater than zero. And this goes high during negative half cycle if Vc is
less than zero. So if this condition is satisfied, this AND gate is enabled and this small pulse, short
duration pulse is passed on to the OR gate.

13:15

If this condition is satisfied, then this short duration pulse which is located at 270 degree position, that is
negative maximum position, that will be passed on to the OR gate. So, one of them will come out here.
So, short duration pulse at 90 degree or 270 degree position depending upon this polarity of the
capacitor voltage.
13:41

Now that short duration pulse will be passed on to this OR gate only if this AND gate is enabled by an
ON-OFF control. So when this unit receives an ON-OFF control, ON means this has to go high. So this line
will go high and only then this AND gate will pass on this trigger pulse to rest of the circuitry.

14:10

this will be disabled, it will be zero here. Okay. And similarly, this AND gate also will be producing zero
only. So this OR gate will have only zero zero condition till this ON-OFF signal is asserted. And therefore,
zero zero condition, zero in the set. So this flip-flop will not be set, it will be in the reset condition and T1
and T2 gate signal will be in the zero condition.

14:38

as long as this line is held low. When this line goes high, indicating that you have received an ON
command, both this AND gate and this AND gate are enabled. This AND gate enabling means a firing
pulse will come either at 90 degree or at 270 degree. Please note that we didn't check whether the
magnitude of capacitor voltage is more than VM or less than VM. If it is more than VM,

15:07

this path is correct. Firing has to be at 90 or 270. If it is less than VM, then firing has to be before 90 or
between 180 and 270. So that firing will take place earlier through this ORGATE channel. The firing has
already taken place, then no harm is done, even if firing comes from this path. That is why I am not
checking the magnitude of VC.

15:37

I am not checking whether the capacitor is holding a voltage more than Vm or less than Vm. If it is more
than Vm, this is the path that should work. If it is less than Vm, this is the path which should work. So if it
is less than Vm, this path will work first, even before this path works. So that's okay. That is, after you
have set the flip-flop, if you give one more set signal, flip-flop is not going to complain. It's already set.
That is why we don't have to check the...

16:07
magnitude. Right. Now coming to the second channel. In the second channel the supply voltage signal VS
and the capacitor signal of course I am assuming that they have been converted to low voltage levels
using the same scale factor so that the comparison will be meaningful. Now these two voltages are
compared in a comparator. So this comparator

16:37

whenever the supply voltage goes above the capacitor voltage.

16:43

Let's assume that capacitor voltage is positive. Okay, and it is less than the peak Vm. In that case, this
comparator will go high somewhere before 90 degree position and then it will go low between 90 and
180. Once more, I'm assuming that Vc is a positive number and it is less than Vm. So it will intersect the
sine wave in the positive half cycle two times.

17:13

So this comparator output will go high first and then go low. When it goes high, this monostable is
triggered. A short pulse and that pulse position will be before 90 degree position if Vc is greater than
zero. If Vc is greater than zero, maybe I should clarify there.

17:36

VC is greater than zero and

17:47

and VC is less than VM.

17:51

So me lolly.

17:55
This channel will produce a short pulse before 270 but after 180 if Vc is less than 0 and mod of Vc is less
than Vm.

18:10

Otherwise these two channels will not produce anything. The monostables will never be triggered. What
do I mean by otherwise? I mean if the magnitude of VC is more than VM, these monostables will never
be triggered. So no harm done by that path. Okay, right. So if the capacitive voltage is less in magnitude
compared to VM, then this channel will work. If the...

18:40

Capacitor voltage is positive then this channel will work and if it is negative then this channel will work
and produce a short duration pulse as a result of mono stable triggering so Now our gate will combine
them and present the short duration pulse to this AND gate Which will be enabled only if ON signal is
high If ON signal is high that pulse will be passed on to this OR gate

19:10

and this OR gate will pass it on to the set of the flip-flop. So the flip-flop will be set and the flip-flop
output is used for further processing before it goes to the T1 and T2 gate. Obviously a flip-flop output
cannot drive a thyristor so you need isolation and driver circuitry and all that. This cannot directly be
given to thyristor but this can be given to the thyristor through isolation

19:41

In fact, we are not even looking at that part of electronics, but that is also a huge part of electronics
needed to fire the thyristors. So remember each thyristor is a combination of many thyristors in series
and many of them in parallel. So all those thyristors will have to be triggered using individual gate
isolation circuits and gate driver circuits, but this is the basic signal used for driving them.

20:08

But then if you are using the short duration pulses coming from the monostables for setting the flip flop
then who is going to reset it? Why would you want to reset it? You would want to reset this flip flop
whenever you want to withdraw the gate of T1 and T2 and why do you want to do that? Because you
want to switch off the TSC. So when you want to switch off the TSC you can do that by resetting the flip
flop.

20:38

and you must ensure that after you reset the flip-flop it won't be set again unless somebody is asking you
to switch on the TSC so it will be a good idea to use this on-off signal itself to do the resetting so when
the on signal which is high goes to low condition then it is off so high to low transition so you go for a
high to low transition monostable that monostable will produce a short duration pulse

21:07

which will reset the flip-pop. And since the signal has gone low, that low will disable the sand gate and
disable the sand gate. So even if these gates are working, their output will not be passed on to the R
gate, and so it will not be passed on to the SET. So you have done a reset, and you're not issuing SET. You
will issue SET only when the next ON signal goes high.

21:34

after sometime maybe you want to switch on the TSE again so you will assert this line you will take it
from low to high and that is when setting action will take place. So this is how the on off controller of a
TSE unit can be implemented to bring the algorithm that we discussed into implementation. I repeat the
algorithm. The capacitor voltage magnitude and polarity should be checked.

22:06

If the magnitude of capacitive voltage is more than the amplitude of the grid voltage, then the right
switching point is 90 degree or 270 degree. It is 90 degree if the capacitive voltage is positive with the
polarity that we have shown in the diagram. And it is 270 degree if the capacitive voltage is negative with
the polarity shown in the diagram.

22:30

If the capacitor voltage magnitude is less than Vm, then the correct switching point is the point at which
the input voltage becomes equal to capacitor voltage. There are two such points. The first point should
be taken. This will result in switching on at an angle between 0 to 90 if the capacitor voltage is positive
and less than Vm. It will also result in switching on at
22:58

an angle between 180 to 270 if the capacitor voltage is less, magnitude is less than Vm and polarity is
negative. This is the algorithm that gets implemented by this analog and digital circuit. Of course it is
possible to implement this algorithm in the form of a microcontroller code or it can be implemented in a
PLC system, it can be implemented in FPGA system.

23:27

it can also be implemented in pure coding in a computer.

23:33

Okay.

23:36

Next, till now we were discussing individual TSC units only. Now let's go to the TSC-TCR system.

23:52

The single phase TSC-TCR system is shown here. A three phase unit is constructed by putting such single
phase units in delta R star. Remember in the case of FC-TCR, we always preferred delta connection for
TCR in view of certain harmonic advantages we could derive from such a delta connection.

24:17

Of course we can derive the same harmonic advantage because there is a TCR here also. So if you put it
in delta, maybe you can derive the harmonic advantages in the form of all triplen harmonics will
disappear in the line etc. So delta is used. But sometimes for convenience it is also used in star.

24:40

Okay. Let's assume that mostly it is delta, but nothing preventing it from using in star. Okay, so the single-
phase TSC-TCR system. Here there's a point that we have to understand clearly. The point is.
24:57

The rating of the TCR unit has to be more than that of the individual TSC unit.

25:08

From the first look of it, we get a feeling that the TCR unit needs to be rated only equally corresponding
to TSE unit. For example, consider a situation of four TSE units, each having Q by 4, where Q is the
maximum desired capacity of the units. So the total desired capacity Q is 4.

25:33

per unit of reactive power, capacity of reactive power is distributed in four units, each one carrying Q by
4.

25:44

Now suppose you switch on the first unit and keep the other three units off, then you get a fixed amount
of reactive power which is Q by 4, capacitive reactive power. Maybe you don't want that. Maybe you
want something between Surya to Q by 4.

26:03

Then you have to cancel part of this by running the TCR.

26:10

You see at each position a TSC-TCR system becomes equivalent to FCTCR. So if only one unit is switched
on, the other three are kept off, then it is a FCTCR system now. So the control logic is the same. So you
can vary the effective reactive power demanded by the system from 0 to Q by 4 capacitive by varying the
TCR demand between Q by 4 to 0.

26:42

If TCR is gated completely and it is taking Q by 4 lagging, then that Q by 4 lagging and Q by 4 leading will
cancel and you will get net zero. And if that CCR is taking zero, then you get Q by 4. So to vary the net
reactive power taken by one TSC unit and one TCR unit.

27:06

you need to have a TCR unit which is exactly equivalent rating to the TSC unit.

27:14

Now suppose you switch on two units.

27:18

they will take Q by 2 and the TCR unit can consume anywhere from 0 to Q by 4. Therefore, the overall
combination can now consume anything between Q by 4 capacitive to Q by 2 capacitive and similarly
when the third unit is switched on and fourth unit is switched off.

27:40

So the coarse control is done by switching on the TSEs and switching them off. And the fine control is
done by alpha control of TCR. And we see no reason why TCR should have a rating more than Q by 4. We
can understand why it has to be a minimum of Q by 4. No harm if it is more than Q by 4. But we don't
see why it should be more than Q by 4. Of course there is no harm if it is more than Q by 4.

28:11

In that case you won't be making use of the full controlled range of TCR. I mean you won't be changing
the alpha from 90 degree to 180 degree if the rating of the TCR is more than Q by 4.

28:27

So no harm done. But then why should it be more? Compulsorily it has to be more. Why?

28:36
Mandatorily, it has to be more than Q by 4 or generally more than Q by N, where N is a number of
identical TSE units. Well, the problem is, again consider the same example. Let's say there are four TSE
units, each rated for 0.25 p.

29:05

each rated for 0.25 pu into 4 and say we have a 0.25 pu TCR. This is the example we are discussing.

29:21

Not suppose.

29:25

the internal controller gets a command to produce point two four.

29:31

What will it do? It will switch on one unit.

29:39

may be the first unit of TSC. And then it will work TCR so that TCR will consume 0.001.

29:52

So 0.25 unit capacity minus 0.001 of GCR will produce 0.249 capacity. Now while it is at 0.249, maybe
somebody switched on an induction motor, or somebody switched off a halogen lamp, or somebody
switched on a CFL lamp somewhere in the power system. And as a result, the reactive power
compensation needed.

30:21

The setting changed.

control systems worked and the setting for this particular unit was changed from 0.249 to 0.251. You see
it's a change by 0.002 only.

30:38

But then with this change, the internal controller of our TSC-TCR system will see that it is more than
0.25, so it will switch on the second unit also.

30:50

because if you want to generate more than 0.25 capacity then you need one more unit in service so
second unit is switched on but if the second unit is switched on the total reactive power taken by the
two units will be 0.5 pu you need only 0.251 therefore the TCR has to now work with 0.249

31:15

So, due to the small change in the setting, what happened was a second unit immediately went on, I
mean second TSC unit, and the TCR firing angle changed from nearly 180 degree to nearly 90 degree,
full.

31:39

Okay, that is this is full firing and this is negligible firing. Here the TCR is not conducting almost virtually
non conducting here it is fully conducting it is full capacity. So TCR alpha will suddenly swing from one
extreme to the other extreme and the second unit will immediately go on.

32:00

Okay, but then right at that moment, maybe somebody switched off a mode.

32:09

because this kind of random events keeps happening in a power system maybe something happened
somewhere in the load and then the controller which was responsible for producing a setting for your
internal controller that controller changed the setting back to let's say point two four nine two

32:36

it doesn't go back to 0.249 maybe it goes back to 0.2492 or 95 again problem now the first unit second
unit has to be switched off okay because the TCR has only 0.25 capacity and the two units together are
producing 0.5 so even if the TCR produces 0.25 the net will be 0.25 capacity but you want only this much

33:03

it is less than 0.25 so there is no other way you have to switch off the second unit so you switch off the
second unit and then TCR has to run from 0.249 to 0.0008 to cancel the remaining once more it has to
change the firing angle rapidly from one position to the other extreme position and maybe within a few
milliseconds

33:31

This 0.2492 becomes 0.2501. So this way, the setting coming from the controller, I mean the outside
world external controller, or I mean the setting received by the internal controller of our TSC-TCR system
is not going to be a rock hard constant, mathematical constant setting. Even if it remains more or less
constant, it has a statistical variation or random component in the setting.

33:59

may be it is around 0.25 but it has some random noise in it. This noise is created by multitude of random
events happening in the power system. So you can never count on a rock hard setting. The setting
changes in milliseconds in a random manner.

34:22

And if the setting happens to be anywhere around 0.25 and then it has a random component, then you
have the problem of second unit frequently coming into service and going out of service in a repeated
manner. On other words, it becomes messy, it chatters. This is the so-called famous chattering problem,
relay chattering problem, right? So the second unit chatters. It comes on, goes off, comes on, goes off.
And correspondingly the TC also chatters.
34:51

its alpha changes from near 0 to or near 180 to near 90 and again 90 to 180. This kind of flipping takes
place. In fact, that kind of rapid flipping of the firing angle from one extreme to the other will result in lot
of noise emanating from the inductor. So this kind of chattering behavior is to be prevented.

35:18

Same chattering can take place when the setting comes close to 0.5. Again it can take place when the
setting comes close to 0.75. And again it can take place when the setting is close to 0. Of course, if the
setting is somewhere in between, for example, 0.17, chattering may not be there because whatever
random noisy changes in that 0.17 comes, maybe it won't take it to 0.25 or 0.

35:48

So you must have an idea about the noise content in the setting or the range of variation of the setting
due to random events in the power system. If you have such a notion, such an idea, then you can solve
this problem. It's a famous problem and the solution is also famous. Use hysteresis switching. Isn't it? So
using hysteresis in a switcher is the classic way to suppress chattering.

36:18

or rattling. This is the answer to the question, why we need TCR to be more than, rating of TCR has to be
more than that of individual TSE unit. This is because we want to implement hysteresis. Okay, what is the
meaning of that? When the demand goes slightly above 0.25, we know we have to switch on the second
unit.

36:44

So when the demand is 0.25 per unit, we will switch on the second unit. Slightly above 0.25, we switch
on. Of course, first unit is already on. We will switch on the second unit. Now suppose the demand after
going above 0.25, suppose it starts coming down. It goes in the opposite direction. Do we switch off the
second unit when the demand is 0.249999, et cetera? No, we don't. We will switch it off.

37:14
after it crosses 0.25 and only after it crosses a value below that, for example, maybe 0.2. So we will
switch on the unit number 2 when the demand goes to 0.25 in the increasing direction. When the
demand is decreasing, we will wait till the demand comes to 0.2, only then we will switch off the unit.
Okay.

37:41

If you switch off the second unit only at point two, that means with the demand of point two, slightly
more than point two, we have two units working. Two units together will produce point five. But right
now we need only slightly more than point two. By slightly more because the second unit is still in
service. When it touches point two, second unit will go out of service. So I would say point two zero one
or something.

38:10

So approximately 0.2 is what you need, but you are producing 0.5. So the TCR should cancel 0.3, isn't it?
So TCR should have a rating 0.3, at least 0.3, if not more than that. Why? Because you put a 0.5
clearance, sorry, 0.05 clearance. So you switched on the second unit when the demand crossed 0.25 in
the increasing direction, but you switched off the TCR, TCS unit, second TCS unit.

38:38

only after the demand crosses point two in the opposite direction.

38:45

So since you are waiting some more to switch it off, the TCR has to be capable of handling the
unnecessary capacity power that the two units are producing. So obviously the TCR rating has to be
rating of TSC unit plus the clearance you are providing, which as you can appreciate now, which is the so
called hysteresis bandwidth. So whatever hysteresis bandwidth you are building into the switching
process.

39:15

that also has to be added to the rating of TSE unit and that number will be the TCR unit rating minimum.
It can be more than that, but it has to be at least that much. So this is the reason why TCR rating has to
be more than two by N. Now the switching strategy, the switching events that will happen when
39:41

The capacitive reactive power demand on the TSC-TCR system is on the rise. That is, you are moving
from left to right. In my diagram, I am taking capacitive reactive power demanded as a positive quantity
and I am representing it on the right half. Some test books do it the other way.

40:07

Okay, so what is on the x-axis is the capacitive reactive power that the system is demanding from or the
internal controller is demanding. Okay, this is the origin zero. This is Q by 4, this is Q by 2, 3Q by 4 and Q,
where Q is the maximum rating of the unit. So, if you are demanding anything which is slightly above
zero, immediately the first unit will go on, the blue line.

40:37

So it goes on and when it goes on it produces Q by 4. The site is Q by 4. But then what you need is only
slightly more than zero. Since it is producing Q by 4 which is not needed that has to be almost fully
cancelled which will be clear in the next diagram. This diagram gives you the TCR reactive power. So as
you increase your demand on the TSE TCR unit.

41:05

what happens to the units individually and what happens to the TCR reactive power. The two graphs will
show you that. The blue traces for the events happening when you are increasing the demand from a
negative number to positive number, you are moving in this direction. So when you cross zero, the first
unit will come on. When you cross Q by four, second unit comes on.

41:34

third unit comes on when you cross three Q by four fourth unit gets switched on when you cross Q by
four second unit comes on the total delivered is Q by two but you need only slightly more than Q by four
so the TCR comes in and Cancels almost completely the reactive power taken by a unit Okay, and then

42:01

the TCR adjusts its reactive power as you increase your demand like this. Once more your demand
crossed Q by 4 just crossed Q by 4 immediately the second unit came on and then the two units together
started producing Q by 2. I should have written here also this is Q by 4 this is Q by 2 3 Q by 4.

42:28

So Q by two total is being generated, but you need slightly more than Q by four only. You are here. So
TCR comes in. TCR produces nearly Q by four and cancels it, cancels the second one. Then as you
increase your demand still further, the second one doesn't have to be canceled fully, so TCR goes back.
And finally, when you are at this edge, the TCR has to produce only zero, because you don't want any
cancellation.

42:58

and then you go above Q by 2, immediately the third one comes on, like that. Now after reaching the full
level, suppose you are going back, I mean you are reducing your demand. When you reduce your
demand, you will switch off the fourth unit, not at 3Q by 4, rather at 3Q by 4 minus QH hysteresis
quantity. I am using the green curve. The arrow tells you the direction in which you are moving. So when
you are reducing the demand...

43:28

you will switch off the fourth unit only if the demand comes below 3Q by 4 not at 3Q by 4 below by how
much by QH similarly at other points so actually I am showing two lines the blue and green they are
collinear actually it's only for visual appreciation that I am separating and showing it there okay so this y-
axis should give you the reactive power

43:57

delivered by the TSE units, right? So, whether you are going this way or this way, the reactive power is Q
only, okay? So, I'm separating the two lines slightly to show clearly. Similarly here also. Okay, and finally,
when the reactive power demanded is zero, no, the first unit is not switched off. It is switched off only
when the reactive power goes below minus QH.

44:25

That means only if you demand inductive QH, only then the first unit is switched off.
44:35

so the two corresponding paths forward path and reverse path for the TCR reactive pore is also shown
here and from this should be clear that required minimum TCR rating is this gap not this gap the second
gap and that is q by 4 plus qH now the amount of qH has to be decided by

45:05

noise that the TSC-TCR system setting can suffer. That depends on where you are installing this unit, how
rapid the load variations are at that bus, et cetera. An experimental study or a simulation study will be
needed to decide the value of QH. Higher the value of QH, okay, better the avoidance of chattering. But
higher the value of QH, higher has to be the rating of TCR.

45:38

Now let's see how this...

45:43

the algorithm which is described in these two graphs can be implemented in hardware. That unit I will
call it the on off signal generator. Here in the first controller unit we are talking about the different parts
of internal controller. The first part was the on off controller which really does the switching on of the
TSE unit and switching off the TSE unit.

46:13

it does the switching on and switching off only by receiving an on-off signal this on-off signal has to come
by obeying this logic so this logic has to be implemented in a circuit block or system block output of that
system block will become input here I am talking about that block now, the on-off signal generator for
TSE unit this is a single unit

46:43

all the TSEs are handled in a single unit. This will produce the on-off signal for all the four units in our
example and distribute the four signals to four separate units of this kind. Okay. Now, I am assuming that
the input to this unit will be the demanded or
47:12

But this current will be represented in the form of a DC voltage as it is usual in control system design. So
we will use a DC voltage to represent that demanded capacitive current. But there has to be a scale
factor. What is the scale factor? Well, the scale factor is the IC reference that you can get when all the
TSC units are on and the system voltage is nominal, I mean rated. Okay.

47:41

So nominal voltage or the rated voltage, all the units are on, what is the maximum capacitive current we
can get? That value is represented by a standard voltage. Let's say VR, reference voltage.

47:57

maybe you can take it as 5 volt. So therefore, full capacity at nominal voltage, rated voltage is
represented by a DC voltage VR. So if you want half the rating, then you will give a voltage of VR by two.
That is the way we are presenting the command to this unit.

48:20

then the grid voltage.

48:25

the phase voltage if it is a star connected unit, the line voltage if it is a delta connected unit. That voltage
is passed through a potential transformer, then it is passed through a potential divider for stepping down
further, then it is given to a full bridge rectifier and then a capacitor filter is used.

48:47

thereby you produce a DC voltage whose magnitude is proportional to the amplitude of grid voltage and
that proportionality constant is alpha who decides alpha? potential transformer ratio, potential divider
ratio and that's it those ratios will decide it now the ratios are such that the quantity alpha is such that
when Vm is nominal value
49:15

Suppose it is a 110 kV bus, nominal value is 110 kV, Vm is 110 into rho 2. So when system is at nominal
condition, the number alpha Vm nominal will be Vr. That is how alpha is set. So we produce a low voltage
DC from the grid using this technique of potential transformer, potential divider, full bridge rectifier and
capacitor filter.

49:46

with the value of alpha adjusted such that under nominal conditions that DC voltage is exactly VR, which
is incidentally the voltage we use to represent the maximum capacitive current we can get from the TSE
units under nominal conditions.

50:05

Now the first block in our controllers we are going to use.

50:11

This is the rail to rail.

50:17

CMOS operational amplifier I have explained this kind of amplifiers in a previous lecture Rail to rail the
meaning of that etc. The function of this is we just want to buffer this voltage and in addition Okay, we
want to saturate the output at alpha Vm The reason for this was explained clearly when I discussed
internal control of TCR the same reason

50:47

It's a problem with system voltage changing. Essentially, all these units are variable impedance type. So
unless the voltage is correct at nominal level, you won't get the rated capacity from them. This point was
explained, and the need for this kind of a saturation was explained in TCR internal controller lecture.

51:15
If you forgot, please listen to that lecture. Please understand why we are saturating the output of this at
alpha Vm. Okay. So right now we may take it that the received references simply buffered in the unity
gain amplifier and passed on to a set of regenerative comparators. Regenerative comparators are
electronic circuits which will produce hysteresis in their switching behavior.

51:44

You must have studied this in the undergraduate electronic courses and you must have done in the lab
also. If not you have to study by yourself. I don't want to spend time on teaching you how to design a
regenerative comparator using LM741 or LM311IC. But you should know. Okay. Now this regenerative
comparator should have this kind of hysteresis curve. The first one should have, it should go high at zero.

52:14

but it should go low at a small negative number. Second one should go high at Q by four. It should go low
at Q by four minus QH, like that, okay? So four of them are arranged. These are op-amp or comparator IC
circuits, properly designed, okay? All of them are given this reference value, and the reference value is
compared with the setting. The first...

52:45

First comparison is with 0 itself. So if the reference value is more than 0, this comparator will go high,
indicating that the first unit has to be switched on. If the reference value here is more than alpha Vm by
4, one-fourth, of course alpha Vm under nominal conditions will be equal to Vr. So it will be Vr by 4. Vr is
full capacity, so Vr by 4 is one-fourth capacity. Okay.

53:17

So it is compared with that number and when the setting goes above that, the second unit comes on like
that. So the logic I explained here is properly implemented with the help of regenerative comparators
designed using analog electronic building blocks. The reference voltages for those comparators are
produced in a resistor chain. The resistor R is in kilo ohms range.

53:46

the supply voltage is alpha Vm which will be numerically equal to Vr if the system voltage is nominal
otherwise it will get adjusted. It will get adjusted such that the problem with variable impedance type
54:02

reactive power compensator refusing to give you the correct value because of voltage change that
problem is solved by making this alpha Vm instead of Vr which is a fixed number okay so here when the
grid voltage changes this supply voltage also will change this reference will change and supply voltage of
this op-amp will change solving the problem of variability in the reactive power taken by a capacitor and
inductor

54:31

to a certain extent. Detailed explanation was given earlier in the context of internal control of TCR.
Please listen to that again if you are not clear about it. Okay. So these four lines will give you the on-off
control signals for the four units. I have labeled the signals here. This is A and this is B, C, D, E. These five
signals will go to the next controller circuit. That is why I am labeling them.

55:01

So this will go to TCR reference generator unit. These four also will go there. In addition to going for on-
off control. Where will they go? They have to go here. So they have to go to the first unit. There are four
such units. BCDE will come to the four units. In addition to that, those lines will also go to the next block.

55:30

And the next block is the TCR reference generator unit. The TCR unit has to be now properly controlled,
right? So you have decided how many units have to go on. That is all you have decided till now. But you
know that those units will produce fixed amount of reactive power, leading reactive power. You may not
want that much. You may want to cancel part of it. That is why TCR is used. So how much should be
canceled? Well, the answer is...

55:59

What are you asking for reference? And what is the total you're getting from the on units? Are they
equal? If not, find the difference. That difference must be given as a reference for TCR. That's what we
are doing here. So from A, that means from the demand. So the demanded quantity is given here. And
one fourth,
56:28

capacity alpha V m by 4 which will be V r by 4 under nominal conditions that is given as a constant here.

56:38

These are CMOS switches, electronic switches, which can be controlled by control signals. Those control
signals are actually these signals, BCDE. So when the first unit has to go on, line B will go high and that
will close this switch also. When the second unit goes on, this switch also will get closed. When the third
one goes on, this will go, get closed. So when the units get switched on...

57:06

corresponding switches here also will get switched on. Suppose only first unit is on. So these three
switches are open, this unit is on. So this is closed. So alpha Vm by four is getting multiplied by a gain of
R by R, which is one. So it becomes minus alpha Vm by four here. Minus, and here it is plus.

57:35

So these two together summed up. So you're taking difference actually and that difference is multiplied
by four. That multiplication is not absolute must but I wanted to give VR volts when IL reference

57:55

is maximum. So I want to use the same scale factor for the TCR control also. So IL reference will be
represented by VR volts for full capacity of TCR under VN nominal condition. That is where I am putting
4. Otherwise you will have to adjust that scale factor in the TCR internal control unit.

58:24

So it's basically a summing unit. The difference between the capacitor delivered reactive power and the
reactive power you wanted to take, these two are different. And that difference is given as the IL
reference with a certain scale factor. That's the gist of it. Now, this reference, IL reference, will go to the
internal controller of TCR. I am not discussing it again because I've spent a lecture on that earlier.

58:54
we have studied TCR internal controller. So that TCR internal controller also will be needed here. That
will also be part of internal controller of TSE TCR. So in short, which are the units in TSE TCR internal
controller? The answer is...

59:15

the first unit will be this the on-off signal generator that is a block and from that block signals will go to
the next block TCR reference generator unit and the output of this reference generator unit will go to
TCR internal controller unit which we have learned earlier that's another big block and these signals will
go to this on-off

59:44

each individual TSE units. So the ON-OFF controller will be on a per TSE basis. For each TSE unit you need
one ON-OFF controller.

59:59

and this one is a common unit which will handle all the TSE units and this one is a single unit because
there's only a single TCR and the last one which is not given in this lecture which is TCR internal
controller that's also only one so these blocks put together will be the internal controller for TSE TCR
system with that I wind up the

01:00:29

coverage on TSC, TCR. There are some more points like the loss behavior, then the similarly some other
points are there. Please read the textbook. I don't want to spend time on those things. Please read and
study it by yourself. I strongly recommend that you read the textbook.

01:00:55

these two, section 5.2 from Understanding Facts by Engrani and more importantly section 3.8 from Dis-
Excellent Book by R. Mohan and Rajiv Verma. So I am leaving out some portions in this TSC-TCR. Please
cover the portions by yourself by reading these two sections thoroughly. In the next lecture we will be
taking up another new topic.