WEBVTT

00:00:02.990 --> 00:00:04.850
GILBERT STRANG: This
video is about one

00:00:04.850 --> 00:00:07.610
of the key applications
of ordinary differential

00:00:07.610 --> 00:00:14.120
equations to electrical flow,
flow of currents in a network.

00:00:14.120 --> 00:00:18.000
And so I drew a network,
a very simple network.

00:00:18.000 --> 00:00:21.930
It's just called an RLC loop.

00:00:21.930 --> 00:00:24.876
It's only got one loop, so
it's a really simple network.

00:00:26.530 --> 00:00:31.550
The R stands for
resistance to the flow.

00:00:31.550 --> 00:00:34.680
The L stands for an inductance.

00:00:34.680 --> 00:00:36.010
And the C is the.

00:00:36.010 --> 00:00:39.000
Capacitance those are
the three elements

00:00:39.000 --> 00:00:47.660
of a simple linear constant
coefficient problem associated

00:00:47.660 --> 00:00:49.240
with one loop.

00:00:49.240 --> 00:00:52.770
And then there is a
switch, which I'll close,

00:00:52.770 --> 00:00:55.180
and the flow will begin.

00:00:55.180 --> 00:00:59.420
And there is a voltage
source, so like a battery,

00:00:59.420 --> 00:01:03.940
or maybe let's make this
alternating current.

00:01:03.940 --> 00:01:08.470
So the voltage source
will be some voltage

00:01:08.470 --> 00:01:13.750
times an e to the i omega t.

00:01:13.750 --> 00:01:17.300
So we're going to have
alternating current.

00:01:17.300 --> 00:01:22.810
And the question is,
what is the current?

00:01:22.810 --> 00:01:25.590
We have to find the
current, I. So the current

00:01:25.590 --> 00:01:34.575
is I of t going around the loop.

00:01:38.040 --> 00:01:40.240
And we saw our
differential equation

00:01:40.240 --> 00:01:44.870
will have that unknown I of
t, rather than my usual y.

00:01:44.870 --> 00:01:47.400
I'm going to use I for current.

00:01:47.400 --> 00:01:56.220
Again, this is an RLC loop that
everybody has to understand,

00:01:56.220 --> 00:02:03.510
as in electrical engineering.

00:02:03.510 --> 00:02:08.699
So I'm going to have a second
order differential equation.

00:02:08.699 --> 00:02:11.300
Well, you'll see what
that equation is.

00:02:11.300 --> 00:02:14.760
So you'll remember Ohm's law.

00:02:14.760 --> 00:02:19.620
That the voltage is the
current times the resistance.

00:02:19.620 --> 00:02:28.560
So this gives me a voltage
across the resistor.

00:02:28.560 --> 00:02:31.920
If the current is I and
the resistance is R,

00:02:31.920 --> 00:02:39.200
then the voltage drop from
here to here is I times R.

00:02:39.200 --> 00:02:40.180
So that's that term.

00:02:42.940 --> 00:02:49.170
But now I have also my
current is changing with time.

00:02:49.170 --> 00:02:50.670
This is alternating current.

00:02:50.670 --> 00:02:52.360
It's going up and down.

00:02:52.360 --> 00:03:00.600
So the current is also going
through the inductance.

00:03:00.600 --> 00:03:05.560
And there, the voltage
drop across the inductance

00:03:05.560 --> 00:03:07.090
has this form.

00:03:07.090 --> 00:03:10.000
The derivative of the
current comes into it.

00:03:10.000 --> 00:03:13.830
And in the capacitance,
which is building up charge,

00:03:13.830 --> 00:03:17.170
the integral of the
current comes in.

00:03:17.170 --> 00:03:21.630
So there, that's the
physical equation that

00:03:21.630 --> 00:03:25.710
expresses this
voltage law, which

00:03:25.710 --> 00:03:31.980
says that around a closed
loop-- this is a closed,

00:03:31.980 --> 00:03:36.900
loops are closed-- add to 0.

00:03:36.900 --> 00:03:42.720
So I have four terms, and
they combine to give us 0.

00:03:42.720 --> 00:03:46.040
So there's an equation
I'd like to solve.

00:03:46.040 --> 00:03:49.520
And how am I going to
solve that equation?

00:03:49.520 --> 00:03:54.830
By the standard
idea which applies

00:03:54.830 --> 00:03:57.430
when I have constant
coefficients

00:03:57.430 --> 00:04:03.680
and I have a pure
exponential forcing term.

00:04:03.680 --> 00:04:09.240
I look for a solution that is
a multiple of that exponential,

00:04:09.240 --> 00:04:09.946
right?

00:04:09.946 --> 00:04:12.540
The solution to
differential equations

00:04:12.540 --> 00:04:14.970
with constant
coefficients, if they

00:04:14.970 --> 00:04:17.480
have an exponential
forcing, then

00:04:17.480 --> 00:04:25.040
the solution is I
equals some, shall

00:04:25.040 --> 00:04:30.520
I say W e to the i omega t.

00:04:30.520 --> 00:04:35.860
Some multiple of the source
gives me the solution

00:04:35.860 --> 00:04:37.520
to that differential equation.

00:04:37.520 --> 00:04:41.280
Well, it's actually a
differential integral equation.

00:04:41.280 --> 00:04:45.140
I can make it a more familiar
looking differential equation

00:04:45.140 --> 00:04:47.920
by taking the derivative
of every term.

00:04:47.920 --> 00:04:49.400
Suppose I do that.

00:04:49.400 --> 00:04:52.050
Suppose I take the derivative
of every term, just to make

00:04:52.050 --> 00:04:54.190
it look really familiar.

00:04:54.190 --> 00:04:58.900
That would be L
times I double prime.

00:04:58.900 --> 00:05:01.280
Taking the derivative
of the derivative.

00:05:01.280 --> 00:05:05.560
This would be RI prime.

00:05:05.560 --> 00:05:08.860
The derivative of the integral
would be just I itself.

00:05:08.860 --> 00:05:13.010
So I'd have 1 over
C I. And I would

00:05:13.010 --> 00:05:19.140
have the derivative here, I
omega V e to the I omega t.

00:05:22.160 --> 00:05:27.110
So it's just a
standard second order

00:05:27.110 --> 00:05:30.410
constant coefficient linear
differential equation.

00:05:30.410 --> 00:05:34.470
And in fact, if you are
a mechanical engineer,

00:05:34.470 --> 00:05:36.570
you would look at
that and say, well, I

00:05:36.570 --> 00:05:39.760
don't know what L, R,
and 1 over C stand for.

00:05:39.760 --> 00:05:44.350
But I know that I should
see the mass, the damping,

00:05:44.350 --> 00:05:46.060
and the stiffness there.

00:05:46.060 --> 00:05:50.490
So we have a complete parallel
between two important fields

00:05:50.490 --> 00:05:54.730
of engineering, the electric
engineering with L, R,

00:05:54.730 --> 00:05:59.590
and 1 over C, mechanical
engineering with M,

00:05:59.590 --> 00:06:04.090
B for damping, and
K for stiffness.

00:06:04.090 --> 00:06:11.920
And actually, that parallel
allowed analog computers--

00:06:11.920 --> 00:06:16.140
which came before
digital computers

00:06:16.140 --> 00:06:21.150
and lost out in
that competition.

00:06:21.150 --> 00:06:27.430
An analog computer was just
solving this linear equation

00:06:27.430 --> 00:06:34.010
by actually imposing the voltage
and measuring the current.

00:06:34.010 --> 00:06:38.990
So an analog computer
actually solved the equation

00:06:38.990 --> 00:06:44.240
by creating the model
and measuring the answer.

00:06:44.240 --> 00:06:48.620
But we're not creating
an analog computer here.

00:06:48.620 --> 00:06:53.240
We're just doing
differential equations.

00:06:53.240 --> 00:06:57.570
So why don't I figure
out what that W is.

00:06:57.570 --> 00:06:58.790
So what am I going to do?

00:06:58.790 --> 00:07:02.880
As always, I have this equation.

00:07:02.880 --> 00:07:04.820
I have a pure exponential.

00:07:04.820 --> 00:07:07.700
I look for a solution
of that same form.

00:07:07.700 --> 00:07:09.070
I plug it in.

00:07:09.070 --> 00:07:11.910
And I get an equation for W.

00:07:11.910 --> 00:07:14.030
That's exactly what I'll
do on the next board.

00:07:14.030 --> 00:07:19.290
I'll put W e to the I
omega t into this equation

00:07:19.290 --> 00:07:24.170
and find W. Let's do it.

00:07:28.430 --> 00:07:31.030
Maybe I'll bring
that down just a hair

00:07:31.030 --> 00:07:35.520
and I'll do it here where
you can watch me do it.

00:07:35.520 --> 00:07:38.130
So I have L times
the derivative.

00:07:38.130 --> 00:07:42.560
So I have L. The
derivative will bring down

00:07:42.560 --> 00:07:51.640
an I omega L. Everything is
going to multiply W and match

00:07:51.640 --> 00:07:58.590
V. When I put this
into the equation,

00:07:58.590 --> 00:08:05.760
the derivative is an I omega
L W e to the I omega t,

00:08:05.760 --> 00:08:08.910
and it's matching V
e to the I omega t.

00:08:08.910 --> 00:08:15.980
Now, what happens when I put
I in for that second term,

00:08:15.980 --> 00:08:22.240
R. I just get an R. R times
W times e to the I omega t.

00:08:22.240 --> 00:08:23.230
No problem.

00:08:23.230 --> 00:08:28.500
And now finally, a 1
over C. The integral.

00:08:28.500 --> 00:08:34.090
The integral of the
exponential brings down--

00:08:34.090 --> 00:08:41.580
let me put it in the denominator
neatly-- I divide by I omega

00:08:41.580 --> 00:08:44.450
when I integrate e
to the I omega t.

00:08:44.450 --> 00:08:48.860
I have a division by I omega.

00:08:48.860 --> 00:08:50.240
That's it.

00:08:50.240 --> 00:08:52.190
That's it.

00:08:52.190 --> 00:08:56.630
Those are the three terms that
come-- times W, the unknown.

00:08:56.630 --> 00:08:57.415
This is to find.

00:09:00.090 --> 00:09:03.010
And of course, we
find it right away.

00:09:03.010 --> 00:09:08.870
We find W is V
over-- and now we're

00:09:08.870 --> 00:09:13.970
seeing this I
omega L plus R. Oh,

00:09:13.970 --> 00:09:17.280
let me combine the I omegas.

00:09:17.280 --> 00:09:21.440
Combine the real part
and the imaginary part.

00:09:21.440 --> 00:09:27.120
The real part is R. And
the imaginary part is I

00:09:27.120 --> 00:09:38.100
omega L minus 1 over I omega C.

00:09:38.100 --> 00:09:39.790
Straightforward.

00:09:39.790 --> 00:09:42.610
And that has a name.

00:09:42.610 --> 00:09:45.080
That is the resistance.

00:09:45.080 --> 00:09:47.370
But when there's
also terms coming

00:09:47.370 --> 00:09:50.680
from an inductance
and a capacitance,

00:09:50.680 --> 00:09:54.800
then the whole thing is
called the impedance.

00:09:54.800 --> 00:10:02.260
So this whole thing,
this whole denominator,

00:10:02.260 --> 00:10:04.625
is called the complex impedance.

00:10:09.790 --> 00:10:14.150
Believe me, all these
ideas are so important.

00:10:14.150 --> 00:10:16.420
There's a whole vocabulary here.

00:10:16.420 --> 00:10:21.020
But you see, we've done
exactly the same thing

00:10:21.020 --> 00:10:24.080
for other constant
coefficient equations.

00:10:24.080 --> 00:10:28.910
We just called the coefficient
A, B, C. Or maybe M, B, K.

00:10:28.910 --> 00:10:32.310
And now we have slightly
different letters,

00:10:32.310 --> 00:10:35.430
but we don't have
a new idea here.

00:10:35.430 --> 00:10:39.690
The idea is this 1 over, that
1 over the impedance, that

00:10:39.690 --> 00:10:48.130
will be the transfer function,
which multiplies the source

00:10:48.130 --> 00:10:54.440
to give the complex number
W. And W is a complex number.

00:10:54.440 --> 00:10:58.810
I have to now think about that.

00:10:58.810 --> 00:11:06.030
And that impedance
is always called Z.

00:11:06.030 --> 00:11:12.910
So we now have a new letter
for the important quantity that

00:11:12.910 --> 00:11:15.550
shows up in the
denominator there.

00:11:15.550 --> 00:11:19.780
And again, its real
part is the resistance.

00:11:19.780 --> 00:11:23.000
Its imaginary part
comes from L and C.

00:11:23.000 --> 00:11:26.200
So we can easily see
how large-- what's

00:11:26.200 --> 00:11:28.560
the size of that impedance?

00:11:28.560 --> 00:11:35.830
What's going to be the
magnitude of this current?

00:11:35.830 --> 00:11:37.640
We want the size of that number.

00:11:37.640 --> 00:11:39.980
V is the size of the voltage.

00:11:39.980 --> 00:11:42.370
Here is the size
of the impedance.

00:11:42.370 --> 00:11:45.360
And the answer will
give us the size of W.

00:11:45.360 --> 00:11:48.660
I'm using size or
magnitude to say

00:11:48.660 --> 00:11:51.500
that when I only
do magnitudes, you

00:11:51.500 --> 00:11:53.930
won't be seeing the phase lag.

00:11:53.930 --> 00:11:58.120
So complex numbers,
like this complex number

00:11:58.120 --> 00:12:00.890
has a magnitude which
we're about to write down.

00:12:00.890 --> 00:12:05.590
And also it has a
phase lag that tells us

00:12:05.590 --> 00:12:07.920
how much is in
the imaginary part

00:12:07.920 --> 00:12:09.710
and how much is
in the real part.

00:12:09.710 --> 00:12:11.750
But the magnitude is easy.

00:12:11.750 --> 00:12:14.300
What's the magnitude
of a complex number?

00:12:14.300 --> 00:12:19.790
It's the real part squared and
the imaginary part squared.

00:12:19.790 --> 00:12:22.830
Oh, that should have been
a plus there, I think.

00:12:22.830 --> 00:12:26.750
I don't know how
it became a minus.

00:12:26.750 --> 00:12:30.860
It will become a minus,
so I was thinking

00:12:30.860 --> 00:12:33.920
if I put the I up there.

00:12:33.920 --> 00:12:35.840
Let me show you what I'm saying.

00:12:35.840 --> 00:12:41.300
The imaginary part
is omega L minus 1

00:12:41.300 --> 00:12:49.260
over omega C. What I'm saying
is that if I put the I up there,

00:12:49.260 --> 00:12:55.520
then 1 over I is minus I.
That's the brilliant step I just

00:12:55.520 --> 00:12:56.610
took there.

00:12:56.610 --> 00:13:00.470
So all that squared.

00:13:00.470 --> 00:13:03.000
Are you OK with that?

00:13:03.000 --> 00:13:05.950
It's the real part squared,
which is the resistance.

00:13:05.950 --> 00:13:08.860
And this combination
gives the imaginary part.

00:13:08.860 --> 00:13:10.140
We square that.

00:13:10.140 --> 00:13:12.630
That's maybe called
the reactants.

00:13:12.630 --> 00:13:16.450
And the sum of those squares
is the impedance squared,

00:13:16.450 --> 00:13:18.870
the magnitude.

00:13:18.870 --> 00:13:25.860
So we have essentially
successfully solved

00:13:25.860 --> 00:13:32.940
a second order constant
coefficient single equation

00:13:32.940 --> 00:13:35.930
for the current.

00:13:35.930 --> 00:13:36.990
What to do now.

00:13:36.990 --> 00:13:40.540
Just let me add a
little bit more.

00:13:43.510 --> 00:13:46.580
Maybe just a comment.

00:13:46.580 --> 00:13:50.620
That video was about one loop.

00:13:50.620 --> 00:13:54.850
When I told Dr. Mohler that
one of the applications, one

00:13:54.850 --> 00:13:57.700
of the real applications
in this series of videos

00:13:57.700 --> 00:14:00.540
would be an RLC
circuit, his reply

00:14:00.540 --> 00:14:04.940
was an RLC circuit is
not an application, not

00:14:04.940 --> 00:14:07.090
a realistic application.

00:14:07.090 --> 00:14:08.980
One loop.

00:14:08.980 --> 00:14:14.620
So how do we proceed
with a full scale circuit

00:14:14.620 --> 00:14:21.420
with many nodes, many resistors,
many conductors, many edges?

00:14:21.420 --> 00:14:24.290
Well, we have a big
decision to make.

00:14:24.290 --> 00:14:26.330
And that's the comment
I want to make.

00:14:26.330 --> 00:14:27.190
They have a choice.

00:14:27.190 --> 00:14:31.770
They can use Kirchoff's
current law at the nodes

00:14:31.770 --> 00:14:35.960
and solve for the
voltages at those nodes.

00:14:35.960 --> 00:14:39.670
Or they can do as
we did for one loop,

00:14:39.670 --> 00:14:43.590
use Kirchoff's voltage
law around that one loop

00:14:43.590 --> 00:14:48.290
which said that the
currents in the loop

00:14:48.290 --> 00:14:52.910
gave a total voltage
drop adding to 0.

00:14:52.910 --> 00:14:57.260
So we solve the current
equation for the unknown I. This

00:14:57.260 --> 00:15:00.050
is what we did for one loop.

00:15:00.050 --> 00:15:05.680
My message is just for a big
system, this is the winner.

00:15:05.680 --> 00:15:10.110
So writing down the
equations in terms

00:15:10.110 --> 00:15:13.480
of Kirchoff's current law,
that the currents-- we

00:15:13.480 --> 00:15:16.860
get the nodal
picture, the picture

00:15:16.860 --> 00:15:20.580
with an equation for every
node instead of the picture

00:15:20.580 --> 00:15:22.700
for an equation for every loop.

00:15:22.700 --> 00:15:27.620
Because it's not so easy to see
which are the loops to consider

00:15:27.620 --> 00:15:32.453
and which loops are
combinations of other loops.

00:15:34.960 --> 00:15:36.740
The linear algebra
is the question.

00:15:36.740 --> 00:15:40.780
And the linear algebra,
to get the loop picture

00:15:40.780 --> 00:15:43.810
independent and clear,
is more difficult

00:15:43.810 --> 00:15:46.620
than the node picture.

00:15:46.620 --> 00:15:51.790
The node picture with
the unknown voltages,

00:15:51.790 --> 00:15:58.740
V at the nodes, is the good one.

00:15:58.740 --> 00:16:03.150
And the matrix that comes into
that is the incidence matrix.

00:16:03.150 --> 00:16:07.340
It connects nodes and edges.

00:16:07.340 --> 00:16:09.960
It says how the network
is put together.

00:16:09.960 --> 00:16:14.130
And that matrix, I'll study with
a little bit of linear algebra.

00:16:14.130 --> 00:16:16.220
So that comes in a later video.

00:16:16.220 --> 00:16:19.180
If you look for
incidence matrices,

00:16:19.180 --> 00:16:23.780
you'll see probably two
videos about those very, very

00:16:23.780 --> 00:16:26.810
important and
beautiful matrices.

00:16:26.810 --> 00:16:28.640
Thank you.