EECS 16B Designing Information Devices and Systems II

Fall 2018 Elad Alon and Miki Lustig Homework 3

This homework is due on Wednesday September 19, 2018 at 11:59PM.
Self-grades are due on Monday, September 24, 2018 at 11:59PM.

1. Mechanical 2nd Order Differential Equation

2 dy
Solve 3 ddt 2y − 12 dy
dt + 24y = 24, where y(0) = 1 and (0) = 2
dt
Solution:
d2 y
First, we’ll divide both sides by 3 to get a coefficient of 1 in front of the dt 2
term:
d2 y
dt 2
− 4 dy
dt + 8y = 8

Since this is a nonhomogeneous differential equation, we will perform a change of coordinates. Choose ỹ
d2 ỹ d2 y
such that 8ỹ = 8(y − 8), which means we need ỹ = y − 1. Then, ddtỹ = dy
dt and dt 2 = dt 2 .

d2 y dy
8= − 4 + 8y
dt 2 dt
2
d ỹ d ỹ
8 = 2 − 4 + 8(ỹ + 1)
dt dt
2
d ỹ d ỹ
0 = 2 − 4 + 8ỹ
dt dt
"#
ỹ(t)
We set up the matrix differential equation with the state vector ~x(t) = d ỹ .
dt (t)
" # " #" #
d ỹ
dt (t) =
0 1 ỹ(t)
d2 ỹ −8 4 ddtỹ (t)
dt 2
(t)

Call this coefficient matrix A. We will solve for its eigenvalues, which are roots of det(A − λ I) = 0.
" #
−λ 1 
det  = λ 2 − 4λ + 8 = 0
−8 4 − λ
√
4 ± 16 − 32
λ= = 2±2j
2
The λ s are non-real complex conjugates, so the general solution is given by:

ỹ(t) = k1 e(2+2 j)t + k2 e(2−2 j)t

Because k1 , k2 are complex conjugate pairs, we can simplify to the form shown below:

ỹ(t) = e2t (c1 cos(2t) + c2 sin(2t))

EECS 16B, Fall 2018, Homework 3 1

 We reverse the change of variables y 7→ ỹ.

y(t) = ỹ(t) + 1
= e2t (c1 cos(2t) + c2 sin(2t)) + 1
dy
(t) = 2e2t (c1 cos(2t) + c2 sin(2t)) + e2t (−2c1 sin(2t) + 2c2 cos(2t))
dt
dy
To find c1 and c2 , we substitute the initial conditions for y and dt . Recall sin(0) = 0 and cos(0) = 1.

c1 + 1 = 1
2c1 + 2c2 = 2

From this we have c1 = 0 and c2 = 1. Returning to the solution for y,

y(t) = e2t sin(2t) + 1

2. Solution to Repeated Roots

In lecture, we claimed that the solution to a second-order differential equation with repeated eigenvalue λ0
is y = c1 eλt + c2teλt . In this problem, we will show why this solution is valid.
2
(a) Given a differential equation ddt 2y + a1 dy
dt + a0 y = 0, assume that both eigenvalues of the A matrix with
" #
y(t)
the state vector defined as dy(t) are λ0 . Find a0 , a1 in terms of λ0 .
dt
" #
y
Solution: Let ~x = dy . Then our vector differential equation is
dt
" # " #" #
dy
dt 0 1 y
d2y = dy
−a0 −a1 dt
dt 2

If we find the characteristic polynomial of the matrix, we get λ 2 + a1 λ + a0 = 0. We need this to

have two roots of λ0 . A polynomial with two roots λ0 looks like (λ − λ0 )(λ − λ0 ) = λ 2 − 2λ0 λ + λ02 .
Therefore, a0 = λ02 , a1 = −2λ0 , so

d2y dy
− 2λ0 + λ02 y = 0
dt 2 dt

(b) Let’s assume the solution to our differential equation is c1 eλ0t + c2teλ0t . Verify that this solution satis-
fies the differential equation you get when using the a0 and a1 you found in part (a).

EECS 16B, Fall 2018, Homework 3 2

 Solution:
y = c1 eλ0t + c2teλ0t
dy
= λ0 c1 eλ0t + c2 eλ0t + λ0 c2teλ0t
dt
d2y
= λ02 c1 eλ0t + λ0 c2 eλ0t + λ0 c2 eλ0t + λ02 c2teλ0t
dt 2
= λ02 c1 eλ0t + 2λ0 c2 eλ0t + λ02 c2teλ0t
d2y dy
− 2λ0 + λ02 y = λ02 c1 eλ0t + 2λ0 c2 eλ0t + λ02 c2teλ0t − 2λ0 (λ0 c1 eλ0t + c2 eλ0t + λ0 c2teλ0t ) + λ02 (c1 eλ0t + c2teλ0t )
dt 2 dt
= c1 eλ0t (λ02 − 2λ02 + λ02 ) + c2 eλ0t (2λ0 − 2λ0 ) + c2teλ0t (λ02 − 2λ02 + λ02 )
=0

(c) Making the same assumption as part (b), show that we can always find constants c1 , c2 such that we
can satisfy initial conditions y(0) = y0 , dy 0
dt (0) = y0
Solution:
y(0) = y0 = c1 e0 + c2 (0)e0
= c1
dy
(0) = y00 = λ0 c1 e0 + c2 e0 + λ0 c2 (0)e0
dt
= λ0 c1 + c2
We have two linearly independent equations and we’re solving for two variables, so we should always
reach a unique solution.

3. Fun with Inductors

L
IL

S1 t = 0 S2 t =0

+ Vin
− +

R1 R2 Vout

Figure 1: Circuit A

(a) Consider circuit A. Assuming that for t < 0, switch S1 is on and switch S2 is off (and both switches
have been in these states indefinitely), what is iL (0)?
Solution: When S1 is on and S2 is off for a long period of time, didtL = 0 because the circuit will have
reached a steady state, and the current through R1 will be equal to iL . We find
Vin −VL −VR1 = 0

EECS 16B, Fall 2018, Homework 3 3

 diL
Vin − L (0) − iL (0)R1 = 0
dt
Vin − iL (0)R1 = 0
Vin
iL (0) =
R1

(b) Now let’s assume that for t ≥ 0, S1 is off and S2 is on. Solve for Vout (t) for t ≥ 0.
Solution:
Vin −VL −Vout = 0
diL
Vin − L − iL R2 = 0
dt
diL R2 Vin
+ iL =
dt L L
This is a non-homogenous first order differential equation in iL . We can solve for iL (t) and then use
Ohm’s law to find Vout (t) after this has been solved.

diL R2 Vin
+ (iL − ) = 0
dt L R2

Let i˜L = iL − VRin2 . We now have:

d ĩL R2
+ ĩL = 0
dt L
The general solution is given by:
R2
ĩL (t) = c1 e− L t
Resubstituting back iL , we have:
Vin R2
iL (t) = + c1 e− L t
R2
Applying initial conditions, we know:
Vin Vin
iL (0) = + c1 =
R2 R1
Vin Vin
c1 = −
R1 R2
Our solution for iL (t) thus becomes:
Vin Vin Vin R2
iL (t) = + ( − )e− L t
R2 R1 R2
Since Vout (t) = iL (t)R2 ,
R2 R2
Vout (t) = Vin (1 + ( − 1)e− L t )
R1

EECS 16B, Fall 2018, Homework 3 4

 (c) If Vin = 1V , L = 1nH, R1 = 1kΩ, and R2 = 10kΩ, what is the maximum value of Vout (t) for t ≥ 0?
R2
Solution: Since the coefficient in front of our time-varying component e− L t , given by RR12 − 1 = 9, is
positive, Vout (t) undergoes decay over time. Therefore, the maximum value is achieved at t = 0:
R2
maxVout (t) = Vout (0) = Vin = 10V
R1

(d) In general, if we want maxVout (t) to be greater than Vin , what relationship needs to be maintained
between the values of R1 and R2 ?
Solution: As long as the coefficient on our exponential term, given by RR21 − 1, is greater than 0 (i.e.
when RR12 > 1) then the maximum value of Vout (t) will be achieved at t = 0 and will have a value of
R2 R2
R1 Vin > Vin . Otherwise, if R1 ≤ 1, the maximum value of Vout (t) is reached at t = ∞, where Vout = Vin
regardless of R2 and R1 . Therefore, our necessary condition for the maximum of Vout to be greater than
Vin is:
R2 > R1

(e) Now assume that at time t = t1 , switch S2 was turned off, and switch S1 was turned back on. Solve for
iL (t) for t > t1 . If R2 > R1 , how does this iL (t) for t > t1 compare with the initial condition iL (0) you
found in part (a)?
Solution: Our new initial condition for t > t1 is given by plugging in t = t1 into the equation for iL (t)
R2
we found in part (b). Thus, iL (t1 ) = VRin2 + ( VRin1 − VRin2 )e− L t1 .
We can write the relationship between the current through the inductor and the current through R1 :

iL = iR1

VR1
iL =
R1
Vin −VL
iL =
R1
Vin L didtL
iL = −
R1 R1
diL R1 Vin
+ iL =
dt L L
This is a first order non-homogeneous differential equation similar to that found in part (b), except
with R1 in place of R2 . Following those steps in part (b), we find the general solution:
Vin R1
iL (t) = + c1 e− L t
R1
To find c1 we apply our initial condition:
Vin R1 Vin Vin Vin R2
iL (t1 ) = + c1 e− L t1 = + ( − )e− L t1
R1 R2 R1 R2
R1 Vin Vin R2
c1 e− L t1 = ( − )(1 − e− L t1 )
R2 R1

EECS 16B, Fall 2018, Homework 3 5

 R2
( VRin2 − VRin1 )(1 − e− L t1 )
c1 = R1
e− L t1
Vin Vin R1 t1 R1 −R2
c1 = ( − )(e L − e L t1 )
R2 R1
Thus, we have
Vin Vin Vin R1 R1 −R2 R1
iL (t) = + ( − )(e L t1 − e L t1 )e− L t
R1 R2 R1
for t > t1 . We also see that as t → ∞, iL (t) for t > t1 becomes:
Vin Vin Vin R1 R1 −R2 R1
iL (t = ∞) = + ( − )(e L t1 − e L t1 )e− L ∞
R1 R2 R1
Vin Vin Vin R1 R1 −R2
iL (t = ∞) = + ( − )(e L t1 − e L t1 )(0)
R1 R2 R1
Vin
iL (t = ∞) = = iL (0)
R1
Thus, if we turn S2 back off and S1 back on as was described in this part, we will eventually revert
back to the initial state from which we started! Specifically, if R2 > R1 , iL (t) at t = t1 will be less than
our initial condition iL (0), and iL (t) will rise to iL (0) over time.

4. Oscillators
In this question, we’ll be looking at an oscillator circuit. There are many types of oscillators, but this circuit
is known as an LC tank. It’s called an oscillator because the circuit produces a repetitive voltage waveform
under the right initial conditions.

In this circuit, we have an inductor L = 10nH and capacitor C = 10pF in parallel:

L Vout C
− ic
iL

(a) If iL (0) = 0A and Vout (0) = 0V, derive an expression for Vout (t) for t ≥ 0. Use Vout and iL as your state
variables.
Solution:
Since the inductor and capacitor are in parallel:

VL = Vc = Vout

KCL gives:
dVout
iL = −ic = −C
dt
dVout 1
= − iL
dt C

EECS 16B, Fall 2018, Homework 3 6

 diL
VL = Vout = L
dt
diL 1
= Vout
dt L
Putting it into matrix form: " # " #" #
dVout
dt 0 − C1 Vout
diL = 1
dt L 0 iL
Finding the eigenvalues: " #
 −λ − C1 = λ2 + 1 = 0
det 1
L −λ LC

1
λ1,2 = 0 ± j √
LC
The eigenvalues are purely imaginary, so the solution to the differential equation takes the form:
   
1 1
Vout (t) = c1 cos √ t + c2 sin √ t
LC LC
dVout (0)
We can find c1 and c2 with initial conditions for Vout (0) and dt

Vout (0) = 0 = c1
   
dVout c1 1 c2 1
= − √ sin √ t + √ cos √ t
dt LC LC LC LC
dVout (0) 1 c2
= ic (0) = 0 = √
dt C LC
c1 = c2 = 0
Vout (t) = 0

(b) Now let’s see how the circuit reacts with non-zero initial current. If iL (0) = 50µA and Vout (0) = 0V,
derive an expression for Vout (t) for t ≥ 0. How does the amplitude of Vout change over time?
Solution:
The only thing that changed from part (a) was the initial conditions, which means we still have:
   
1 1
Vout (t) = c1 cos √ t + c2 sin √ t
LC LC
Plugging in initial conditions:
Vout (0) = 0 = c1
ic (0) = −iL (0) = −50 × 10−6
dVout (0) 1 −50 × 10−6 c2
= ic (0) = −11
=√
dt C 10 10 × 10−11
−8

c1 = 0
5 √
c2 = − √ × 10−3 = −5 10 × 10−4
10
 √  √ 
Vout (t) = −5 10 × 10−4 sin 10 × 109t
Notice that the amplitude of Vout is constant.

EECS 16B, Fall 2018, Homework 3 7

 (c) In order to ensure an initial condition where we get non-zero output, some switches and a voltage
source have been added to the circuit. For t ≤ 0, switch S1 is on while S2 is off. Find Vc (0) and iL (0).
Use component values Vs = 3V, L = 10nH and C = 10pF.

t =0 t =0

S2 S1
+

L Vout C + Vs
−
− ic
iL

Solution: For t ≤ 0, the circuit looks like:

L Vout C + Vs
−
− ic
iL

There is no path for current to flow through the inductor, so

iL (0) = 0A

Since the capacitor is in parallel with Vs :

Vout (0) = Vs = 3V

(d) At t = 0, the switches flip state (S1 turns off and S2 turns on). Derive an expression for Vout (t) for t ≥ 0.
Use the same component values as part (c).
Solution:
For t ≥ 0, the circuit looks like:

L Vout C + Vs
−
− ic
iL

The voltage source is in series with an open circuit, so it doesn’t affect the circuit. This means the
circuit is the same as in part (a), so we can say:
   
1 1
Vout (t) = c1 cos √ t + c2 sin √ t
LC LC
Plugging in initial conditions:
Vout (0) = 3 = c1
dVout (0) 1 c2
= ic (0) = 0 = √
dt C 10 × 10−11
−8

EECS 16B, Fall 2018, Homework 3 8

 c1 = 3
c2 = 0
√ 
Vout (t) = 3 cos 10 × 109t

(e) Let’s see what happens when there is a parasitic resistance R in parallel with the LC tank.

R L Vout C
− ic
iR iL

If iL (0) = 0µA and Vout (0) = 3V, derive an expression for Vout (t) for t ≥ 0. Use component values
C = 10pF, L = 10nH, and R = 100kΩ. How does the amplitude of Vout change over time?
Solution: For this problem, we’ll use the same state variables as part (a), Vout and iL .
Since all three components are in parallel, we can say:

VR = VL = Vc = Vout
dVout diL
To solve the differential equation, we need to get dt and dt in terms of Vout and iL

dVout
C = ic
dt
Using KCL:
iR + iL + ic = 0
Vout dVout
+ iL +C =0
R dt
dVout 1 1
= − Vout − iL
dt RC C
diL
VL = Vout = L
dt
diL 1
= Vout
dt L
Putting into matrix form: " # " #" #
dVout 1
dt − RC − C1 Vout
diL = 1
dt L 0 iL
Finding eigenvalues: " #
1
 − RC − λ − C1 = λ2 + 1 λ + 1 = 0
det 1
L −λ RC LC
s 2
1 1 1
λ1,2 = − ± −
2RC 2RC LC

EECS 16B, Fall 2018, Homework 3 9

 s
1 1
λ1,2 = − −11
± − 1019
5
2 × 10 × 10 4(10 × 10−11 )2
5

p
λ1,2 = −5 × 105 ± 25 × 1010 − 1019
√
Since 25 × 1010 << 1019 , we can estimate the square root term as −1019 :
√
λ1,2 = −5 × 105 ± j 10 × 109
5
 √ √ 
Vout (t) = e−5×10 t c1 cos( 10 × 109t) + c2 sin( 10 × 109t)

Plugging in initial conditions:

Vout (0) = 3 = c1

iL (0) + ic (0) + iR (0) = 0

dVout (0) Vout (0)
0 +C + =0
dt R
dVout (0) 3
=− 5 = −3 × 106
dt 10 × 10−11

dVout 5
 √ √ 
= (−5 × 105 )e−5×10 t c1 cos( 10 × 109t) + c2 sin( 10 × 109t)
dt  √
5 √ √ √ 
+ e−5×10 t −( 10 × 109 )c1 sin( 10 × 109t) + ( 10 × 109 )c2 cos( 10 × 109t) (1)

dVout (0) √
= −3 × 106 = (−5 × 105 )c1 + ( 10 × 109 )c2
dt
−15 × 10−4 √
c2 = √ = −15 10 × 10−5
10
This gives us:
5t
 √ √ √ 
Vout (t) = e−5×10 3 cos( 10 × 109t) − 15 10 × 10−5 sin( 10 × 109t)

The amplitude of Vout decays over time. This is due to energy being dissipated across the resistor as
energy transitions between the inductor and capacitor.
(f) In order to counteract the parasitic resistance, we create a negative resistance (which can be done using
transistors, but will not be covered in this class) in parallel with the other components:

− R2 R L Vout C
− ic
iL

EECS 16B, Fall 2018, Homework 3 10

 If iL (0) = 0µA and Vout (0) = 3V, Derive an expression for Vout (t) for t ≥ 0. Use component values
C = 10pF, L = 10nH, and R = 100kΩ. How does the amplitude of Vout change over time?
Solution:
Since the negative resistance and positive resistance are in parallel, we can combine the two:

R −R
 
−R 2
Req = R|| = = −R
2 R + −R2

Now we have the same circuit as part (e), but with R as negative. We can use the previous equations
from part (e), and just plug in −R for R:
s 2
1 1 1
λ1,2 = − ± −
2(−R)C 2(−R)C LC
√
λ1,2 = 5 × 105 ± 10 × 109
5
 √ √ 
Vout (t) = e5×10 t c1 cos( 10 × 109t) + c2 sin( 10 × 109t)

Plugging in initial conditions:

Vout (0) = 3 = c1

iL (0) + ic (0) + iR (0) = 0

dVout (0) Vout (0)
0 +C + =0
dt −R
dVout (0) 3
= 5 = 3 × 106
dt 10 × 10−11

dVout 5
 √ √ 
= (5 × 105 )e5×10 t c1 cos( 10 × 109t) + c2 sin( 10 × 109t)
dt  √
5 √ √ √ 
+ e5×10 t −( 10 × 109 )c1 sin( 10 × 109t) + ( 10 × 109 )c2 cos( 10 × 109t) (2)

dVout (0) √
= 3 × 106 = (5 × 105 )c1 + ( 10 × 109 )c2
dt
15 × 10−4 √
c2 = √ = 15 10 × 10−5
10
This gives us:
5t
 √ √ √ 
Vout (t) = e5×10 3 cos( 10 × 109t) + 15 10 × 10−4 sin( 10 × 109t)

The amplitude of Vout grows over time.

EECS 16B, Fall 2018, Homework 3 11

 (g) (BONUS)
What value should the negative resistor have if we want to maintain a constant amplitude on Vout (t)?
Solution:
We saw in part (b) that the amplitude of Vout is constant when the only connection to the LC circuit is
through an open circuit, i.e. when Req = ∞. For calculation purposes, we’ll call the negative resistor
Rn .
RRn
Req = ∞ =
R + Rn
For Req to go to infinity, then we need
Rn = −R

5. Write Your Own Question And Provide a Thorough Solution.

Writing your own problems is a very important way to really learn material. The famous “Bloom’s Tax-
onomy” that lists the levels of learning is: Remember, Understand, Apply, Analyze, Evaluate, and Create.
Using what you know to create is the top level. We rarely ask you any homework questions about the lowest
level of straight-up remembering, expecting you to be able to do that yourself (e.g. making flashcards). But
we don’t want the same to be true about the highest level. As a practical matter, having some practice at
trying to create problems helps you study for exams much better than simply counting on solving existing
practice problems. This is because thinking about how to create an interesting problem forces you to really
look at the material from the perspective of those who are going to create the exams. Besides, this is fun. If
you want to make a boring problem, go ahead. That is your prerogative. But it is more fun to really engage
with the material, discover something interesting, and then come up with a problem that walks others down
a journey that lets them share your discovery. You don’t have to achieve this every week. But unless you try
every week, it probably won’t ever happen.

Contributors:

• Jaymo Kang.

• Arda Sahiner.

• Kyle Tanghe.

EECS 16B, Fall 2018, Homework 3 12