CAMOSUN COLLEGE

ELECTRONICS DEPARTMENT
________________________________________________________________

ECET 221: LAB # 03-04

Exploration of the second order sallen key STEP
response AT different gain settings
_____________________________________________________________________________________

I. Objective:

• The objective of this lab-session is to investigate the response of a typical

second order system modeled by a Sallen-Key circuit: The reason of this
particular choice is that second order systems behave like the majority of
continuous control systems. A set point is represented by a step function
and the response may be overdamped (slow process), critically damped
(ideal), oscillatory or underdamped (decaying oscillations)
• you will build a simple op amp circuit and investigate its behavior for
various gains. Your experiments in the lab will be supported by PSpice
circuit simulation as well as MATLAB transfer function analysis.

II. Background:

In many systems, gain will determine how the system responds. This is because
the damping factor ζ , and therefore the poles, depend on the gain of the system.
In passive second order systems ζ depends on the components R,L, and C. A
general second order system may be described as having the transfer function:
2
𝐾𝜔𝑛
𝐻(𝑠) = 2
𝑠2 + 2 𝜁 𝜔𝑛 𝑠 + 𝜔𝑛

where ωn is the natural frequency, ζ is the damping factor, and K is the DC gain.

For damping factors greater than one (ζ>1), the system is said to be
overdamped. This means that when a system is hit with an input step function,
the system output will step up rather sluggishly.

1/26 ECET 221_Lab_3/4

When the damping factor is equal to one(]ζ=1), the system is said to be
critically damped. This means the output will step up as quickly as it can
without overshooting.

When the damping factor is less than one(0<ζ<1), the system is said to be
underdamped. This means the system output steps up rapidly, but overshoots
the mark before settling to its final value.

When the damping factor is equal to zero(ζ=0), the system is said to be

oscillatory. This means the system output is a sinusoidal signal.

When the damping factor is negative(ζ<0), the system is said to be unstable.

This means the system output is infinite (clamped by op amps).

If the second order transfer function is known, values for ζ and wn may be
determined algebraically. Percent overshoot is the maximum percentage
overshoot over steady state response. It may be measured from a graph of
system step response. Theoretically,

−𝜋𝜁
( )
−𝜋𝜁 √1−𝜁2
% overshoot = 100 exponential ( ) = 100𝑒
√ 1-𝜁 2

Settling time is the time required for the response to come permanently within a
2% band around the steady state value. It may be measured from a graph of
system step response. Theoretically,
5
Ts =
 n

2/26 ECET 221_Lab_3/4

Part I: Theory

The Sallen-Key circuit circuit of interest is:

20k pot

Figure 1: Circuit Laboratory 3

The circuit may be transformed using Laplace techniques. The impedance of a

1
resistor is R; the impedance of a capacitor is . Notice that all resistors (except
sC
R4) have the same value, and all capacitors have the same value. Your analysis
will be easiest if you let R1 = R2 = R3 = R. Similarly use C1 = C2 = C.
Numerical values can be substituted in at the end. In other words, your
expressions will be functions of R, C, and R4.

• From the transformed circuit (using Laplace), find the following relationships:

(a) Vin in terms of Va, Vb and Vout (KCL equation at node Vb)
(b) Vb in terms of Va (KCL equation at node Va)
(c) Va in terms of Vout (voltage divider at - terminal) NOTE: For this
R + R4
relationship, let K = .
R
(d) Use (b) and (c) to solve for Vb in terms of Vout.
(e) Substitute (c) and (d) into (a) to solve for Vin in terms of Vout.
Vout
(f) From (e), find the transfer function . Be sure the leading coefficient in
Vin
the denominator is one.

You should end up with:

3/26 ECET 221_Lab_3/4

 𝑅4 1
(1+ )∗( )2 𝐾∗(𝜔𝑜 )2
𝑅 𝑅𝐶
𝑇(𝑠) = 𝑅4 𝑠 1 or 𝑇(𝑠) = 𝑠2 +2𝜍𝜔 2
𝑠2 +(2− ) +( )2 𝑜 𝑠+𝜔𝑜
𝑅 𝑅𝐶 𝑅𝐶
• Now consider the transfer function found previously

- Calculate the natural frequency ωo .

- For each of the following values of K, determine:
The damping factor and the corresponding behavior of the response.
When appropriate, calculate the settling time and the % overshoot.
3−𝐾
- The poles of the transfer function. (remember for Sallen Key 𝜉 = 2 ;
1
𝜔𝑛 = 𝑅𝐶 and the poles are s1,2 = - 𝜉𝜔𝑜 +/- 𝜔𝑜 √(𝜉)2 − 1. Plot the poles in the s
plane

(a) K=1
(b) K=1.5
(c) K=2.9
(d) K=3
(e) K=4

Theorical formulas

Figure 2: General circuit Sallen Key

In this part, I will show how can I get the transfer function for further calculations:
PART A
It is known by KCL in note Vb:
𝐼𝑏 − 𝐼𝑖𝑛 − 𝐼𝑎 − 𝐼0𝑢𝑡 = 0
𝐼𝑏 = 𝐼𝑖𝑛 − 𝐼𝑎 − 𝐼0𝑢𝑡
1 1 𝑉𝑖𝑛 𝑉𝑎
𝑉𝑏 ∗ ( + + 𝐶𝑆) = + + 𝑉𝑜𝑢𝑡 𝑆
𝑅 𝑅 𝑅 𝑅
PART B
It is known by KCL in note Va:
𝐼𝑎 − 𝐼𝑏 − 𝐼𝑜𝑝𝑎𝑚𝑝 − 𝐼𝑔𝑟𝑜𝑢𝑛𝑑 = 0, 𝐼𝑜𝑝𝑎𝑚𝑝 & 𝐼𝑔𝑟𝑜𝑢𝑛𝑑 𝑎𝑟𝑒 0.
𝐼𝑎 = 𝐼𝑏
1 1
𝑉𝑎 ∗ ( + 𝐶𝑆) = 𝑉𝑏 ∗ ( )
𝑅 𝑅
𝑉𝑏 = 𝑉𝑎 ∗ (1 + 𝑅𝐶𝑆)

4/26 ECET 221_Lab_3/4

PART C
The current form the negative voltage of the op-amp same as the output current.
𝐼− = 𝐼𝑜𝑢𝑡
1 1 1
𝑉− ∗ ( + ) = 𝑉𝑜𝑢𝑡 ∗ ( )
𝑅 𝑅4 𝑅4
𝑅
𝑉− = 𝑉𝑜𝑢𝑡 ∗ (1 + )
𝑅4
It is necessary to remember:
𝑅
𝑉𝑎 = 𝑉− 𝑘 = (1 + )
𝑅4
PART D
From part B and C, we will calculate Vb:
𝑅4
𝑉𝑏 = 𝑉𝑜𝑢𝑡 (1 + ) (1 + 𝑅𝐶𝑆)
𝑅
𝑂𝑟 𝑐𝑜𝑛𝑠𝑖𝑑𝑒𝑟𝑖𝑛𝑔 𝐾:
𝑉𝑏 = 𝑉𝑜𝑢𝑡 ∗ 𝐾 ∗ (1 + 𝑅𝐶𝑆)
PART E
Part C and D into A, so I will obtain Vin:
𝑅4 𝑆 1 2
𝑆 2 + (2 − 𝑅 ) ∗ 𝑅𝐶 + (𝑅𝐶 )
𝑉𝑖𝑛 = ∗ 𝑉𝑜𝑢𝑡
𝑅4 1 2
(1 + 𝑅 ) (𝑅𝐶 )
𝑂𝑟 𝑐𝑜𝑛𝑠𝑖𝑑𝑒𝑟𝑖𝑛𝑔 𝐾:
1 𝑆 1 2
𝑆 2 + (3 − −1 ) ∗ 𝑅𝐶 + (𝑅𝐶 )
𝑉𝑖𝑛 = 𝐾 ∗ 𝑉𝑜𝑢𝑡
1 1 2
( −1 ) (𝑅𝐶 )
𝐾

PART F
Transfer function:
𝑅4 1 2
𝑉𝑜𝑢𝑡 (1 + 𝑅 ) (𝑅𝐶 )
=
𝑉𝑖𝑛 𝑅4 𝑆 1 2
𝑆 2 + (2 − 𝑅 ) ∗ 𝑅𝐶 + (𝑅𝐶 )

𝑂𝑟 𝑐𝑜𝑛𝑠𝑖𝑑𝑒𝑟𝑖𝑛𝑔 𝐾:
1 1 2
𝑉𝑜𝑢𝑡 ( ) ( )
= 𝐾 −1 𝑅𝐶
𝑉𝑖𝑛 1 𝑆 1 2
𝑆 2 + (3 − −1 ) ∗ 𝑅𝐶 + (𝑅𝐶 )
𝐾
It is also necessary to consider:
1 3−𝐾
𝜔0 = = 21.3𝑘𝑟𝑎𝑑 𝜉 = 𝑆12 = − 𝜉𝜔0 +/− 𝜔0 ∗ √𝜉 2 − 1
𝑅𝐶 2

5/26 ECET 221_Lab_3/4

Data Table
In the following Table 1: Data Table, I showed the main data required for the
following calculations, and further calculations:
ω₀ 21300 ω₀^2 453690000
K Damping(ξ) R4(Ω) R4(kΩ) 2*ξ * ω₀ K *(ω₀^2) #roots
1 1 0 0 42600 453690000 1
1.5 0.75 2350 2.35 31950 680535000 2
2.9 0.05 8930 8.93 2130 1315701000 2
3 0 9400 9.4 0 1361070000 2
4 -0.5 14100 14.1 -21300 1814760000 2
Table 1: Data Table

It is possible to arrange part F:

𝑉𝑜𝑢𝑡 (𝐾)(𝜔0 )2
= 2
𝑉𝑖𝑛 𝑆 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2

Equations
a) Considering K=1, and the calculations from the Table 1: Data Table.

𝐴 = (𝐾)(𝜔0 )2 = 1 ∗ 453690000 = 453690000

𝑆 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 => 𝑆 2 + 42600𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 1
2

𝑆12 = − 𝜉𝜔0 +/− 𝜔0 ∗ √𝜉 2 − 1 = −21300

In this case, it is critically dumped. The system will return to the ideal state
more quickly than the overdamped system but will not overshoot that ideal state
as the underdamped system will do. Poles equals.

𝑆12

Figure 3: Critically Damped

The poles are the same, this means is critically dumped and the damping factor
is 1.

6/26 ECET 221_Lab_3/4

 1 −1
τ1 = − = − = 46.95𝑢𝑠
𝑆 −21300
𝑠𝑒𝑡𝑡𝑙𝑖𝑛𝑔 𝑡𝑖𝑚𝑒: 𝑡𝑠 = 5 ∗ τ1 = 234.74𝑢𝑠
𝑁𝑜𝑡 𝑜𝑣𝑒𝑟𝑠ℎ𝑜𝑜𝑡𝑒𝑟!

b) Considering K=1.5, and the calculations from the Table 1: Data Table.

𝐴 = (𝐾)(𝜔0 )2 = 1.5 ∗ 453690000 = 680535000

𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 31950𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 2
𝑆12 = − 𝜉𝜔0 +/− 𝜔0 ∗ √𝜉 2 − 1 = −15975 +/−14089𝑖

An underdamped system will return to its ideal state more quickly than under a
critical system. However, it is noticeable the overshoot in this case.

ω0 = 21300 𝑟𝑎𝑑 𝜉 = 0.75

𝑑𝑎𝑚𝑝𝑒𝑑 𝑜𝑠𝑐𝑖𝑙𝑙𝑎𝑡𝑖𝑛𝑔 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 ω𝑑 = √1 − (𝜉)2 ∗ ω0 = √1 − (0.75)2 ∗ 21300
= 14089 𝑟𝑎𝑑
2𝜋 2𝜋
𝑡𝑟𝑎𝑛𝑠𝑖𝑒𝑛𝑡 𝑡𝑖𝑚𝑒: 𝑡𝑛 = = = 445.96𝑢𝑠
ω𝑑 14089
𝜋 𝜋
𝑝𝑒𝑎𝑘 𝑡𝑖𝑚𝑒: 𝑡𝑝 = = = 222.98𝑢𝑠
ω𝑑 14089
𝜉 0.75
−( )𝜋 −( )𝜋
√1−𝜉 2
% 𝑜𝑣𝑒𝑟𝑠ℎ𝑜𝑜𝑡 = 100𝑒 = 100𝑒 √1−0.752 = 2.84%
5
𝑠𝑒𝑡𝑡𝑙𝑖𝑛𝑔 𝑡𝑖𝑚𝑒: 𝑡𝑠 = = 313𝑢𝑠
𝜉ω0

𝑆1

𝑆2

Figure 4: underdamped

It can be seen two poles far from the imaginary axis, so it is kind of
underdamped.

7/26 ECET 221_Lab_3/4

 c) Considering K=2.9, and the calculations from the Table 1: Data Table.

𝐴 = (𝐾)(𝜔0 )2 = 2.9 ∗ 453690000 = 1315701000

𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 2130𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 3
𝑆12 = − 𝜉𝜔0 +/− 𝜔0 ∗ √𝜉 2 − 1 = −1065 +/−21273𝑖

An underdamped system will return to its ideal state more quickly than under a
critical system. However, it is noticeable the overshoot in this case.

ω0 = 21300 𝑟𝑎𝑑 𝜉 = 0.05

𝑑𝑎𝑚𝑝𝑒𝑑 𝑜𝑠𝑐𝑖𝑙𝑙𝑎𝑡𝑖𝑛𝑔 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦 ω𝑑 = √1 − (𝜉)2 ∗ ω0 = √1 − (0.05)2 ∗ 21300
= 21273𝑆1𝑟𝑎𝑑
2𝜋 2𝜋
𝑡𝑟𝑎𝑛𝑠𝑖𝑒𝑛𝑡 𝑡𝑖𝑚𝑒: 𝑡𝑛 = = = 295𝑢𝑠
ω𝑑 21273
𝜋 𝜋
𝑝𝑒𝑎𝑘 𝑡𝑖𝑚𝑒: 𝑡𝑝 = = = 148𝑢𝑠
ω𝑑 21273
𝜉 0.05
−( )𝜋 −( )𝜋
√1−𝜉 2 2
% 𝑜𝑣𝑒𝑟𝑠ℎ𝑜𝑜𝑡 = 100𝑒 = 100𝑒 √1−0.05 = 85%
5
𝑠𝑒𝑡𝑡𝑙𝑖𝑛𝑔 𝑡𝑖𝑚𝑒: 𝑡𝑠 = = 4.69𝑚𝑠
𝜉ω0

𝑆2

Figure 5: underdamped

Similar to the last case, however, this underdamped is extremely.

d) Considering K=3, and the calculations from the Table 1: Data Table.

𝐴 = (𝐾)(𝜔0 )2 = 2.9 ∗ 453690000 = 1814760000

𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 0𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 4
𝑆12 = − 𝜉𝜔0 +/− 𝜔0 ∗ √𝜉 2 − 1 = +/−21300𝑖

8/26 ECET 221_Lab_3/4

It can be seen that the coefficient is zero, this generates a constant sine
amplitude. Again, it does not show overshoot. This considerable marginal stable
or oscillator.

Oscillator
(ξ)=0

Figure 6: Marginal stable

These poles are almost to pass the imaginary axis, so closed to be unstable.
e) Considering K=4, and the calculations from the Table 1: Data Table.

𝐴 = (𝐾)(𝜔0 )2 = 2.9 ∗ 453690000 = 1361070000

𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 − 21300𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 5
𝑆12 = − 𝜉𝜔0 +/− 𝜔0 ∗ √𝜉 2 − 1 = 10650 +/−18446𝑖

This case is instable, the poles are in the right side.

Unstable

Figure 7: Unstable

9/26 ECET 221_Lab_3/4

Part II: MultiSim Simulation
NI MultiSim software allows you to draw the circuit diagram and simulate it. Build
the Sallen-Key circuit.

You may use an input positive going square wave of 1Vp@100Hz.

Calculation of resistances for R4.

I made the simulation from the figure 1:
Sallen Key 2nd Order
C1

XSC1
0.01µF
VCC
Ext Trig
15V +
_
U1 A B
7
5
1

R1 R2 + _ + _
3

4.7kΩ 4.7kΩ C2 741

6
2
XFG1
0.01µF
4

COM VEE
-15V R4
R3
9500Ω
4.7kΩ

Figure 8: Circuit from Multiism

We will explore the 5 cases for K mentioned above:

• For each of the values of K listed above, calculate R4 and simulate the
circuit to confirm the corresponding behavior. Run the simulation and
produce a plot of Vout and Vin versus time for each case. Record the
graph for your documentation(report).

This data calculated from the Table 1: Data Table.

1- K=1.0, R4=…0.00kΩ…. Behaviour:……Critically damped…

2- K=1.5, R4=…2.35kΩ …. Behaviour:……Underdamped……..
3- K=2.9, R4=…8.93kΩ …. Behaviour:……Underdamped……..
4- K=3.0, R4=…9.40kΩ …. Behaviour:……Marginally Stable….
5- K=4.0, R4=…14.10kΩ …. Behaviour:……Unstable……………

10/26 ECET 221_Lab_3/4

VALUE 1
The value is shown in the Table 1, but the calculation is the following with K=1:
𝑅4
𝐾 = 1+
𝑅
𝑅4 = 𝑅(𝐾 − 1) = 0Ω

VALUE 2
The value is shown in the Table 1, but the calculation is the following with K=1.5:
𝑅4 = 𝑅(𝐾 − 1) = 2350Ω

VALUE 3
The value is shown in the Table 1, but the calculation is the following with K=2.9:
𝑅4 = 𝑅(𝐾 − 1) = 8930Ω

VALUE 4
The value is shown in the Table 1, but the calculation is the following with K=3:
𝑅4 = 𝑅(𝐾 − 1) = 9400Ω

VALUE 5
The value is shown in the Table 1, but the calculation is the following with K=4:
𝑅4 = 𝑅(𝐾 − 1) = 14100Ω

Simulation Multisim
The following graphs will help me for the experiment:

Graph 1 (connected by wire)

Figure 9: Critically Damped

11/26 ECET 221_Lab_3/4

Graph 2 (2200Ω+150Ω)

Figure 10: underdamped

Graph 3 (6800Ω+2200Ω)

Figure 11: underdamped

12/26 ECET 221_Lab_3/4

Graph 4 (6800Ω+2200Ω+560Ω)

Figure 12: Marginal stable

Graph 5 (6800Ω+6800Ω+470Ω)

Figure 13: Unstable

• Why do the Vout curves have different steady state values?

This happened because of the change of the resistor R4, this generates
different kinds of gains as the voltage divider in the feedback. The function
of the low pass filter causes different kind of responses.
• For the oscillatory behavior. record the frequency of oscillation. Compare it
1
with fc = 2𝜋𝑅𝐶.
This happened in the case4:

13/26 ECET 221_Lab_3/4

 𝑅4 = 9400Ω

From the figure 14, I calculated the frequency of oscillation:

1 1
𝐹= = = 3060.87𝐻𝑧
𝑇 326.705𝑚𝑠

Figure 14: Period of the oscillation response.

Finally, the calculation of fc:

1 1
𝑓𝑐 = = = 3386.3𝐻𝑧
2𝜋𝑅𝐶 2𝜋 ∗ 4700 ∗ 0.01𝑢𝑓
So, the values are similar, 𝐹 = 𝐹𝑐
I can say that the frequency is determined by R and C. The values of the
resistors does not change anything in the frequency.

• Do the curves support the damping factor calculations you performed

previously?
Yes, the behavior is congruent, take reference table 1. It can be seen that
with different values of resistor R4 the damping coefficient changed. This
means that the poles start moving and gave me the responses.

14/26 ECET 221_Lab_3/4

• Produce an expanded plot for K=2.9, so that % overshoot and settling time
may be measured. Measure % overshoot and settling time on the graph (as
per the definitions given in the introduction).

Figure 15: underdamping response at K=2.9

I had the following calculations:

8.3 − 3.2
%𝑂𝑆 = ( ) ∗ 100 = 85%
3.2 + 2.8
𝑆𝑒𝑡𝑡𝑙𝑖𝑛𝑔 𝑡𝑖𝑚𝑒 = 1.2249 − 1.22 = 0.0049 = 4.9𝑚𝑠

• Using the formulas given in the introduction, determine the theoretical values
for % overshoot and settling time for K=2.9. Do they agree with your
measurements?
−𝜋𝜉 −𝜋(0.05)
√1−𝜉 2 2
%𝑂𝑆 = 𝑒 = 𝑒 √1−0.05 = 85%
5 5
𝑆𝑒𝑡𝑡𝑙𝑖𝑛𝑔 𝑡𝑖𝑚𝑒 = = = 4.69𝑚𝑠
𝜉 𝜔0 0.05 ∗ 21.3𝑘

*increase resolution by decreasing max time step exp(-5)

15/26 ECET 221_Lab_3/4

Part III: Matlab Simulation
Matlab is a convenient tool which allows you to analyze systems for which you
already know the transfer function.
2
𝐾𝜔𝑛
𝐻(𝑠) = 2
𝑠2 + 2 𝜁 𝜔𝑛 𝑠 + 𝜔𝑛

• For each of the following values of K, use Matlab to enter the transfer function
for the Sallen and Key filter and obtain (and plot) a step response:

(a) K=1 (b) K=1.5 (c) K=2.9 (d) K=3 (e) K=4

If t = (as2 + bs +c) / (ds2 + es + f)

The MATLAB commands will be:

t=tf([a b c],[d e f]);

step(t)
Example K=1 means R4=0 and ξ =1 (C.D. behavior) so
2
𝜔𝑛 𝑎
𝐻(𝑠) = 2 =d 𝑠2 + 𝑒
𝑠 2 + 2 𝜔𝑛 𝑠 + 𝜔𝑛 𝑛𝑠 + 𝑓
>> R=..;
>> C=..;
>> w=1/(R*C);
>>W=1/(R*C)^2 ;
>> K=1;
>>z=(3-K)/2;
>> t=tf([W ],[1 2*z*w W]);
>> step(t)
Try
>>impulse(t) /*response tells you about stability*/

• Confirm that this Laplace analysis supports your simulation results from Part
II, including measurements of %overshoot and settling time for K=2.9.
Yes, it was demonstrated the values for overshoot and the values for those
values of K.
%𝑂𝑆 = 85% 𝑆𝑒𝑡𝑡𝑙𝑖𝑛𝑔 𝑡𝑖𝑚𝑒 = 4.69𝑚𝑠
• Account for any differences between the Matlab and PSpice results.

16/26 ECET 221_Lab_3/4

Simulation Matlab
Part a K=1
Regarding 𝑆 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 => 𝑆 2 + 42600𝑆 +
2

453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 1 & Table 1.

t=tf([453690000 ],[1 42600 453690000]);

Figure 16: Step Response of Critically damped.

17/26 ECET 221_Lab_3/4

 Figure 17: Impulse Response of Critically damped.

Part b K=1.5
Regarding 𝑆 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 31950𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 2
2

& Table 1.
t=tf([680535000 ],[1 31950 453690000]);

Figure 18: Step Response of Underdamped.

18/26 ECET 221_Lab_3/4

 Figure 19: Impulse Response of Underdamped.

Part c K=2.9
Regarding 𝑆 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 2130𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 3
2

& Table 1.
t=tf([1315701000 ],[1 2130 453690000]);

Figure 20: Step Response of Underdamped.

19/26 ECET 221_Lab_3/4

 Figure 21: Impulse Response of Underdamped.

Part d K=3
Regarding 𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 0𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 4 &
Table 1.
>> t=tf([1361070000 ],[1 0 453690000]); this case the coefficient of S is 0.

Figure 22: Step Response of Marginal Stable.

20/26 ECET 221_Lab_3/4

 Figure 23: Impulse Response of Marginal Stable.

Part e K=4
Regarding 𝑆 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 0𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 4 &
2

Table 1.
>> t=tf([1814760000 ],[1 -21300 453690000]);

Figure 24: Step Response of Unstable.

21/26 ECET 221_Lab_3/4

 Figure 25: Impulse Response of Unstable.

Part IV: Building the Circuit

Construct the Sallen and Key filter in the lab using an LF 411 op amp and a 20K
pot for R4. Drive the circuit with a suitable positive-going square wave (it is
convenient to use the same input settings as you used in Pspice) and observe
how the output compares to the input.

• Vary the value of R4 to obtain each of the values of gain K listed above.
• Sketch the display of output and input for each value of K.
• Are the scope displays what you expected? Account for any differences from
your previous analyses (e.g. %overshoot, settling time, steady state output).

Figure 26: Physical circuit

22/26 ECET 221_Lab_3/4

It was decided not to use the potentiometer of 20kΩ, because of the precision.
That is why I used the resistance values from Part II: MultiSim Simulation and
confirm the simulations for each case.

Part a K=1 (connected by wire)

Figure 27: Critically Damped similar to figure 9.

Part b K=1.5 (2200Ω+150Ω)

Figure 28: underdamped similar to figure 10.

Part c K=2.9 (6800Ω+2200Ω+560Ω)

23/26 ECET 221_Lab_3/4

 Figure 29: underdamped similar to figure 11.

Part d K=3 (6800Ω+2200Ω+560Ω)

Figure 30: Marginal stable similar to figure 12

Part e K=4 (6800Ω+6800Ω+470Ω)

24/26 ECET 221_Lab_3/4

 Figure 31: Unstable

Part V: Conclusiones
This part in the laboratory, I will give all the conclusions from the experiment:

1. In this laboratory, it was demonstrated each response of control system of

second grade Sallen Key in the theory, Multisim, and MATLAB.
2. The behavior for each response, it depends on the change of the
resistance R4 and therefore of K and damping factor (𝜉). Referred to the
table 1.
𝜉 = 1 𝑐𝑟𝑖𝑡𝑖𝑐𝑎𝑙𝑙𝑦 𝑑𝑎𝑚𝑝𝑒𝑑 𝑓𝑜𝑟 𝐾 = 1
0 < 𝜉 < 1 𝑢𝑛𝑑𝑒𝑟𝑑𝑎𝑚𝑝𝑒𝑑 𝑓𝑜𝑟 𝐾 = 1.5 & 𝑘 = 2.9
𝜉 = 0 𝑚𝑎𝑟𝑔𝑖𝑛𝑎𝑙 𝑠𝑡𝑎𝑏𝑙𝑒 𝑓𝑜𝑟 𝐾 = 3
𝜉 < 0 𝑢𝑛𝑠𝑡𝑎𝑏𝑙𝑒 𝑓𝑜𝑟 𝐾 = 4
3. It was demonstrated the formula of the transfer function by the theorical
formulas at Part 1.
𝑉𝑜𝑢𝑡 (𝐾)(𝜔0 )2
= 2
𝑉𝑖𝑛 𝑆 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2
4. The table 1 was used to the calculations for each response.
5. In each value of K, I had equation:
𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 => 𝑆 2 + 42600𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 1(K=1)
𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 31950𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 2(K=1.5)
𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 2130𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 3(K=2.9)

25/26 ECET 221_Lab_3/4

𝑆 2 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 2130𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 3𝑆 2 + 2 ∗ 𝜉 ∗
𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 + 0𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 4(K=3)
𝑆 + 2 ∗ 𝜉 ∗ 𝜔0 ∗ 𝑆 + (𝜔0 )2 = 𝑆 2 − 21300𝑆 + 453690000 … 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 5(K=4)
2

6. The poles obtained were:

𝑆12 = −21300 … from 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 𝟏
𝑆12 = −15975 +/−14089𝑖 … from 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 𝟐
𝑆12 = −1065 +/−21273𝑖 … from 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 𝟑
𝑆12 = +/−21300𝑖 … from 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 𝟒
𝑆12 = 10650 +/−18446𝑖 … from 𝑬𝒒𝒖𝒂𝒕𝒊𝒐𝒏 𝟓
The behavior of the poles is showed in the figures 3, 4, 5, 6, and 7.
7. The parameters were set at the equations section at Part 1: Theory.
8. When K=1.5, it is underdamped, but this case is kind of underdamped
because there are far from the imaginary axe. On the other hand, when
K=2.9, this extremely underdamped closed to be marginal stable. Moreover,
it is noticeable the over shooter is bigger when it’s closed to the imaginary
axe.
9. There weren’t overshooter for critically damped, marginal stable or unstable.
10. Regarding table 1, I obtained the following values of R4:
K=1 R4=0 Ω
K=1.5 R4=2350 Ω
K=2.9 R4=8930 Ω
K=3 R4=9400 Ω
K=4 R4=14100 Ω
However, it was taken approximate resistors with help of the commercial
resistors show in the figures 9, 10, 11, 13, and 13.
11. With help of the simulations of Multisim, I had an idea of the responses I will
have at the physical part of the experiment. Moreover, it was demonstrated
the parameters for an underdamped control system like overshooter and
settling time. This is shown in figure 15.
12. Considering the equations 1, 2, 3, 4, and 5. It was run the commands for the
MATLAB part. It was shown the response from a step and an impulse. It can
be seen at figures 16, 17, 18, 19, 20, 21, 22, 23, 24, and 25.
13. I found similitude between:
o Figure 9 with figure 16.
o Figure 10 with figure 18.
o Figure 11 with figure 20.
The next cases are slightly different because MATLAB only was inserted a
step:
o Figure 12 with figure 22.
o Figure 13 with figure 24.
This confirms the similitude of K and factor damping.
14. Then, it was the hout of the physical experiment:
I can also demonstrate that my answer are correct because I found the
following similitudes.
o Figure 9 with figure 27.
o Figure 10 with figure 28.

26/26 ECET 221_Lab_3/4

 o Figure 11 with figure 29.
o Figure 12 with figure 30.
The next case is slightly different because of rail voltages of the op amp.
o Figure 13 with figure 31.
15. As I mentioned before, in the experiment, I must use commercial resistors as:
K=1 R4=0 Ω (connected by wire)
K=1.5 R4=2350 Ω (2200Ω+150Ω)
K=2.9 R4=8930 Ω (6800Ω+2200Ω)
K=3 R4=9400 Ω (6800Ω+2200Ω+560Ω)
K=4 R4=14100 Ω (6800Ω+6800Ω+470Ω)
16. In the experiment, there errors presented because of the errors of the
components like resistors and capacitors. Nevertheless, the capacitors were
the main part in the behavior of the responses.
17. In the Graph 4, I have to use the resistor 560 Ω in (6800Ω+2200Ω+560Ω),
because 470Ω did not work!

27/26 ECET 221_Lab_3/4