LEP

1.3.27
-11

Forced Oscillations Pohls pendulum

Determination of resonance frequencies by Fourier analysis

Related topics
Angular frequency, characteristic frequency, resonance frequency, torsion pendulum, torsional vibration, torque, restoring torque, damped/undamped free oscillation, forced oscillation, ratio of attenuation/decrement, damping constant, logarithmic decrement, aperiodic case, creeping.
Principle
If an oscillating system is allowed to swing freely it is observed
that the decrease of successive maximum amplitudes is highly dependent on the damping. If the oscillating system is stimulated to swing by an external periodic torque, we observe that
in the steady state the amplitude is a function of the frequency and the amplitude of the external periodic torque and of the
damping. The characteristic frequencies of the free oscillation
as well as the resonance curves of the forced oscillation for
different damping values are to be determined.
Therefore, the oscillations are recorded with the Cobra3 systemin connection with the movement sensor. The curves of
the different oscillations are displayed and the necessary
quantities for the determination of the characteristic values
can easily becalculated.
Equipment
Torsion pendulum after Pohl
Power supply, universal
Bridge rectifier, 30 V AC/1 A DC
Stopwatch, digital, 1/100 sec.
Digital multimeter
Connecting cord, l = 250 mm, yellow

11214.00
13500.93
06031.10
03071.01
07134.00
07360.02

Connecting cord, l = 750 mm, red

Connecting cord, l = 750 mm, blue
Cobra3 Basic Unit
Power supply, 12 VRS 232 data cable
Cobra3 Translation/Rotation Software
Movement sensor with cable
Adapter, BNC-socket/4mm plug pair
Adapter, BNC-socket-plug, 4 mm
Silk thread, l = 200 m
Weight holder, 1 g
Tripod base PASS
Support rod PASS, l = 400 mm
Support rod PASS, l = 250 mm
Right angle clamp PASS
PC, Windows 95 or higher

07362.01
07362.04
12150.00
12151.99
14602.00
14512.61
12004.10
07542.27
07542.20
02412.00
02407.00
02005.55
02026.55
02025.55
02040.55

2
3
1
1
1
1
1
1
1
1
1
1
1
1
2

Tasks
A. Free oscillation

1
1
1
1
1
2

1. To determine the oscillating period and the characteristic frequency of the undamped case.
2. To determine the oscillating periods and the corresponding characteristic frequencies for different damping values. Successive, unidirectional maximum amplitudes are to be plotted as a function of time. The corresponding ratios of attenuation, the damping constants
and the logarithmic decrements are to be calculated.
3. To realize the aperiodic case and the creeping.

Fig. 1a. Experimental set-up for free and forced torsional vibration

PHYWE series of publications Laboratory Experiments Physics PHYWE SYSTEME GMBH & Co. KG D-37070 Gttingen

21327-11

LEP
1.3.27
-11

Forced Oscillations Pohls pendulum

Determination of resonance frequencies by Fourier analysis

B. Forced oscillation
1. The resonance curves are to be determined and to be
represented graphically using the damping values of A.
2. The resonance frequencies are to be determined and
are to be compared with the resonance frequency values found beforehand.

Use a paper-fastener or the weight holder and hang it at the

end of the thread to tighten it. So, the thread runs parallel and
always leaves the pendulum perpendicular.

Fig. 2. Connection of the movement sensor to the Cobra3

Basic Unit

3. The phase shifting between the torsion pendulum and

the stimulating external torque is to be observed for a
small damping value assuming that in one case the
stimulating frequency is far below the resonance frequency and in the other case it is far above it.
Fig. 1b: Electrical connection of the experiment.

red
black
yellow
BNC1
BNC2

Procedure
FREE OSCILLATION UNDAMPED CASE
To determine the characteristic frequency v0 of the torsion
pendulum without damping (IB = 0), the time for several oscillations
is measured repeatedly and the mean value of the peri
od T0 calculated.

Set-up
The experiment is set up as shown in Fig. 1a und 1b. The DC
output U_ of the power supply unit is connected to the two
upper sockets of the DC motor. The eddy current brake also
needs DC voltage. For this reason a rectifier is inserted
between the AC output U_ of the power supply unit and the
entrance to the eddy current brake. The DC current supplied
to the eddy current brake, IB, is indicated by the ampmeter.
Perform the electrical connection of the movement sensor to
the Cobra3 Basic Unit according to Fig. 2.

Set the measuring parameters in measure according to Fig. 3.

Select 12 mm for the axle diameter of the movement sensor.
The axis diameter in the Rotation menu item is twice the distance from the pivot point of the pendulum to the attachment
point of the silk thread that runs to the movement sensor, i.e.
here about 180 mm.

Fig. 3. Measuring parameters

To obtain a connection between the movement sensor and the

pendulum do as follows:
Place the movement sensor at the end of the table to ensure
that the used thread can easily swing.
Take about 100 cm of thread (dependent on the distance
between pendulum and sensor), place it in the groove of the
copper-pendulum, rotate the pendulum half and fix one end
with some adhesive tape.
Note: Making a knot at the end of the thread helps keeping it
fixed with the tape.
Make sure, that the tape lies close on the copper-disc.
Take the other end of the thread and place it across the larger of the two cord grooves on the movement sensor. Wound it
around the cord groove once.

21327-11

PHYWE series of publications Laboratory Experiments Physics PHYWE SYSTEME GMBH & Co. KG D-37070 Gttingen

Forced Oscillations Pohls pendulum

Determination of resonance frequencies by Fourier analysis

LEP
1.3.27
-11

Chose the icon continue to enter the measurement window.

Here, the actual value of the movement sensor is displayed.
Set the pendulum in motion (oscillation amplitude up to 15
scale divisions) and click on the Start measurement icon.

If the values (50 ms) in the Get value every (50) ms (Fig.3)
dialog box are too high or too low, noisy or non-uniform measurements can occur. In this case adjust the measurement
sampling rate appropriately.

After approximately 5-10 oscillations click on the Stop measurement icon.

There are different opportunities to determine the oscillation

period and the characteristic frequency of the undamped
case.
The period duration can either be calculated with the aid of the
cursor lines, which can be freely moved and shifted onto the
adjacent maxima or minima of the oscillation curve (Fig. 5).
Alternatively, use the item curve analysis in the analysis
menu, where the extrema of the chosen curve can be calculated.
Or use Fourier analysis, where the peak displays directly the
period duration of the oscillation.

Fig.4 shows the recorded curve of this undamped oscillation.

FREE OSCILLATION

DAMPED CASE

In the same way the characteristic frequencies for the damped

oscillations are found.
Rebuild the experimental set-up as described above, now
using the following current intensities for the eddy current
brake (controlled by the ammeter):
Table for the current values for the eddy current brake:
Remarks
Initially, it has to be ensured that the pendulum pointer at rest
coincides with the zero-position of the scale. This can be
achieved by turning the excentric disc of the motor.

IB
IB
IB
IB
IB

~
~
~
~
~

0.16
0.34
0.52
0.70
0.88

A,
A,
A,
A,
A,

(U~
(U~
(U~
(U~
(U~

=
=
=
=
=

2 V)
4 V)
6 V)
8 V)
12 V)

Try to rotate the pendulum carefully and release it quickly without friction of your hand. This avoids disturbances of the
thread and damping of the pendulum, which can lead to measurement errors.
For amplitudes up to 13 scale divisions the thread is well tightened.

To realize the aperiodic case (IB ~ 1.5 A) and the creeping

Case (IB ~ 1.7 A) the eddy current brake is briefly connected
directly to the DC output U_ of the power supply unit.

Fig. 5: Calculation of the period duration with cursor lines

Fig. 6: Recorded curve of the damped oscillation

Fig. 6 shows the curve of the damped case with IB ~ 0.25 A.

PHYWE series of publications Laboratory Experiments Physics PHYWE SYSTEME GMBH & Co. KG D-37070 Gttingen

21327-11

LEP
1.3.27
-11

Forced Oscillations Pohls pendulum

Determination of resonance frequencies by Fourier analysis

The Calculation of the oscillating periods is the same as

described above. To evaluate the characteristic damping values, determine the magnitude of successive amplitudes on
the same side (i.e. either minima or maxima). Therefore, use
the cursor lines, again or chose the curve-analysis function in
the analysis menu. Here, you can calculate the extrema of the
curve and use them directly for the determination of the
damping ratio (Fig.7).

Forced oscillation
To stimulate the torsion pendulum, the connecting rod of the
motor is fixed to the upper third of the stimulating source. The
DC voltage U_ of the power supply unit must be set to maximum.
The stimulating frequency va of the motor can be found by
using a stopwatch and counting the number of turns.
The amplitudes of the forced oscillation are recorded in the
same way as for the free oscillations. The measurement
begins with small frequencies. va is increased by means of the
motor-potentiometer setting coarse. In the vincinity of the
maximum amplitude in the resonance case va is changed in
small steps using the potentiometer setting fine. In each
case, only values should be taken into account after a stable
pendulum amplitude has been established. The higher the
damping values are the faster this steady state is reached.

Fig. 8 shows the measured curve near the resonance frequency for a small damping value (IB ~ 0.16 A).

Remarks
The weight hanging on the rotational axis of the movement
sensor should not exceed 2 g to avoid greater influence on
the pendulum system.
In extremely short oscillation periods, signal transients or
deformations can occur. These can be reduced if the sampling rate is changed. In any case, error-free recorded
intervals can be selected from the measuring signal after
completion of the measurements.
Sickle-shaped deformation of the oscillations are due to
slippage of the thread across the cord groove on the
movement sensor. This is avoided if the thread is wound
around the cord groove once.
Since the movement recording is not performed without
contact, slight damping of the measured oscillations does
occur, but the difference between nearly undamped and
damped oscillations is very significant.
Theory and evaluation
A. Undamped and damped free oscillation In case of free and
damped torsional vibration torques M1 (spiral spring) and M2
(eddy current brake) act on the pendulum. We have
M1 = D0 f and M2 = Cf
f

f
D0
C

=
=
=
=

angle of rotation
angular velocity
torque per unit angle
factor of proportionality depending on the current
which supplies the eddy current brake

The resultant torque

In the absence of damping or for only very small damping values, va must be chosen in such a way that the pendulum does
not exceed its scale range (resonance catastrophe).
Therefore, in the resonance regime the measurement has to
be stopped before the maximum values are reached.
Again, the maximum oscillation amplitudes can be calculated
with the aid of the cursor lines or with the curve analysis.
The resonance frequencies can be determined by Fourier
analysis.

M = D0f Cf
leads us to the following equation of motion:
..
If + C = f + D0f = 0

(1)

I.. = pendulums moment of inertia

f = angular acceleration

Fig. 8: Forced oscillations for a small damping value

Fig. 7: Curve analysis parameters

21327-11

PHYWE series of publications Laboratory Experiments Physics PHYWE SYSTEME GMBH & Co. KG D-37070 Gttingen

Forced Oscillations Pohls pendulum

Determination of resonance frequencies by Fourier analysis

where

Dividing Eq. (1) by I and using the abbreviations

d=

C
D0
and v20 =
2I
I

fa 

results in
..
f + 2 df + v20f = 0

f0
va 2 2
d va 2
e1  c d f  c2
d
v
v
A
0
0 v0

(9)

(2)
and f0 =

d is called the damping constant and

v0 =

LEP
1.3.27
-11

D0
B I

F0
v20

Furthermore:

the characteristic frequency of the undamped system.

tan a =

2 dva
v20 v2a

The solution of the differential equation (2) is

respectively
f(t) = f0e-dt cos vt

(3)
a = arc tan

with
v = 2v20  d2

(4)

Eq. (3) shows that the amplitude f(t) of the damped oscillation
has decreased to the e-th part of the initial amplitude f0 after
the time t = 1/6 has elapsed. Moreover, from Eq. (3) it follows
that the ratio of two successive amplitudes is constant.
fn
= K = edT
fn1

(5)

2 dva

(10)

v20 v2a

An analysis of Eq. (9) gives evidence of the following:

1. The greater F0 , the greater fa
2. For a fixed value F0 we have:
f S fmax for va  v0
3. The greater d, the smaller fa
4. For d = 0 we find:
fa S if va = v0

K is called the damping ratio and the quantity

- = ln K = dT = ln

fn
fn1

(6)

is called the logarithmic decrement. Eq. (4) has a real solution only if
v20 x d2.
v20

For
= d , the pendulum returns in a minimum of time to its
initial position without oscillating (aperiodic case). For v20 < d2,
the pendulum returns asymptotically to its initial position
(creeping).

Fig.9: Phase shifting of forced oscillation for different damping

values

B. Forced oscillation
If the pendulum is acted on by a periodic torque Ma =
M0 cos vat Eq. (2) changes into
..
f + 2 df + v20f = F0 cos vat

where F0 =

(7)

M0
I

In the steady state, the solution of this differential equation is

f(t)  fa cos (vat a)

(8)

PHYWE series of publications Laboratory Experiments Physics PHYWE SYSTEME GMBH & Co. KG D-37070 Gttingen

21327-11

LEP
1.3.27
-11

Forced Oscillations Pohls pendulum

Determination of resonance frequencies by Fourier analysis

Results
Typical values for this experimental set-up are:

For very small frequencies va the phase difference is approximately zero, i.e. the pendulum and the stimulating torque are
inphase. If va is much greater than v0, pendulum and stimulating torque are nearly in opposite phase to each other.

T0 = (1.75 0.05) sec;

= (3.59 0.10) sec1
v
0
I

1/d

0.16

11.1

0.09

3.58

1.17

0.16

0.34

3.8

0.26

3.57

1.58

0.46

0.52

2.0

0.51

3.54

2.47

0.91

0.70

1.2

0.85

3.50

4.60

1.53

0.88

0.8

1.30

3.42

10.89

2.39

Fig. 9 shows the phase difference of the forced oscillation as

a function of the stimulating frequency according to Eq. (10).

Fig. 10 illustrates the decrease in the unidirectional amplitude

values as a function of time for free oscillation and different
damping values.
Fig. 11 shows the resonance curves for different damping values.
Evaluating the curves leads to medium resonance frequency
v = 3.59 s1.
Furthermore, the software allows to calculate the Fourierspectra of the recorded oscillations. With this opportunity it
can easily be shown that for increasing damping values the
frequency of maximum amplitude shifts to smaller values (see
Fig.12).

Fig.10: Maximum values of unidirectional amplitudes as a function of time for different damping values.

21327-11

PHYWE series of publications Laboratory Experiments Physics PHYWE SYSTEME GMBH & Co. KG D-37070 Gttingen

Forced Oscillations Pohls pendulum

Determination of resonance frequencies by Fourier analysis

LEP
1.3.27
-11

Fig.11: Resonance curves for different damping values.

Fig.12: Fourier-spectra for different damping values, increasing left to right.

PHYWE series of publications Laboratory Experiments Physics PHYWE SYSTEME GMBH & Co. KG D-37070 Gttingen

21327-11

LEP
1.3.27
-11

Forced Oscillations Pohls pendulum

Determination of resonance frequencies by Fourier analysis

21327-11

PHYWE series of publications Laboratory Experiments Physics PHYWE SYSTEME GMBH & Co. KG D-37070 Gttingen