ENGN4625 & ENGN6625
Power Systems and Power Electronics
Design Assignment
Power Supply Design
1 Assignment Deliverable
A well-formatted and laid out report containing details of your power supply design, with
justifications for decisions made, and validated test results including analysis of output
reliability and impacts on power quality.
Maximum of 10 pages, including design calculations, figures, tables, key schematics
and waveform images. Additional waveform images may be included in an appendix and
commented on in the main body of the report.
More details on project report expectations can be found in Section 6.
2 Overview
You work for a consumer electronics design company that is developing a Smart-device
Docking Station capable of providing power to a range of smart devices (phones, tablets,
music players, etc), audio amplifiers, as well as providing an auxiliary power supply. The
docking station is required to be able to plug into any single-phase power outlet anywhere
in the world, and should be compliant to applicable power quality standards. Your role at
the company is to design and test a power supply system that is capable of supplying the
various parts of the docking station with the necessary quantities of power at the required
voltages. The requirements are detailed in Section 3.
There are four main parts to the project which you should complete: Converter Design,
Stage 1 Converter Build and Test, Stage 2 Converter Build and Test, and then finally
combining the two converter stages together into one Complete Power Supply Circuit and
testing.
3 Power Supply Specifications
The basic power supply system and requirements are represented by the block diagram
shown in Figure 1. You are required to use a standard diode bridge rectifier, and then
design two separate DC-DC Converters (Stage 1 shown in green, Stage 2 in red in the
block diagram). Your company requires that the second (red) DC-DC Converter that
you design be flexible enough such that a single design can be manufactured which sat-
isfies both separate end-use voltage and power requirements. The Stage 2 converter
should be able to be powered via the mains/Stage 1 Converter. In order to help re-
duce manufacturing costs and to keep the product as lightweight as possible, you should
1
design all converters to operate at a switching frequency of at least fsw kHz where
fsw = (30 +
∑
(your numeric UNI− ID numbers)) kHz 1, and avoid using excessively
large components. It makes sure that there isn’t a single solution for the design. So it is
not like doing an assignment with a fixed solution (like early year AQF Level 7 work). It
is closer to a real-world design requirement. Everyone will end up with a different design.
Figure 1: Power supply block diagram.
4 Project Requirements
Part 1 - Converter Design
Rectifier/Stage 1 DC-DC Converter: Start by designing the Rectifier and Stage 1 DC-DC
converter. Unlike the rectifiers mostly discussed in the notes, you won’t have a large enough
load to enable inductive filtering on its own to smooth the output (without using a very
large inductor). And like the hardware labs, if you use too large a capacitive filter it will
result in large amount of harmonic distortion on the input side. You should instead allow
the rectifier output to have some capacitive filtering but still fluctuate over a reasonable
range, and rely upon the Stage 1 DC-DC converter to provide suitable regulation and a
constant 36 V output.
Stage 2 DC-DC Converter: You need to come up with a single converter design that
will work for both loads. The converter should be operated at least fsw kHz, although it
does not have to be the same switching frequency as the stage 1 rectifier.
1For example if your UNI-ID is u5664317, then switching frequency will be 30+(5+6+6+4+3+1+7) =
62 kHz.
2
Part 2 - Implement and Test Stage 2 Converters
Use LTSpice to build the Stage 2 converter with open loop control (unregulated), and
then test the converter for all three load types and load sizes. Verify that the converter
produces the required outputs, stays in continuous conduction. Measure the input power
required in each load scenario and measure converter efficiency under a range of different
load scenarios. Optional: Implement closed loop/feedback PWM control to automatically
manage the duty cycle to achieve the required output. Note: it can be harder to get this
right than for the Stage 1 converter - in fact it is recommended to do stage 1 converter
feedback control first.
Part 3 - Implement and Test Rectifier and Stage 1 Converter
Use LTSpice to build the Rectifier and Stage 1 converter, complete with closed loop PWM
control of the DC-DC converter section (regulated supply). Validate that the circuit works
for both maximum and minimum output loads and for the maximum and minimum AC
voltages that it can experience. Check that the converter produces the required 24V output,
that it stays in continuous conduction and also that there are no high current spikes in the
circuit. Measure the input power required in each load/input scenario, identifying where
the losses are, and calculate the Rectifier/Converter efficiency. Observe and comment on
the AC supply current waveform.
Part 4 - Join the Rectifier/Stage 1 converter and three Stage 2
converters and test
Connect the Rectifier/Stage 1 converter to two copies of the Stage 2 converter and test
under all real conditions. Comment particularly on whether you are still able to achieve
the desired output voltages and low ripple, and if required what you think might do to
address that. Calculate the combined/overall efficiency of supply to the loads.
Comment on what you observe at start-up, that is when you first plug in the power
supply and all output voltages are zero. Do some research and see what can be done in
practice to address this.
5 Notes/hints on LTSpice Components
Diode: You may pick any available diode, as long as it can handle a large enough forward
current and withstand a large enough reverse bias voltage. The MUR460 is recommended.
Inductor: Use the standard inductor, but to make it more realistic include in series
with it (or define in the inductor specs) a resistance which you must scale with inductance
value, using 20 mΩ per 100 µH, Figure 2.
Switch: Start with a simple generic Switch model (SW), rather than using a FET cor
IGBT. You won’t then have to worry about maximum current or reverse bias or about
3
ENGN4625/6625 Major Project, 2015
p4/4
Notes / hints on LTSpice Components:
Diode: You may pick any available diode, as long as it can handle a large enough forward
current and withstand a large enough reverse bias voltage. I recommend using MUR460.
Inductor: Use the standard inductor, but to make it more realistic
include in series with it (or define in the inductor specs) a resistance
which you must scale with inductance value, using 20 mΩ per 100 µH.
Switch: I recommend starting with a simple generic Switch model
(SW), rather than using a FET or IGBT. You won’t then have to worry
about maximum current or reverse bias or about switching transients
(or possible simulation hang-ups). To make it a little more realistic
though you should accompany your switch with a 100 mΩ resistor in
series with it (this represents the ‘ON’ resistance of the switch).
PWM control: When it comes to implementing PWM control of your
DC-DC converter you will need to compare the output voltage
against a desired value and then against a triangle waveform. For
similar reasons as for the switch, I recommend to start with that you
use a simple behavioural voltage source (bv) to generate the PWM
pulse train. You can input into this any equation you like... you can
consider this to be a legitimate software implementation of PWM
feedback control.
Node labels: It will help, both for plotting and for defining feedback
equations, if you label some nodes:
Transient simulation time: Make sure that when you measure average values and ripple, and
FFT that you ‘ignore’ the transient start-up component. I recommend for the case of the
Rectifier and stage 1 converter simulating for at least 5 – 10 lots of 50 Hz cycles. For the
Stage 2 converter on its own you will probably not need to simulate for as long.

What your project report should contain:
Your project report should contain an abstract or summary which describes in summary form
what you did: your power supply design, any special features you uses, your observations
and results, comments on any difficulties that need addressing, and which extra bits you
might have attempted and with what result.
Your report should then ideally contain a section for each separate part of the project,
detailing for example the design calculations and choices, schematics from LTspice of your
final circuit, summary of validated results and measurements, include key waveforms and
any unusual observations and what you did to fix them.
You should include a short conclusion or discussion section which wraps it all up.
Figure 2: Realistic inductor model.
switching transients (or possible simulation hang-ups). To make it a little more realistic
though you should accompany your switch with a 100 mΩ resistor in series with it (this
represents the ON resistance of the switch), Figure 3.
ENGN4625/6625 Major Project, 2015
p4/4
Notes / hints on LTSpice Components:
Diode: You may pick any available diode, as long as it can hand e a large enough forward
current and withstand a large enough reverse bias voltage. I recommend using MUR460.
Inductor: Use the standard inductor, but to make it more realistic
include in series with it (or define in the inductor specs) a resistance
which you must scale with inductance value, using 20 mΩ per 100 µH.
Switch: I recommend starting with a simple generic Switch model
(SW), rather than using a FET or IGBT. You won’t then have to worry
about maximum current or reverse bias or about switching transients
(or possible simulation hang-ups). To make it a little more realistic
though you should accompany your switch with a 100 mΩ resistor in
series with it (this represents the ‘ON’ resistance of the switch).
PWM control: When it comes to implementing PWM control of your
DC-DC converter you will need to compare the output voltage
against a desired value and then against a triangle waveform. For
similar reasons as for the switch, I recommend to start with that you
use a simple behavioural voltage source (bv) to generate the PWM
pulse train. You ca input into this any equation you like... you can
consider this to be a legitimate software impleme tation of PWM
feedback control.
Node labels: It will help, both for plotting and for defining feedback
equations, if you label some nodes:
Transient simulation time: Make sure th t when you measure average values and ripple, and
FFT that you ‘ignore’ the transie t start-up component. I recommend for the case of the
Rectifier an stage 1 converter simulating for at le st 5 – 10 lots of 50 Hz cycles. For the
Stag 2 converter o its ow you wil probably not need to simulate for as long.

What your project report should contain:
Your project report should contain an abstract or summary which describes in summary form
what you did: your power supply design, any special features you uses, your observations
and results, comments on any difficulties that need addressing, and which extra bits you
might have attempted and with what result.
Your report should then ideally contain a section for each separate part of the project,
detailing for example the design calculations and choices, schematics from LTspice of your
final circuit, summary of validated results and measurements, include key waveforms and
any unusual observations and what you did to fix them.
You should include a short conclusion or discussion section which wraps it all up.
Figure 3: Realistic switch model.
PWM control: When it comes to implementing PWM control of your DC-DC converter
you will need to compare the output voltage against a desired value and then against a
triangle wavefo m. For similar reasons as for the switch, it is recommended to start with
that you use a simple beh vioural voltage source (bv) to generate the PWM pulse train,
Figure 4. You can input into this any equation you like... you can consider this to be a
legitimate software implementation of PWM feedback control.
ENGN4625/6625 Maj r Project, 2015
p4/4
Notes / hints on LTSpice Components:
Diode: Y u may pick any available diode, as long as it can handle a large enough forward
current and withst nd a large ough reverse bias voltage. I recommend using MUR460.
Ind ctor: Use the standard i ductor, but to make it more realistic
include in series with it (or define in the inductor specs) a resistance
which you must scale with inductance value, using 20 mΩ per 100 µH.
Switch: I recommend starting with a simpl g neric Switch model
(SW), r ther than using a FET or IGBT. You won’t then hav to w rry
about maximum current or reverse bias or about swi ching tr nsients
(or po sible simulati n hang-ups). To make it a little more r alistic
though you should accompany your switch with a 100 Ω r sistor in
series wit it (this represents the ‘ON’ resistanc of the switch).
PWM control: When it comes to implementing PWM control of your
DC-DC converter you will need to compare the output voltage
against a desired value and then against a triangle waveform. For
similar reasons as for the switch, I recommend to start with that you
use a simple behavioural voltage source (bv) to generate the PWM
pulse train. You can input into this any equation you like... you can
consider this to be a legitimate software implementation of PWM
feedback control.
Node labels: It will help, both for plotting and for defining feedback
equations, if you label some nodes:
Transient simulation time: M ke sure that when yo measure average values and ripple, and
FFT that you ‘ignore’ the transient start-up c mpon nt. I recommend for the case of the
Rectifier and stage 1 converter simulating for at least 5 – 10 lots of 50 Hz cycles. For the
Stage 2 converter on its own you will probably not need to simulate for as long.

What your project report should contain:
Your project report sh uld contain an abstract or summary which describes in summary form
what you did: your power supply design, any special features you uses, your observations
and results, comments on any difficulties that need addressing, and which extra bits you
might have attempted and with what result.
Your report should then ideally contain a section for each separate part of the project,
detailing for example the design calculations and choices, schematics from LTspice of your
final circuit, summary of validated results and measurements, include key waveforms and
any unusual observations and what you did to fix them.
You should include a short conclusion or discussion section which wraps it all up.
Figure 4: Behavioural switch.
Node labels: It will help, both for plotting and for defining feedback equations, if you
label some nodes, Figure 5.
Transient simulation time: Make sure that when you measure average values and ripple,
and FFT that you ign re the transient start-up component. For the case of the Rectifier
a d stage 1 converter simulating for at ast 5 - 10 lots of 50 Hz cycles is recommended.
F r the Stage 2 converter on its own you will probably not need to simulate for as long.
4
ENGN4625/6625 Major Project, 2015
p4/4
Notes / hints on LTSpice Components:
Diode: You may pick any available diode, as long as it can handle a large enough forward
current and withstand a large enough reverse bias voltage. I recommend using MUR460.
Inductor: Use the standard inductor, but to make it more realistic
include in series with it (or define in the inductor specs) a resistance
which you must scale with inductance value, using 20 mΩ per 100 µH.
Switch: I recommend starting with a simple generic Switch model
(SW), rather than using a FET or IGBT. You won’t then have to worry
about maximum current or reverse bias or about switching transients
(or possible simulation hang-ups). To make it a little more realistic
though you should accompany your switch with a 100 mΩ resistor in
series with it (this represents the ‘ON’ resistance of the switch).
PWM control: When it comes to implementing PWM control of your
DC-DC converter you will need to compare the output voltage
against a desired value and then against a triangle waveform. For
similar reasons as for the switch, I recommend to start with that you
use a simple behavioural voltage source (bv) to generate the PWM
pulse train. You can input into this any equation you like... you can
consider this to be a legitimate software implementation of PWM
feedback control.
Node labels: It will help, both for plotting and for defining feedback
equations, if you label some nodes:
Transient simulation time: Make sure that when you measure average values and ripple, and
FFT that you ‘ignore’ the transient start-up component. I recommend for the case of the
Rectifier and stage 1 converter simulating for at least 5 – 10 lots of 50 Hz cycles. For the
Stage 2 converter on its own you will probably not need to simulate for as long.

What your project report should contain:
Your project report should contain an abstract or summary which describes in summary form
what you did: your power supply design, any special features you uses, your observations
and results, comments on any difficulties that need addressing, and which extra bits you
might have attempted and with what result.
Your report should then ideally contain a section for each separate part of the project,
detailing for example the design calculations and choices, schematics from LTspice of your
final circuit, summary of validated results and measurements, include key waveforms and
any unusual observations and what you did to fix them.
You should include a short conclusion or discussion section which wraps it all up.
Figure 5: Node label.
6 What your pr ject report should contain
Your project report should contain an abstract or summary which describes in summary
form what you did: your power supply design, any special features you uses, your observa-
tions and results, comments on any difficulties that need addressing, and which extra bits
you might have attempted and with what result. Your report should then ideally contain a
section for each separate part of the project, detailing for example the design calculations
and choices, schematics from LTspice of your final circuit, summary of validated results
and measurements, include key waveforms and any unusual observations and what you did
to fix them. You should include a short conclusion or discussion section which wraps it all
up.
5