Lecture II

Types of Diode
Signal Diode
-- Power Diode Fast recovery diode
Schottky Diodes
LED Diode
--Zener Diode
--LASER Diode
Max properties: similar to regular power diodes but recover
time as low as 50ns
The following is a graph of a diodes recovery time. trr is
shorter for fast recovery diodes
Fast Recovery Diodes
Max properties: 1500V, 400A, 1kHz
Forward voltage drop of 0.7 V when on
Power Diodes
What do we use diodes for?
Used to Block DC Voltage
Non-linear mixing of two voltages (e.g. AM)
Turn AC into DC (voltage rectifier)
Voltage multiplier (ex. double input voltage)
Protect circuits by limiting the voltage (clipping and clamping)
Max properties: 400V, 400A
Schottky Diodes
Lower voltage drop when conducting than regular diodes
Very fast recovery time
Ideal for high current low voltage applications
BJTBipolar Junction Transistors is a current controlled device
Types of Transistors
IGBTInsulated Gate Bipolar Transistor is a voltage controlled device
FETField Effect Transistor is a voltage controlled device
MOSFETMetal Oxide Semiconductor is a voltage controlled device
IGBT-NPN IGBT + Diode
R4
10K
+10V
100K
R3
10V
R2
1K
hfe = 100
Q1
2N2222
Inverting Amplifier
Inverting Amplifier Simulation
I
C
= h
fe
I
B
then transistor is in active mode

I
b
> I
c
/h
fe
then transistor is in saturation

b
c
fe
I
I
h
.

0.000us 100.0us 200.0us 300.0us 400.0us 500.0us
12.50 V
10.00 V
7.500 V
5.000 V
2.500 V
0.000 V
-2.500 V
-5.000 V
-7.500 V
A: v1_1
B: q1_3
C: c2_2
When transistor is driven hard in saturation then it acts as a
unidirectional switch.
Transistor is driven in Active Mode.
Linear Voltage Regulator
C
B
+
C1
100uF
A
+10V
10kHz
V1
-200m/200mV
C2
1uF
Q1
BC547A
R5
20K
R4
1k
R3
2K
R2
2K
R1
1k
C
B
A
MOSFET
Measured input characteristics
( ID vs. VGS ) for (a) nMOS transistor
As Can be seen from its input characteristics
MOSFET is a voltage controlled device
Extremely fast switching speed with lower switching losses compare to BJT & IGBT
but act as small resistance R
DS(on)
in on state

In High Voltage Range, due to high R
DS(on)
conduction losses becomes
very high compare to BJT and IGBT and hence MOSFET are used in low voltage
Switching circuits.
IGBT
IGBT-NPN IGBT + Diode
Equivalent circuit model of an IGBT
IGBT has Input characteristics of MOSFET and Output Characteristics of BJT
It is rugged like BJT with low conduction losses but switching speed similar
to MOSFET with lower switching losses than BJT
High Efficiency Power Switches
It all started with the mighty Bipolar Junction Transistor (BJT) in 1970, the
first fully controllable and commercially viable power switch. BJTs gave rise to
switchmode power conversion by late 1970s and made this class of converters a
reality. Then came the Metal Oxide Field Effect Transistor (MOSFET) by the
late 1970s, a new class of power devices that offered significant performance
improvements

compared to BJTs and extended the limits of power conversion technology as we
know it today. In the 1980s, a new addition to the family of power devices was
made with the invention of the Insulated Gate Bipolar Transistor (IGBTs).
The IGBT was built on the core MOSFET technology and extended
MOSFET operation into high power applications.
A BJT is a current controlled device where the device is turned on and off
by controlling the current into its base junction (IB). A base drive circuit
is normally employed to achieve that (Fig. 1). When a BJT is turned
on, a forward voltage drop is developed across the device terminals
(collector-to-emitter), which gives rise to power loss (conduction loss) and
thus heat loss while the device is conducting current.

However, this voltage drop is relatively low and is easily tolerated by the
external circuitry.
The BJT with drive and control circuitry
Although BJTs were the only real power transistors throughout the 1970s,
they had considerable performance and control limitations, namely:
2. BJTs have relatively slow switching characteristics limiting their maximum
operating switching frequencies to the low kHz range.
3. The BJT is a negative temperature coefficient device, meaning that the
device is liable for thermal runaway as the device temperature increases
1. The drive circuitry was complex and quite sizeable since the base current
require to turn-on and off the device is quite high.
The above limitations accelerated the pace of searching for a new power device
and yielded to the development of the power MOSFET.
Metal Oxide Field Effect Transistors (MOSFETs)
A MOSFET is a voltage controlled device where the device is turned on and off
by controlling the voltage across the gate-source junction (VGS) (Fig. 2).
Unlike BJTs, MOSFETs have high input impedances and thus require
minimal gate currents to turn-on and off. This greatly simplifies the drive
circuitry and reduces its cost. In addition, MOSFETs have very fast switching
characteristics allowing them to operate into the MHz range.
MOSFET with drive and control circuitry
In general, MOSFETs extended the performance advantages of BJTs in terms
of high-current handling capability while eliminating their drawbacks in terms
of:
1. Ease of control: Simpler and lower cost drive circuitry.
2. Faster switching behavior: allowing operation into the MHz range. This
resulted in the miniaturization of power converters as the size of the passive
and filtering components was greatly reduced.
2. High on-resistance especially at high breakdown voltages, which limits their
current handling capabilities. The resistive like behavior of MOSFETs entails
that the voltage drop increases with increasing currents (IR drop) as well as
increasing temperatures (increased resistance). One way to alleviate the
forward drop limitation is to parallel MOSFETs to reduce the equivalent
resistance, similar to connecting resistors in parallel (Fig. 3). Just like
resistors, the equivalent resistance of the parallel configuration is divided
down by the number of parallel devices. This greatly reduces the
conduction losses and increases the current carrying capability of power
MOSFETs.
3. High current carrying capabilities: The positive temperature coefficient
characteristics allow MOSFETs to be paralleled which further increases their
current handling capability.
The above has lead to MOSFETs becoming the devices of choice for power
designers and to the elimination of BJTs in low to medium power applications.
Well it is all good to be true for MOSFETs. MOSFETs do have limitations, namely:
1. Relatively low breakdown voltages: The breakdown voltage limit of the metal
oxide junction is limited to 1000-1200V. This has limited the use of power
MOSFETs in high voltage applications.
Parallel connected MOSFETs for increased current handling capability and reduce
conduction loss
An IGBT is a combo device derived from the power MOSFET and BJT
technologies. In fact, the IGBT is a spinoff from power MOSFETs invented to
extend the performance limitations of power MOSFETs into high voltage and
high power applications. It combines the low conduction loss of a BJT with the
relatively high switching speed of a MOSFET.

The IGBT physical structure is a true representation of the combo technology
the device incorporates having a MOSFET input stage and a BJT output stage
(see Fig. 4).
Insulated Gate Biploar Transistors (IGBTs)
IGBT symbol and equivalent internal structure showing the input MOSFET and the
output BJT
Similar to power MOSFETs, the IGBTs are voltage controlled devices and have
high input impedances thus requiring minimal gate drive circuitry. In addition,
IGBTs and can turn-on and off at speeds comparable to MOSFETs. Unlike
MOSFETs, however, IGBTs exhibit forward voltage drops and current densities
similar to BJTs while switching much faster. In addition, their breakdown voltages
are much higher than MOSFETs with readily available devices having breakdown
voltages in excess of 5000V.
In summary, the performance improvements of IGBTs over MOSFETs and BJTs
are summarized below:
5. High breakdown voltages approaching few thousand volts. The above
performance improvements make IGBTs the devices of choice in high
voltage and high power applications. Specifically, IGBTs are better suited for
voltages in excess of 600V and for power levels in excess of 5-10kW.
1. MOSFET input stage which entails high input impedance, minimal drive
requirements, and simple and low cost drive circuitry
2. Fast switching behavior comparable MOSFETs
3. Low forward voltage drop resulting in lower conduction losses
4. High current carrying capabilities. Presently, IGBTs are available in current
ratings in excess of 1000A.
AC to DC and AC to DC to again AC
Why do we require to rectify ac voltage ? then again invert back to AC?
Most of the electronic circuits work on DC power supply and incoming power to
factory or house being AC such as computer TV etc., we need to convert it to DC
Also circuits like battery chargers, mobile chargers do need to convert AC in to DC
In AC drives, motor speed is controlled by changing motor supply frequency to
adjust its speed
In UPS or Uninterruptible Power Supply, AC input is rectified and then inverted
to give out a clean output free from mains disturbances and battery is used in
parallel with rectified voltage to give AC output to sensitive or critical load in
absence of Mains supply such as computer supply, medical equipment etc.
When AC input is rectified and filtered by capacitors or L-C filter it produces
harmonic currents as rectifier load acts as a non-linear load.
This requires reducing the harmonics below acceptable limit which is achieved
By Unity Power Factor Circuit or commonly called as UPF circuit.
Fundamentals of Switch Mode Power Supply
VOUT = D VIN where
BUCK BOOST (Inverting)
Comparison of Linear Versus Switch Mode Power Supply
The advantages and drawbacks of both technologies
Size: - A 50W linear power supply is typically 3 x 5 x 5.5, whereas a 50W
switch-mode can be as small as 3 x 5 x 1. Thats a size reduction of 80%.
Weight: - A 50W linear weighs 4lbs, a corresponding switcher is 0.75lb. As the
power level increases, so does the weight. I personally remember a two-man lift
needed for a 1000W linear. Today one can carry a 2000W in his carry on luggage
when he flies!
Input Voltage Range: - A linear has a very limited input range requiring
that the transformer taps be changed between different countries.
Normally on the specification you will see 100 to 240VAC.
This is because when to input voltage drops more than 10%, the DC
voltage to the shunt regulator drops too low & the power supply cannot
deliver the required output voltage. At input voltages greater than 10%,
too much voltage is delivered to the regulator resulting in over heating.
If a piece of equipment is tested in the US and shipped to Europe, or
even to Mexico in some cases, the transformer taps have to be
manually changed. Forget to set the taps? The power supply will most
certainly blow the fuse, or may well be damaged.
Most switch-mode supplies will operate anywhere in the world (85 to
264VAC), from industrial areas in Japan to the outback of Australia
without any adjustment. The switch-mode supply will also be able to
withstand small losses of AC power in the range of 10-20ms without
affecting the outputs. A linear will not. No one will care if the AC goes
missing for 1/100th of a second when charging your phone, it will
take 100 of these interruptions to delay the charge by one second!
Having a piece of equipment reboot 100 times a day will cause some
heartbreak!
Efficiency: - A linear power supply because of its design will
normally operate at around 60% efficiency for 24V outputs, where
as a switch-mode is normally 80% or more. Efficiency is a measure
of how much energy the power supply wastes. This has to be
removed with fans or heat sinks from the system.
For a 100W output linear, that waste would be 67W. A 100W switch-
mode would be just 25W. 67W 25W = 42W is the extra power lost
Doesnt sound much, but dont try touching a 40W light bulb! If the
equipment were running 24 hours a day, then the extra losses would
be 367kW hours, even at $0.1 per kW hour, thats an extra $37 a
year for a power supply that costs around $80.
As a quick note, in Europe, they are trying to limit those losses of all
power supplies used by consumers particularly when operating off
load (as many products are left plugged in 24 hours a day). Imagine
250 million power supplies eating up a couple Watts. That equates to
the output of a whole power station!
Reliability: - If reliability is calculated using a part count method, then
the linear power supply will win. With the design & quality
improvements made in the last few years with switch-mode parts &
technology, in reality this advantage has been negated. I have
demonstrated life testing data showing no failures after over 1,000,000
hours on some Lambda (reputed company)products.
Electrical Ripple and Noise: - This is where the linear really
scores! The linear obviously is a lot quieter, by up to a 10,000
times. The topology of the switch-mode supply with its high frequency
switching technology had to have a downside right? So if the noise is
10,000 times worse, how can anyone use it? Sounds so bad. In truth,
there are some applications (studio mixers and very sensitive test
equipment) where low electrical noise is critical. The others?
One of my first sales calls in the USA was to a manufacturer who built
semiconductor fabrication equipment. They used 8 really big linear
units in a large box measuring 2x3x4 feet, it was heavy & actually was
dictating the size of their end equipment. I told the engineer that I
could replace all eight units with two modular products measuring
5x5x10. He laughed and said the noise would be too great. I sent him
samples and went to visit three weeks later. He was delighted with the
performance and has been a long term Lambda customer ever since.
5V linear 5V switch-mode Transient Response:
Transient response is how a power supply reacts to a (fast) change
in load. If the output load quickly changes from say full load to half
load, the output voltage of the power supply will rise (overshoot)
before the internal control circuit has time to compensate, and then
undershoot a little less as the circuit over compensates. The length
of time is takes from the instant of the load change to the time the
output voltage settles back into the load regulation limits can be
critical to some loads. Here the linear again outperforms the switch-
mode.
For a 50% change in load the switch-mode will often take 3000us to
recover. A linear supply will recover in 50us. Is this critical for all
applications? There are a few specialized technologies where this is
important and most engineers will advise you if this is critical. For
the other instances on board capacitors at the end load & the
inductance of cables is enough to reduce overshoot down ten-fold
to where it no longer is a concern.
As linear power supplies are quieter and do not need these
capacitors, they simplify the system filtering, and allow more of the
system leakage current budget to be used for other parts like
monitors. The overall size of the system filter can also be reduced.
How much that impacts cost & performance varies from customer to
customer.
Low leakage currents and Conducted EMI: -
A widely used technique in the design of switch-mode power supplies is
to connect special capacitors from the AC input terminals to Earth. This is
an cost effective method to reduce noise from being fed back through the
input wires and potentially affecting other equipment. These capacitors
have a side effect of allowing a leakage current to be passed through
the Earth or ground cable. Many safety specifications have limits on the
amount of this current that is allowed. UL1950 allows 3mA, medical
industrials less than a tenth of that. The gaming industry is even tighter.
Some switch-mode power supplies (like Lambdas Vega series) are
now available with increased internal filtering that allows for low
leakage versions to be offered to meet medical specifications.
Linear Switching
Function
Only steps down (buck) so input voltage
must be greater than output voltage
Step up (boost), step down (buck), inverts
Efficiency
Low to medium, but actual battery life
depends on load current and battery voltage
over time. Efficiency is high if difference
between input and output voltages is small
High, except at very low load currents (A),
where switch-mode quiescent current (I
Q
) is
usually higher
Waste heat
High, if average load and/or input to output
voltage difference are high
Low, as components usually run cool for
power levels below 10 W
Complexity
Low, usually requiring only the regulator and
low-value bypass capacitors
Medium to high, usually requiring inductor,
diode, and filter caps in addition to the IC;
for high-power circuits, external FETs are
needed
Size
Small to medium in portable designs, but
may be larger if heatsinking is needed
Larger than linear at low power, but smaller
at power levels for which linear requires a
heat sink
Total cost Low
Medium to high, largely due to external
components
Ripple/Noise
Low; no ripple, low noise, better noise
rejection
Medium to high, due to ripple at switching
rate
Table 1: Comparison of the characteristics of switching and linear regulators.
As you can see, depending upon what is critical to the Customer will
influence the decision to go with either a switch-mode or a linear
power supply. It is often worth challenging the use of a linear. Sales
of linear supplies (>10W) fall every year as technology adapts and
improves.
Sinusoidal voltage and non-sinusoidal
current give a distortion power factor of
0.75 for this computer power supply load.
AC power flow has the three components: real
power (also known as active power) (P),
measured in watts (W); apparent power (S),
measured in volt-amperes (VA); and reactive
power (Q), measured in reactive volt-amperes
(var).
The power factor is defined as:

In the case of a perfectly sinusoidal
waveform, P, Q and S can be expressed as
phasors that form a phasor triangle such that:

If is the phase angle between the current
and voltage, then the power factor is equal to
the cosine of the angle,

displacement power factor DPF =
distortion factor DF
And PF = DF*DPF
Unity Power Factor