Lecture_6Supp_Power_Devices_OHP

Lecture_6Supp_Power_Devices_OHP
Section 2: Power Semiconductor Devices
Overview
Power switches are the “work-horses” of PE systems
Will operate in two states
• Fully on – i.e. switch closed
o Conducting state (saturation)
• Fully off – switched open
o Blocking state (cut-off)
Power switches never operate in linear mode
Can be categorised into three groups
1. Uncontrolled: Diode
2. Semi-Controlled: Thyristor (SCR)
3. Fully Controlled: Power transistors
(BJT, MOSFET, IGBT, GTO)
Diode
IGBT
1
Two-terminal devices (diodes): state is completely dependent
on the external power circuit
Three-terminal devices: state is not only dependent on their
external power circuit, but also on the signal on their
driving terminal (gate or base)
Majority carrier devices (use one type of charge carrier) (e.g.
Schottky diode, MOSFET) are fast
The charge injection of minority carrier devices (Thyristor,
bipolar transistor, IGBT) allows for better On-state
performance (has higher current driving ability)
ƒ
Thyristors (SCRs, GTOs, MCTs)
o Appeared in 1957, able to withstand very high
reverse breakdown voltage and high current
2
ƒ
Bipolar Junction Transistors (BJTs or BTs)
o Invented in 1947, not capable of substantial power
handling capabilities until the 1960s
ƒ
Power MOSFETs (late 70’s)
o Operate
at
higher
frequencies
than
bipolar
transistors, but at low voltages
ƒ
Insulated Gate Bipolar Transistors (IGBTs) (1990’s)
o Has the power handling capability of the bipolar
transistor, with the advantages of the isolated gate
drive of the power MOSFET
o Has almost completely replaced the bipolar transistor
in power applications, now rivalling MOSFETS
Most discrete power devices are built using a vertical structure;
small-signal devices employ a lateral structure
With the vertical structure, the current rating of the device is
proportional to its area, and the voltage blocking
capability is achieved in the height of the die.
3
Diodes
(MUR 2.1a)
When forward biased, conducts with
small forward voltage across it
(0.2 → 3 V, typically ~0.7 V)
(MUR 2.1b)
When
reversed
biased,
only
small
leakage
current
(~μA→mA) until reverse breakdown voltage is reached
(normally should not happen)
Diodes turn on very fast – almost an ideal switch
DC v-i characteristics given by:
ID = Is (e
(VD / nVT ) −1
VT =
)
kT
q
ID = current through diode
VD = diode voltage
IS = leakage (or reverse saturation) current ~10-6 →10-15 A
n = empirical constant 1 < n < 2
4
When a diode is switched quickly from forward to reverse
bias, it continues to conduct due to minority carriers
which remain in the p-n junction
These carriers require finite time trr (reverse-recovery time)
to recombine with opposite charge and neutralise
(MUR 2.2)
The peak reverse current IRR is
I RR ≈
2QRR
trr
I RR ≈
2QRR
di
= 2QRR
trr
dt
QRR is the reverse recovery charge – the amount of charge
carriers that flow across the diode in the reverse direction
due to changeover from forward conduction to reverse
blocking condition
5
Example
The reverse recovery time of a diode is trr = 3 μs and the rate
of fall of the diode current is 30 A/μs. Determine:
(a) the storage charge
(b) the peak reverse current
Solution
(a)
QRR
1 di 2
−6 2
trr = 0.5 × 30 A/μs × ( 3 × 10 ) = 135 μC
=
2 dt
(b)
I RR = 2QRR
di
= 2 × 135 × 10−6 × 30 × 106 = 90 A
dt
Effects of reverse recovery include:
• Increase in switching losses
• Increase in voltage rating
• Over-voltage (spikes) in inductive loads
6
Schottky Diodes
• Have low forward voltage drop (~0.3 V)
• Can block voltages to 50 – 100 V
• Used in low voltage, high current applications
(e.g. SMPS)
Fast Recovery Diodes
• Very low trr (< 1 μs)
• Power levels at several hundred volts and amps
• Used in high frequency circuits
Line Frequency Diodes
• Have low on-state voltage (< 1 V) but large trr ~25 μs)
• Can block ~5 kV and pass ~5 kA
• Used in line-frequency 50 – 60 Hz applications such as
rectifiers
7
Thyristor/SCR
If VAC (text uses VAK) is positive, then with Vgate = 0, junctions
J1 and J3 are forward biased, but J2 is reverse biased so
no conduction takes place (OFF state)
The SCR is a thyristor commercialised by GE in 1957
The SCR has three operating regions:
Region 1: VAC is negative and iA is
negligible
Region 2. VAC is positive and iA is
negligible
Region 3. VAC is positive and iA
(similar to MUR 2.3b)
has exceeded the latching current (IL); the SCR is ON
If VAC is increased beyond the breakdown voltage of the
thyristor, avalanche breakdown of J2 takes place and the
thyristor starts conducting
8
An SCR can make the transition from region 2 to region 3 by:
(1) Increase the anode voltage until the forward breakdown
voltage is reached and the device turns on
(2) Apply a positive voltage pulse to the gate input to
“trigger” the SCR ON (normal method in control apps)
(3) Rapidly increase the anode-to-cathode voltage VAC.
Once a thyristor is on, it is latched on – it cannot be turned
off by external control – major limitation overcome by
the GTO (Gate Turn-Off Thyristor)
SCRs/Thyristors have voltage ratings of 50 → 7 kV, and
current ratings up to 4000 amperes. They are still the
primary choice for very high power applications
Forward voltage drop in on state ~ 1 – 3 V
The turn-off time tq is the time between when the anode
current has decreased to 0 and when the thyristor is
capable of withstanding a positive VAC without turning
on again
9
Limits thyristor applications to low frequencies - in low and
medium power applications (< 10 kW) they have mostly
been replaced by devices with superior switching
characteristics like MOSFETs or IGBTs.
Once the SCR is turned ON, it can be turned OFF by:
(1) Reducing the anode current below a
minimum value called the holding
current (typically ~1% of the rated
current).
Causes a transition from
region 3 to region 2
If the anode voltage is reversed (i.e., VAC < 0), the SCR
goes from region 3 to region 1. Used in AC circuits
where the voltage reverses each half-cycle (Zero Cross
operation)
(2) In small current applications, the SCR can be turned OFF
by supplying a negative gate current to increase the
holding current. When the increased holding current
exceeds the load current, the SCR switches into region 2.
10
Types of Thyristors
•
SCR – silicon controlled rectifier (SCR)
•
ASCR – asymmetrical SCR
•
RCT – reverse conducting thyristor
•
LASCR – light activated SCR, or LTT – light triggered
thyristor
•
DIAC & SIDAC – both forms of trigger devices
•
BOD – breakover diode –a gateless thyristor triggered by
avalanche current, used in protection applications
•
TRIAC – a bidirectional switching device containing
two thyristor structures
•
GTO (thyristor) – gate turn-off thyristor
ƒ MA-GTO – Modified anode gate turn-off thyristor
ƒ DB-GTO – Distributed buffer GTO
•
MCT – MOSFET controlled thyristor containing two
additional FET structures for on/off control.
•
BRT – Base Resistance Controlled Thyristor
•
SITh – Static induction thyristor, or FCTh – Field
controlled thyristor containing a gate structure that can
shut down anode current flow.
11
TRIACs - TRIode for Alternating Current
Approximately equivalent to two SCRs joined in inverse
parallel and with their gates connected together
Results in a bidirectional electronic switch which can conduct
current in either direction when triggered (unlike SCRs)
Top figure is a SCR, bottom is a Triac
Triggered by either a positive or a negative gate voltage
Once triggered, the device continues to conduct until the
current through it drops below a certain threshold value
The triac is a very convenient switch for AC circuits (the
motivator for their development), allowing the control of
very large power flows with mA control currents
12
The first trigger pulse turns the triac ON during the positive
half-cycle, and it remains on until the AC voltage
reverses and turns the triac OFF
The next triggering pulse turns the triac ON during the
negative half-cycle
The triac again remains ON until the AC voltage reverses and
turns the triac OFF
The trigger circuit then determines how much current is
delivered to the load
13
Controllable Switches
Devices that can be turned on or off by control signal (GTO)
Ideally
• Block large forward and reverse voltages when off
• Conduct large currents with zero voltage when on
• Switch from on→off→on instantaneously
• Only small control signal required
• Operate in saturation and cutoff modes only – linear
operation dissipates excessive power
Power dissipation enormous concern
Devices have finite turn-on and turn-off times during which
power loss occurs
(i.e. power loss ∝ switching frequency)
Steady state power loss ∝ on-state voltage (Von)
14
Bipolar Junction Transistors
Current controlled devices
Recall that IE = IC + IB where IB is effectively the input
current and IC the output current
(MUR 2.7b)
The forward current gain βF = hFE = IC/IB
A sufficiently large IB (continually applied) turns device on
IB >
IC
hFE
The maximum collector current in the active region occurs
when VCB = 0 and is:
I C max =
VCC − VCE VCC − VBE
=
RC
RC
And the corresponding base current is:
I
I B max = C max
βF
15
If IB goes above IBmax, VBE increases, IC increases and VCE falls
below VBE
When CBJ is forward biased with
VBC ~ 0.4 – 0.5V the transistor
enters saturation (defined as point
where increase in IB does not
significantly increase IC)
I C ( sat ) =
VCC − VCE ( sat )
RC
Base drive circuit is complex
On-state voltage (VCE(sat)) ~ 1 – 2 V
Vce < 1 kV, Ic < 400 A
Speed limited ~ 100 ns → 2 μs (frequencies < 5 kHz)
hFE normally 5 – 10 in high power transistors so can be
connected in Darlington configuration to get larger
current gain: β = β1 + β2 + β1β2
Darlingtons typically have higher
VCE(sat) & slower switching speeds
Old technology, replaced by FETs and IGBTs
16
(MUR 2.8a)
Power MOSFET
•
Voltage controlled device
•
Has high commutation speed and good efficiency
•
Has an isolated gate that makes it easy to drive
•
Is the most widely used low-voltage (< 200 V) switch
(MUR 2.9a)
Basic Structure
Commonly based on the Vertical Diffused MOS (VDMOS)
structure (also called Double-Diffused MOS or DMOS)
(the P wells are created by a diffusion process)
The source electrode is placed over the drain: the current flow
is mainly vertical when the transistor is ON
A cell is very small (ones to tens of μm wide): a power
MOSFET is constituted of several thousand cells
Power MOSFETs with lateral structures exist
17
Switching Operation
No gate current except during switching
(MUR 2.9b)
Because of their unipolar nature (there is no need to remove
minority carriers), the power MOSFET can switch at
very high speed ~ 10 ns → 100 ns hence employed for
frequencies >100 kHz and even into MHz) – is the
dominant high frequency switch
Speed limited by the internal capacitances of the MOSFET
which must be charged or discharged – determined by
the external driver circuit
Gate Oxide Breakdown
The gate oxide is very thin (100 nm or less), so it can only
sustain a limited voltage.
Manufacturers often state a maximum Vgs, ~ 20 → 500 V
(Ids < 300 A)
Exceeding this voltage limit can result in a significant
reduction in the lifetime (or even destruction) of the
MOSFET, (with little to no reduction of RDSon)
18
Breakdown voltage/on-state resistance trade-off
Breakdown voltage and RDSon are determined by the doping
level and the thickness of the N- epitaxial layer
The thicker the layer and the lower its doping level, the
higher the breakdown voltage
But the thinner the layer and the higher the doping level, the
lower the RDSon (and therefore the lower the conduction
losses of the MOSFET)
There is a trade-off in the design of a MOSFET, between its
voltage rating and its ON-state resistance
RDSon increases signficantly with voltage rating
RDSon for P-MOSFETs often three times higher than a
N-MOSFET with the same dimensions
19
IGBT (Insulated Gate BJT)
A bipolar transistor driven by a power MOSFET:
Combines the advantages of being a minority carrier device
(good performance in on-state, even for high voltage),
with the high input impedance of a MOSFET
• Voltage controlled device
• Turn on/off times ~ 1 μs (f ~< 100 kHz)
• Von ~ 2 – 3 V (1 kV device)
Applications in:
(MUR 2-12b)
•
High voltage hobbyists (Tesla coils)
•
Electric vehicles and hybrid cars (Toyota’s Prius has a
50
kW
IGBT
inverter
controlling
(Vce < 3 kV more typical with Ic < 1.2 kA)
•
Audio amplifiers
20
the
motors),
The IGBT turns on when VCE is positive (> Vf) and a positive
signal is applied at the gate (Vg > 0)
It turns off when VCE is positive and a 0 signal is applied at
the gate (Vg = 0), or when VCE is negative.
The turnoff characteristic of the IGBT is approximated by two
segments:
•
When the gate signal falls to 0, the collector current
decreases from Imax to 0.1Imax during the fall time (Tf)
•
It then falls from 0.1Imax to 0 during the tail time (Tt)
21
GTO – Gate Turn Off Thyristor
As per thyristor, can be turned ON by gate current pulse
But can be turned off at any time by the application of a
negative gate-cathode voltage
• short time (~ μs)
• large (~ 1/5 → 1/3 anode current)
Can handle large voltages (~5 kV)
(MUR 2-10b)
and currents (~ 5 kA)
GTOs are slow (~ 2 – 25 μs) used in range of high
voltage/current in frequency range ~ 100 Hz – 10 kHz
Cannot handle inductive turn-offs because of large dv/dt
Facing stiff competition from IGBTs in this niche
22
MCT – MOS Controlled Thyristor
New device – voltage controlled
Similar properties to GTO – low voltage drop, and latching
(MUR 2.13b)
Advantages:
• no large negative gate current required for turn on/off
(unlike GTO)
• faster switching speeds (~ μs) than GTO
• smaller on-state voltage drop than IGBT
• voltages to ~ 3 kV, currents ~ 100’s A (smaller current
than GTO)
23
Summary
(Similar to
MUR 2.14)
The power MOSFET can achieve very high operating
frequency, but can't be used with high voltages
This is a physical limit; no improvement is expected from
silicon MOSFETs , but is often the device of choice for
applications below 200 V
The IGBT is steadily improving. It is becoming available at
higher voltage levels but has a relatively low operating
frequency (few devices are rated over 50 kHz)
The major limitation of the BJT is its relatively high voltage
drop in the on-state (2 to 4 V).
At very high power levels, thyristor-based devices are still the
only choice.
24
MCT and GTO devices constructed to overcome this
Type of
Control
Control
Switching
Switch
Signal
Type
Frequency Rating
SCR
Current
Trigger
Low
High
GTO
Current
Trigger
Low
High
BJT
Current
Linear
Medium
Medium
MOSFET Voltage
Linear
Very High
Low
IGBT
Linear
High
Medium
Voltage
Power
A power device is usually attached to
a heatsink to remove the heat
caused by operation losses.
Wide Band-Gap Semiconductors
The major breakthrough is expected from the replacement of
silicon by a wide band-gap semiconductor.
Silicon carbide (SiC) is considered to be the most promising.
Schottky diodes with a breakdown voltage of 1200 V
SiC can operate at higher temperatures (~ 400°C) and has a
lower thermal resistance than Si, allowing better cooling
25
Power Switch Losses
Want to avoid bulky and costly heat sinks and other heat
removal mechanisms
Main power losses
• forward conduction losses
• blocking state losses
• switching losses
Forward Conduction Loss
An ideal switch has zero voltage drop across it during turn-on
Real switch has forward conduction voltage (on state)
between 1 – 3 V.
Power loss is product of volt-drop over device with current
(averaged over the period)
Major loss at low frequency (and DC)
26
Blocking State Loss
During turn-off ideally no current should flow through switch
In reality, small leakage current may flow
Normally leakage current minimal and turn-off losses are
usually neglected.
Switching Loss
Ideal switch
Real Switch
Ideal switch has zero transition time so voltage and current
are switched instantaneously
Real switch has finite voltage fall time and current rise time
Switching loss is product of voltage and current during
transition and is a major loss at high frequency
27
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement