Power Electronics I

Lecture 2: Switches

Dr.-Ing. Moustafa Adly

Senior Automotive Development Engineer – Volkswagen AG
Senior Member – IEEE
Associate Editor – IET
 Thyristor Triggering

Thyristor triggering (turning ON):

1) Forward voltage triggering

2) Gate triggering

𝑑𝑣
3) Voltage gradient triggering
𝑑𝑡

4) Light triggering (light/photons-activated,

based on electron-hole pair generation in
the pn-junctions)

5) Temperature triggering (expensive &

harmful)

Dr.-Ing. Moustafa Adly Power Electronics I 2

 Thyristor Triggering

Thyristor triggering (turning ON):

1) Forward voltage triggering

When 𝑉𝐴𝐾 exceeds the forward

breakdown voltage 𝑉𝐵0 (avalanche
breakdown). The conduction of current
continues even if 𝑉𝐴𝐾 reduces below 𝑉𝐵0
till thyristor current gets below the
holding current. The gate current is not
needed in this case.
This technique has faster switching
effect (than gate triggering) but is
harmful for the device.

Dr.-Ing. Moustafa Adly Power Electronics I 3

 Thyristor Triggering

Thyristor triggering (turning ON):

2) Gate triggering

Turning ON the thyristor through controlling

the gate current 𝑖𝐺 . This technique is reliable
and efficient. In this technique, a positive
gate voltage is applied (not permanently).
The gate signal can be DC (disadvantage is
power loss), AC, or pulse (short or long single
pulse or train of pulses).

Dr.-Ing. Moustafa Adly Power Electronics I 4

 Thyristor Triggering

Thyristor triggering (turning ON):

𝒅𝒗
3) Voltage gradient triggering
𝒅𝒕

When SCR is forward biased, J1 and J3

are forward biased and J2 is reversed
biased. So J2 behaves as an insulator
between two conducting plates (junction
capacitor). Increasing the rate of change
of forward voltage (not magnitude), J2
breaks and starts conducting. High
values of changing current may damage
the SCR. Therefore, this technique may
destruct the device.

Dr.-Ing. Moustafa Adly Power Electronics I 5

 Thyristor Commutation

Thyristor commutation (turning OFF):

Thyristor cannot be turned off by applying

negative gate current. In all commutation
techniques, a reverse voltage is applied
across the thyristor during the turn OFF
process. There are two methods by which a
thyristor can be turned OFF.

1) Natural commutation

2) Forced commutation

Dr.-Ing. Moustafa Adly Power Electronics I 6

 Thyristor Commutation

Thyristor commutation (turning OFF):

1) Natural commutation

No external circuit is required for this

purpose. Natural commutation occurs when
the current flowing in the thyristor circuit
gets naturally below the holding current (as
in traditional AC circuits).

2) Forced commutation

An external “commutation circuit” is required

to force the thyristor current to decrease
below the holding current. This is intended to
handle DC circuits with current always over
the holding current.

Dr.-Ing. Moustafa Adly Power Electronics I 7

 Thyristor Family Triac & GTO

Triac

The Triac is composed of two back-to-back

thyristors (therefore bidirectional) with one
gate. Accordingly, the current can flow though
it in both directions (as in AC circuits).

Gate Turn Off Thyristor (GTO)

GTO behaves like a normal thyristor but is

fully controlled. It can be turned on through
positive gate current. Moreover, it can be
turned off through a reverse gate current.
However, turning off is difficult as it needs
very large reverse gate current

So far “Semi-controlled switches”

Dr.-Ing. Moustafa Adly Power Electronics I 8

 Fully-Controlled Switches BJT-Transistor

Power NPN Transistor

It is formed of 3 layers (n-p-n) and thus two

junctions and has three terminals. The base
current 𝐼𝐵 controls its operation and must be
permanently available (non-latching switch as
it turns off with zero 𝐼𝐵 ). Commonly used in
the past, replaced by IGBTs and MOSFETs.

ON-state: 𝒗𝑪𝑬 > 𝟎 & 𝒊𝑩 > 𝟎

OFF-state: 𝒊𝑩 < 𝟎

Dr.-Ing. Moustafa Adly Power Electronics I 9

 Fully-Controlled Switches BJT-Transistor
Quasi-saturation

I-V Characteristics Saturation

Active region

Operation region Characteristics

Higher
Cut-off region (𝑖𝐵 = 0) Transistor is OFF
Active region (𝑖𝐵 > 0) Current amplifier

Quasi-saturation region Intermediate stage Cut-off region

Saturation region (𝑖𝐵 > 0) Power switch

Dr.-Ing. Moustafa Adly Power Electronics I 10

 Fully-Controlled Switches MOSFET
N-Channel MOSFET

MOSFET Transistor

MOSFET is characterized by: ease of control through the

gate, optimality for low-voltage operation at high
switching frequencies, but on-state resistance is a concern
at higher voltage ratings. It is controlled by the gate
voltage 𝑣𝐺𝑆 in a non-latching form.
Applying a positive voltage (higher than threshold) at the
gate attracts the electrons of the p-region and creates an
electrons channel between the two n-regions. Thus, it
allows for current flow.

ON-state:
𝒗𝑮𝑺 > 𝒗𝑻𝒉 to build the channel & 𝒗𝑫𝑺 > 𝟎 to get 𝒊𝑫 > 𝟎

OFF-state:
𝒗𝑮𝑺 < 𝒗𝑻𝒉 so no channel and thus 𝒊𝑫 ≅ 𝟎

Dr.-Ing. Moustafa Adly Power Electronics I 11

 Fully-Controlled Switches MOSFET

Ohmic region
I-V Characteristics
Active region
Operation region Characteristics
Cut-off region (𝑣𝐺𝑆 < 𝑣𝑇ℎ ) MOSFET is OFF

Higher
Voltage controlled current
Active region (𝑣𝐺𝑆 > 𝑣𝑇ℎ )
source
Power switch with a resistive
Linear/ohmic region
effect between drain and source Cut-off region
(𝑣𝐺𝑆 > 𝑣𝑇ℎ )
(losses)

So far “Fully-controllable switches”

Dr.-Ing. Moustafa Adly Power Electronics I 12

 Power Losses

Types of power losses in power electronic devices

Loss type Discription Relevance / Comment

Switching loss During turning the switch ON and OFF High (dominant)
Conduction loss During the conduction period High (dominant)

Blocking loss During the forward or reverse blocking states Low

Driving loss Loss in the control circuit Negligible

Leads loss Loss in the terminals of the device Negligible

Dr.-Ing. Moustafa Adly Power Electronics I 13

Power Losses MOSFET

Dr.-Ing. Moustafa Adly Power Electronics I 14

Power Losses BJT

Linear approximation

Turning on time Conduction time Turning off time Blocking time 𝑡𝑂𝑁
Duty cycle 𝐷 = 𝑇
1
Periodic time = : 𝑓𝑠 switching frequency
𝑓𝑠

Dr.-Ing. Moustafa Adly Power Electronics I 15

Power Losses Switching loss

Instantaneous power loss

during turning on

Energy loss during turning on

Instantaneous power loss

during turning off

Energy loss during turning off

Total energy loss during

switching

Average power loss during

switching
Dr.-Ing. Moustafa Adly Power Electronics I 16
Power Losses Cond. / Blocking Loss

Instantaneous power loss

during conduction

Energy loss during conduction

Average power loss during

conduction

Instantaneous power loss

during blocking

Total energy loss during

blocking

Average power loss during

blocking
Dr.-Ing. Moustafa Adly Power Electronics I 17
Power Losses Cond. / Blocking Loss
Average power loss during
conduction

Average power loss during They are all function of 𝑓𝑠 , so

blocking which 𝑓𝑠 to use?

Average power loss during

switching

The switching frequency 𝑓𝑠 that causes 𝑷𝒔𝒘 = 𝑷𝑶𝑵 + 𝑷𝑶𝑭𝑭 (equal switching and conduction/blocking
losses) is called the corner frequency 𝒇𝒄

▪ At frequencies less than corner frequency (𝑓𝑠 < 𝑓𝑐 ),

switching losses are less than conduction/blocking losses (𝑃𝑠𝑤 < 𝑃𝑂𝑁 + 𝑃𝑂𝐹𝐹 )
▪ At frequencies higher than corner frequency (𝑓𝑠 > 𝑓𝑐 ),
switching losses are higher than conduction/blocking losses (𝑃𝑠𝑤 > 𝑃𝑂𝑁 + 𝑃𝑂𝐹𝐹 )
▪ Therefore, for larger device lifetime,
it is recommended to reduce the switching losses through using (𝑓𝑠 < 𝑓𝑐 )
Dr.-Ing. Moustafa Adly Power Electronics I 18
 Thermal Design

Power losses in the semiconductor devices must be removed to

limit temperature rise within the device. The reliability of
devices and their lifetime depend on the operating temperatures,
which should be well below their maximum allowed values. On
the other hand, letting them operate at a high temperature
decreases the cost and the size of the heat sinks required.

There are several cooling techniques, but for general-purpose

applications, circuits are often designed for cooling by
normal air convection, without the use of forced air or liquid.

Semiconductor devices come in a variety of packages (for e.g. with

heat sinks to improve heat transfer to the ambient and cool
them down), which differ in cost, ruggedness, and thermal
conduction.

Dr.-Ing. Moustafa Adly Power Electronics I 19

 Thermal Design

A thermal circuit representation is employed to model

the heat flow from the device to the ambient is employed,
where 𝜃 represents the temperature in °C (as voltage in
electrical circuits), 𝑟 is the thermal resistance of the path
(junction – casing – heat sink – ambient) in °C/W, and 𝑃𝑑
is the total power loss in the device in W (switching,
conduction, and blocking)

In case of utilizing one heat sink for two devices, the

thermal circuit will have parallel paths as shown

Dr.-Ing. Moustafa Adly Power Electronics I 20

Thank You
Best wishes!