Transcript Elektroniczne Układy i Systemy Zasilania

SWITCH-MODE POWER
SUPPLIES AND SYSTEMS
Lecture No 10
Switching transformer design rules.
Power losses analysis in switching regulators
Silesian University of Technology
Faculty of Automatic Control, Electronics
and Computer Sciences
Ryszard Siurek Ph.D., El. Eng.
Flyback converter transformer
D1
UIN
Zp
ZS
UIN
n
I0
IC
C
UIN  zp d
dt
UIN
t
Lp
Ipmax
U0
R0
t
IT
T
iT (t) 
IT
energy storing during cycle I
B
BS
B
H
  B  Se
he nce:
UIN  zp  Se  dB  zp  Se  B
t
dt
 zp 
UIN  t
B  S e
Minimum number of zP turns assuming t = tmax, B = Bs, UIN = UINmax:
zpm in 
UINm ax  tm ax
Bs  Se
Assuming required output power equal to P0
2
2
LpIpm
LpIpm
ax
ax
E
 P0 
2
2T
U2w e
P0 
t
2L
U
p
Im ax  w e tm ax
Lp
zpmin – is set for the chosen core
lg   Lp   P0 
Certain air-gap is necessary to
achieve required output power
Cycle II
-
transistor T is off
ID D1
ZS
ID
I0
IDmax
IC
C
U0
U0
E2 
2
E
2  LSIDm ax  U I  P
00
0
2T
T
2 2
UIN
t
 2
LS U0 t'2

zp2
IDm ax 

2 2
UIN
t
U0
t
LS
T
t
R0
Energy stored in the core is transfered to the output during cycle II
Lp
iD (t)  IDm ax 
t’
B
BS
I2
LS Dm ax
2
H
U0
LS

LS 
t'
zp

UIN t
U0 t'
U20
2P0 T
t'2
 zS  zp
U0 t'
UIN t
z 2S U20 t'2
zS
Selecting t’ < T-t for the maximum output power Po one decides to work with the
discontinuous magnetic flux flow in the whole range of load changes. To increase t’
one must also increase LS, and it is related to higher number of turns of the secondary
winding zS. When t’ = T-t transformer starts to operate with continuous magnetic flux
2
flow.


U 
 IN   2
 n 
For discontinuous
For continuous


UIN 

U0  
flux flow
flux flow
U

0
2Lp fI0
n 1 
Flyback transformer design simplified procedure
1.
2.
3.
4.
Select maximum (nominal) output power Po
Select switching frequency – basing on specifications of available magnetic material,
semiconductors etc.
Calculate tmax, current Imax and required value of Lp
Select core dimensions accordind to Hahn diagrams or using „AP” method (same as in
inductor design procedure)
5.
Calculate (find from diagrams) the air-gap
6.
Calculate minimum number of primary turns, calculate required number of turns zP
7.
Select operating pronciple (continuous or discontinuous flux flow)
8.
Calculate secondary number of turns zS
Usually discontinuous flux flow is observed in flyback converters due to the following
reasons :
-
lower number of winding turns (lower „copper” power losses)
-
lower level of EMC disturbances (transistor switches on with current equal to 0)
-
self-oscillating converter is very easy to design (low-cost solution)
Forward converter transformer
Ip
UIN
IS
n
IM
Zp
ZS US
Lp
transformer
2.
equivalent circuit
UIN
t
Lp

02
ΙS

Lp 
Calculation of minimum number of turns for the primary winding to avoid saturation
in most unfavourable operating conditions
zpm in 
3.
US  n
UIN  t
I
n
0,2  S
n
Selection of the core - basing on diagrams (nomograms etc.) relating core
dimensions to total power for certain converter topology
IM 
1.
ΙS
Ip
UINm ax  tm ax
Bs  Se
Equation identical for any
converter topology
Selection of wire cross section (diameter) taking into account primary current RMS
value and calculation of number of turns for required winding inductance Lp (using
Al constant for selected core) – the following condition must be performed: zp > zpmin
4.
Calculation of secondary winding (windings) number of turns
U0  
5.
6.
zS
zp
UIN
 zS  zp
U0 1
UIN 
Calculation of wire cross section area (copper strip, litz wire) for secondary winding
resulting from secondary current RMS value ISrms=nIprms
Checking if it is enough space to place windings in the core (bobbin) window area –
required isolation and winding arrangement according to safety standards must be
considered
bobbin
secondary
windind
magnetic core
leakage distance
(6 mm)
Safety insulation
(3 layers)
primary
winding
3 mm
functional insulation
(between winding layers)
General notes
1.
2.
3.
4.
Core power loses are higher when frequency and flux density amplitude increase that is why the high value of primary inductance Lp is desirable
High Lp value is related to more primary turns – more trouble with placing the winding
in the bobbin and higher „copper” losses - look for optimum settlemet!
Chose the magnetic core with best available performance – high saturation flux
density Bs, lowest power losses, smallest dimensions
Small air-gap in transformer core may be considered (forward converter) – better
utilisation of the core may be achieved by lowering magnetic remanence
B
B - without air-gap
B - with small air -gap
H
5.
Remember that Bs value decreases with temperature – at 100oC it is lower by 20% – 25%
in comparison to the value specified at 25oC
Switching regulator power losses analysis
1. Switching power losses (dynamic)
LL
UT
L
IT
IL
IL
I0
ILmax
I0
ID
T
UIN
D
C Ro
t
T
Ucontr
t
0
ILmax
IT
ILmin
ID
ts
td
t1
UT
UIN
-IRmax
Eloss
dIT
dt
QR - diode reverse charge [mC]
tf
,
t1
ILmin
ITrmsrds
IRmax  2QR
ILmax
overvoltage due to leakage
inductance
Discharging of transistor capacitances CBCand CBE
(bipolar transistors)
Swith-mode power supply power losses - review
1.
Power losses in passive components
2.
- winding resitances (skin and proximity effects)
- capacitor series resitance (ESR) – output filer electrolytic capacitors
- magnetic core losses (hysteresis and eddy currents)
- power losses in snubbar circuits
Static power losses in semiconductors:
- related to ON- resistance of MOSFET transistor or saturation voltage drop
across bipolar transistor
- related to voltage drop across rectifier diodes (mains input rectifier)
and fast swiching output diodes
IMPORTANT!
for bipolar transistors and diodes
Ploss  Iav Uces
3.
for MOSFET trnasistors
2 r
Ploss  Irms
dson
Switching (dynamic) power losses
- related to semiconductor switching times, reverse charges
- depandant on base (gate) drive circuits
How to minimize power losses – general rules
1.
2.
Power losses in passive components
- select proper wire diameter, use copper stripes or litz wire
- select low ESR capacitors (for switching applications), as big (dimensions)
as possible, connected in parallel,
- make wide and thick copper paths on the PCB
- select modern ferrite cores with best performance at specified operating
frequency and smallest dimensions
- avoid high amplitude of flux density changes
- recover the magnetising energy – do not dissipate it
- use converter topologies with low overvoltages – decrease the influence of
leakagae inductances (eg. two-transistor „forward” topology)
Static power losses in semiconductors:
- select MOSFET trnasistors with low on-resistance
- in high power and high voltage applications use IGBT modules (simple
MOSFET drive circuits, low saturation voltage drop as for bipolar transistors)
- use Shottky diodes if possible (voltage drop below 0,5V)
- use synchronous rectification technique
After switching on the internal body diode the
transistor with very low on-resistisance switches on –
the voltage drop across the conducting transistor is
much lower than across the diode
3.
Transistor dynamic power losses
- select fast transistors (low tr anf tf times)
- use special converter topologies with zero-current or zero-voltage switching
(eg. resonant topologies)
- design carefully snubbar circuits
Voltage UT rise without snubbar circuit
IT
Zp
UT
CIN
UIN
IT
Cs
UT
T
Charging of the capacitor
delays voltage rise across the
switching transistor
decreasing significantly
transistor power losses
t
Ds
It is possible to select such value of the capacitor Cs, that the overall power loss in
the transistor and snubbar circuit reaches minimum.