
A
6/96
AN-19
3
and discrete MOSFET, even though the discrete design may use
a MOSFET with an R
 much lower than the equivalent
TOPSwitch
. The reasons for this become apparent when the
power loss distribution for a discrete supply and 
TOPSwitch
supply are examined and compared in detail. For this purpose,
the power loss budget was measured for a commercially available
24V, 34W universal input switching power supply at an input
voltage of 120 VAC. This discrete design, shown in Figure 2,
uses a 3842 PWM IC controller and a 600 V, 1.2 ohm MOSFET,
operating at a switching frequency of 76 KHz. The primary
circuit of the supply was modified with a TOP214  IC, using the
same transformer, rectifier, and output filter.  The resulting
design is shown in Figure 3.  The power loss budget was
measured  for  the  converted  supply  at  an  input  voltage  of
120  VAC.   The results   for  both  designs  are  compared  in
Figure 1. 
The efficiency of the 
TOPSwitch
 design is slightly
better than the discrete design, even though the R
 of the
TOP214 at 3.6 ohms is three times that of the MOSFET used
in the discrete design.
Part of the efficiency difference is due to the lower total losses
in  the 
 TOPSwitch
  as  compared  to  a  discrete  MOSFET.   In
Figure 1, the switch losses for the discrete MOSFET design and
the 
TOPSwitch
 design are divided into two major components,
conduction loss and switching loss. The conduction loss of the
power MOSFET in the discrete design is 0.37 W, with an
additional loss term of 0.16 W due to the current sense resistor.
The 
TOPSwitch
 conduction loss is much higher at 1.07 W, due
to the higher R
 of the TOP214. However, since the
TOPSwitch
 uses the R
 of its internal MOSFET to sense
current, there is no current sense resistor or its associated power
loss. The switching losses are divided into two components,
CV
2
f  losses and crossover  losses. The CV
2
f  term represents the
dissipation due to the stored  energy in the parasitic capacitance
of the transformer and the output capacitance of the MOSFET
switch, which must be discharged by the MOSFET at the
beginning of each cycle. The crossover loss term is due to the
finite switching time of the MOSFET. During turn on and turn
off, there is a short period when there is significant overlap of
voltage and current across the MOSFET. A slow MOSFET will
have a long overlap period at turn on and turn off, resulting in
higher losses. The CV
2
f
 losses of the 
TOPSwitch
  circuit are
only 74% of the losses of the discrete circuit, even though the
TOPSwitch
 is running at 100 KHz as compared to 76 KHz for
the discrete design. Both designs use the same transformer, so
the difference in CV
2
f  loss is due to the lower output capacitance
of the TOP214, which is about one tenth of the output capacitance
of a comparable discrete MOSFET. The switching crossover
loss of the 
TOPSwitch
 is negligible, compared to the 1.08 W
crossover loss of the discrete MOSFET. This difference is due
Figure 4. Schematic Diagram of the ST204A Power Supply.
PI-1680-112795
15 V
RTN
BR1
400 V
C1
47 
μ
F
400 V
C5
47 
μ
F
C4
0.1 
μ
F
U1
TOP204YAI
R3
6.2 
R2
200 
1/2 W
D2
FMB-29L
D3
1N4148
C2
1000 
μ
F
35 V
T1
T1204
1
9, 10
6,7
4
5
2
D1
BYV26C
C7
1.0 nF
Y1
DRAIN
SOURCE
CONTROL
C3
120 
μ
F
25 V
U2
NEC2501
U3
TL431
R4
49.9 k
R5
10 k
C9
0.1 
μ
F
R1
510 
VR1
P6KE200
L1
3.3 
μ
H
F1
3.15 A
J1
C6
0.1 
μ
F
L2
33 mH
L
N