
AN-19
A
6/96
4
to the low output and Miller capacitance of the 
TOPSwitch
internal MOSFET, resulting in fast switching times. The sum of
the conduction, CV
2
f, and crossover losses is 1.39 W for the
TOP214, as compared to 2.04 W for the discrete MOSFET. The
higher conduction losses of the 
TOPSwitch
 are offset by lower
switching losses.
The 
TOPSwitch
 and discrete designs also differ in the amount
of energy that is consumed by the startup and control circuitry.
These losses are shown in Figure 1.  The losses in the 
TOPSwitch
startup circuit are negligible, compared to 30 mW  for the
discrete design. The 
TOPSwitch
 has an internal startup supply
that is automatically switched off when the 
TOPSwitch
 starts
up, so that there are no losses due to the startup circuitry when
TOPSwitch
 is in operation.  Most 3842 and MOSFET power
supply designs use a resistor connected to the high voltage bus
to provide startup bias rather than the more sophisticated startup
bias circuit shown in Figure 2.  
Conventional 3842 and
MOSFET supplies using  resistive startup bias will dissipate
a constant 1-2 watts in the bias resistor, making the
advantages conferred by 
TOPSwitch
 even more apparent.
Controller power consumption  for 
TOPSwitch
 is much less
than a comparable 3842-based circuit (50 mW vs. 300 mW).
The losses in the controller are due to the power consumed by
the control circuit and the power required for the controller to
drive the MOSFET switch. The 
TOPSwitch
 MOSFET is a low
threshold device with low gate capacitance and almost negligible
Miller capacitance, resulting in very low drive power
requirements. The 
TOPSwitch
 controller is a low power CMOS
design typically requiring only 5.7 V, 2.5 mA at maximum duty
cycle, and 6.5 mA at minimum duty cycle. All of these factors
contribute to the low power consumption of the 
TOPSwitch
controller.
Output diode losses for the discrete and 
TOPSwitch
 supplies are
roughly the same. Primary clamp circuit losses are lower for the
TOPSwitch
 circuit, due to the higher operating frequency of the
TOPSwitch
 design. For a fixed value of primary inductance,
higher operating frequency results in lower peak operating
current in the primary, reducing the amount of energy stored in
the leakage inductance and therefore, the power lost in the
clamp circuit. Miscellaneous losses in the 
TOPSwitch
 circuit
are higher, due to the higher operating frequency of 
TOPSwitch
(100 KHz vs. 76 KHz for the discrete design) and the higher
output pre-load current.
TOPSwitchPower Supply Efficiency
For purposes of illustration, a more detailed efficiency analysis
was performed using the ST204A reference design board. The
ST204A is a 15 V, 30 W universal-input flyback power supply
using the TOP204. A complete schematic diagram of this
design is shown in Figure 4.  As shown in Figure 5, this design
has an efficiency of  80% or greater at full load over most of the
90-264 VAC operating range. At input voltages greater than
120 VAC, efficiency is between 85 and 87%. Various parts of
this design will be examined with regard to their effect on the
overall efficiency of the supply, and techniques will be presented
for efficiency measurement and optimization.
Elements of Power Consumption
A  power  dissipation  budget  for  the  ST204A  is  shown  in
Figure 6. Power dissipation was measured for selected
components  at  input  voltages of  90, 120, and 240 VAC, for
30 W output power.  Due to inevitable inaccuracies in measuring
and estimating the dissipation of the various components, the
sum of all the individual power loss components is different
from  the  total power  loss as measured at the supply input by
 3-4%.  A relatively small group of components is responsible
for most of the power loss in the ST204A.  These are the input
common mode inductor (L2),  input rectifier bridge (BR1),
TOPSwitch
 (U1), drain voltage clamp Zener (VR1), and the
output rectifier (D2). Other components dissipate a relatively
small amount of power, but have a large effect on the overall
efficiency of the supply. These are the input filter capacitor (C1)
and transformer (T1).
Measurement Techniques
Measuring the true efficiency contribution of each component
in a power supply is not always a straightforward process.  In the
case of the ST204A and the comparison study cited above, three
measurement techniques were used to obtain the power budgets
shown  in Figures 1 and 6: direct measurement with a wattmeter,
calculation from voltage and current measurements, and the DC
thermal equivalent method.
Direct Measurement
Direct measurement of power dissipation is useful mostly for
measuring the overall efficiency of a power supply. This
measurement is best performed with a wattmeter which is
designed to provide the average reading necessary to obtain
100
70
0
100
200
300
Input Voltage (VAC)
O
80
90
P
Po = 30 W
ST204A
Figure 5. Efficiency vs. Input Voltage, 30 W Output.