APT9903

APPLICATION NOTE By: Richard Frey, P.E.

500W, CLASS E 27.12 MHz AMPLIFIER

USING A SINGLE PLASTIC MOSFET

1 APT9903
 500W, Class E 27.12 MHz Amplifier
Using A Single Plastic MOSFET

Richard Frey, P.E.

Advanced Power Technology, Inc.
Bend, Oregon 97702 USA

ABSTRACT system level benefits operating at higher voltage.

Power Combining and matching are easier since the
In this paper, we report on the design and evolution drain impedance is higher. Lower RF current is less
of a 500W, 27 MHz Class E amplifier. It doubles stress on the series elements and enables the power
the operating frequency of previous high efficiency supply design to be more efficient, smaller and
amplifiers using MOSFET transistors in the TO-247 lighter.
package. Device criteria, circuit design, and
amplifier performance characteristics are presented This RF application challenges most standard
and compared to a HEPA computer model. switchmode devices due to their packaging and die
layout. The drain of the MOSFET is connected to
INTRODUCTION metal back heat spreader, used for heat sinking the
device. This requires either an insulator between
the package heat spreader and the heat sink or
As the semiconductor industry moves to larger size
grounding the drain and changing the circuit
wafers to raise its productivity, they need more
topology to accommodate it. Either method adds
control of the processing. This translates into the
assembly cost of the PA and increases the potential
need to produce a “harder” and more uniform plasma
for failures due to incorrect installation not to
in vapor deposition and etching operations that rely
mention the poor thermal transfer characteristics of
on RF plasma. Typical operation is at 13.56 MHz
the insulator.
but 27.12 MHz produces better results and is gaining
increased popularity.
The package construction results in high source
inductance, limiting the frequency response of these
One of the challenges facing designers is finding
devices. In some cases, the circuitous gate
devices suitable for service at the higher frequency.
metallization layout caused gate signal propagation
Around 1990, plasma equipment designers
delay, limiting the frequency response and reducing
discovered that some of the inexpensive plastic high
the effective power dissipation of the device.
voltage MOSFETs they used in their switchmode
power supplies were capable of operating in high
Raising the bar to 27 MHz reduces the number of
efficiency Class E service at 13.56 MHz. Since the
suitable switchmode parts. This is primarily due to
switchmode devices operate at high voltage and cost
resistive losses associated with the polysilicon gate
much less than the alternative purpose-built RF
structure in the closed cell geometry used in most
parts, acceptance was almost immediate.
switchmode devices. Input capacitance is essentially
fixed by the device's power rating. Doubling the
Purpose-built RF devices operate on supply voltages
operating frequency quadruples the gate dissipation.
below 50 Vdc. Suitable switchmode devices with
If the ESR of the gate capacitor is large, it cannot
drain breakdown voltage of 1 kV can run on supply
carry the RF current required to drive the device.
voltages up to 300 Vdc. They also demonstrated
APT9903
2
The use of metal gate conductors dramatically * PSPICE MODEL RF N-CHANNEL POWER
reduces gate ESR. It is not unusual to reduce the MOSFET
gate resistance by a factor of 100 -- from 4 to 0.04 * ARF448A/B 27 July 1998
ohms -- on devices of equal power rating. This ** GDS
* .SUBCKT ARF448 6 4 1
complicates RF matching to the gate, but it is now CISS 3 5 1450P
rugged enough to operate reliably. CRSS 5 2 65P
LG 7 6 4 6N
At 13 MHz, the device COSS is combined with an M 8 5 3 3 125-050M L=2U W=1.4; DGSB LEVEL 1
external capacitor to obtain the required shunt output J1 8 3 2 125-050J
D 3 2 125-050D
capacitance needed for proper class E operation [1]. LS 1 3 2 3N
At 27 MHz, the COSS is typically larger than that REGATE 7 5 .29
required for optimum Class E service. This LD 4 2 4 5N
compromises optimum Class E efficiency and output .MODEL 125-050M NMOS (VTO=3.4 KP=14u
power capability. However, this would be a design Lambda=1m Gamma=.2 RD=130m RS=13m)
.MODEL 125-050J NJF (VTO=-25.5 BETA=.01
consideration for any RF device, not just Lambda=.5)
switchmode MOSFETs. .MODEL 125-050D D (BV=550 RS=230M CJO=422P
VJ=670M M=330M)
Vdd .ENDS
C4 C9 L4 C5 C10

L1
J2

Q1
L2
J1 C3
T1
C6-C8
C1 R1
Figure 2. SPICE model for ARF448 MOSFET
Figure 1: 27 MHz Class E Amplifier Schematic

High voltage MOSFETs are now available that C1,C3 75-380 pF mica trimmer, ARCO 465
combine the best practices from the RF world with C4-C8 .01 uF 1 kV disc ceramic
the economy of the switchmode devices and C9,C10 .1 uF 500V disc ceramic
packaging. They are available in mirror image pairs L1 6 uH. 25t #24 ga.enam. 0.5" dia.
and the heat spreader of the plastic TO-247 package L2 210 nH. 4t #8 ga. .75" id, 1" long
L4 2t #20 PTFE on .5" ferrite bead µ=850
is connected to the source. They operate up to 300V Q1 APT ARF448A
Vdd, at frequencies up to 100 MHz. R1 25Ω 5W non-inductive
T1 Pri: 4t #20 PTFE, Sec: 1t brass tube
AMPLIFIER DESIGN on 2 hole balun bead
Fair-Rite #2843010302 µ=850
The device used for this amplifier is an APT
Table 1: 27 MHz Class E Amplifier Part Values
ARF448A. It has a BV of 450V and RθJC of 0.55°C/
W. The class C gain is more than 25dB at 27 MHz.
COSS is 125 pF at Vdd=125V. RD(ON) is 0.4Ω. These
data sheet parameters were entered in the HEPA [1]
3 APT9903
Class E design program and it generated starting copper shim stock.
design values for the circuit. The circuit is common
to most Class E amplifiers. [2] See Figure 1. In
Class E amplifiers there is a capacitor shunted across
the drain to source. There is none used here because
the output capacitance of the device is slightly larger
than the optimum value for that component. This
defines the upper frequency limit for efficient Class
E operation of a particular device. The 3Ω ESR of
COSS is one of the primary loss mechanisms in the
circuit. However, COSS exhibits little of the parasitic
inductance that caused VHF ringing in the circuit
described by Davis. The output circuit values were
adjusted slightly for maximum efficiency at an
output power of 490 W but basically, it worked "as
advertised" right off the bat.

While this description of the output circuit sounds

straightforward, it was not reduced to practice very
easily. The problem was providing enough rf voltage Figure 4: Drain voltage waveform of 500W
to the gate to drive the drain into saturation. The Class E amplifier
input impedance of the gate at 27 MHz is 0.1 -j2.7.
Ciss is 1400 pF. If 10V of peak gate drive is needed, The four turn primary is wound inside the tubes for
a reasonable match between the drive source and maximum coupling and minimum leakage which
gate is required. There is approximately 9 nH of measured 19 nH referenced to the secondary side.
parasitic gate inductance. This is enough inductance
to make it impossible to observe the actual voltage High quality passive components are required in the
applied to "the gate" with an oscilloscope. output network. Most important of these is L2. It
was wound from #8 ga. bare copper wire, its Q
SPICE was used to model the gate drive circuit. A measured 375, and the calculated dissipation is 4.2
SPICE macro model for the device is shown in W. This coil is not capable of continuous duty
Figure 2. The goal was to design a network to match operation unless it is attached with high temperature
the gate impedance sufficiently to permit sine wave solder and/or separate mechanical termination
drive as in reference [2]. A circuit using a 4:1 support is used. It was necessary to parallel three
transformer and an L-network was designed using 10 nF Z5U ceramic coupling capacitors to carry the
winSmith. [3] This worked, but the circuit was not rf current.
stable. The parallel equivalent of the gate is 2200
pF in parallel with 210 ohms. A 25 ohm 5W padding COMPUTER SIMULATION
resistance was placed across the gate. This raises
the effective input impedance to 0.38 -j2.6, lowers A simple SPICE model of the amplifier, similar to
the network Q, and makes it much easier to match that used by Davis in [2], was compared with HEPA
and drive properly. results and the measured results of the operating
amplifier. Overall, the agreement was good,
The input transformer has been used before [4]. It especially between HEPA and the ideal circuit
is made from a two hole ferrite "binocular" bead SPICE model. Attempts to insert the SPICE macro
balun. The secondary winding consists of two 7/8" model of the transistor into the amplifier was not
pieces of 3/16" diameter brass tubing connected with successful. A much more sophisticated model is

APT9903
4
needed to adequately simulate the effects of the current foldback would be needed in the power
nonlinear capacitances of the MOSFET. However, supply of a practicable amplifier.
the SPICE model was very useful for understanding
the gate drive problem mentioned earlier.
90
1000

800
135 45
600
400

200
180 0 0
rel ef f
Figure 3. Drain voltage waveform from HEPA P out
% Id
HEPA assumes that the device is being driven into
-135 -45
saturation, the only input parameter it considers is
input drive power -- used for the overall efficiency
calculation. The gate drive was the biggest problem - 90

in the design because a large RF-capable power Figure 5: Load pull result for 500 W nominal output
MOSFET has a very small input impedance. at 2:1 VSWR. Efficiency is displayed as % x 10.

One of the best tests for reliability of an amplifier is Class E amplifiers are used commercially in RF
mismatch load testing. Called "load pull" in some sources of plasma generators and as exciters for CO2
circles, it is a test which describes what the amplifier lasers. In plasma generator applications it is known
does when operating into a load other than 50 ohms. that the load varies over a particular range. The
The ARF448 has about 175 watts of available feed line length is chosen so the load presented to
dissipation in the test amplifier with its air cooled the amplifier does not enter the +45° quadrant. Other
heat sink. The performance at eight points around a plasma generators have some form of active
2:1 VSWR mismatch circle was calculated using matching network between the amplifier and the
HEPA. Then the same test was run on the amplifier load to handle the load pull issue. The drive or the
itself with the drive duty cycle reduced to 50%. The drain voltage is reduced until a reasonable match is
various loads were obtained using a variable L- achieved. In CO2 lasers, the drive is pulse width
network and adjusting each load impedance with a modulated and under high VSWR conditions, the
vector impedance meter. The differences between pulse width is reduced until the load stabilizes.
the calculated and measured results were very small.
EXPERIMENTAL RESULTS
The measured load pull results are shown in Figure
5. The most obvious conclusion is that there is a The efficiency of the Class E amplifier is 13% better
region in the high impedance inductive quadrant that with 30% more output power than was obtained
should be avoided. Protection would be required to from an initial Class C amplifier designed by
keep the transistor within its safe operating area if classical methods using the same components. [5]
operated in this region. The schematic did not change, only the values of
L2 and C3 are different.
Without protection, the output power soared to 750
W and the drain current was almost twice normal. The performance of the amplifier is summarized in
The efficiency stayed quite constant, never losing the table below along with the calculated values from
more than 11% at any load angle. Some form of HEPA. The results are fairly close.
5 APT9903
 Clearly, the device is operating somewhere past the
Measured HEPA Units point where C OSS provides the right value for
optimum Class E operation. There are tradeoffs to
RF Output Power 490 494.93 W be made between efficiency and maximum output
Drive Power 10 10 W power. The goal was 500W. 83% was the highest
Drain Voltage 125 125 V efficiency obtainable.
Drain Current 4.72 4.47 A
Peak Drain Voltage 430 410.25 V SUMMARY
Peak Drain Current 27.27 A
Loss in L2 4.2 W This article has presented the design of a 27 MHz
Second Harmonic -24 dB Class E amplifier using a high voltage RF MOSFET.
Third Harmonic -57 dB The initial circuit values were obtained using HEPA,
Drain Efficiency 83 88.45 % a commercial Class E design program. 27 MHz was
chosen because it represents a frequency above that
The drain voltage waveform is shown in Figure 4 deemed optimum for the parameters of the device.
and the gate waveform in Figure 6. The peak drain A comparison of the measured and calculated
voltage measures 430V. The approximately 10 nH performance is presented along with several
of parasitic inductance between the "drain" and the practical design issues and application caveats.
scope probe prohibits a true measurement. This is a
problem similar to that already mentioned for the REFERENCES
gate. It is as high as prudent design will allow. The
achieved efficiency is 5.5% below the model. This [1] N. Sokal, HEPA-PLUS/WB v3.29, Design
could be due to the COSS being higher than the data Automation, Inc., Lexington, MA
sheet value, or its ESR being higher than 3 ohms or [2] J. F. Davis, D. B. Rutledge, "Industrial Class E
the switching times being longer than the values power amplifiers with low-cost power MOSFETs
used, and sine-wave drive" Conference Papers for RF
Design '97, Santa Clara, CA, Sept 1997,pp. 283-
297.
[3] winSMITH 2.0, Noble Publishing Corp., Tucker,
GA
[4] R. Frey, "A Push-Pull 300 Watt Amplifier for
81.36 MHz", Applied Microwave and Wireless, vol.
10 no. 3, pp. 36-45, April 1998
[5] Krauss, Bostian, and Raab, Solid State Radio
Engineering. New York: John Wiley & Sons, 1980

Figure 6. Measured gate voltage waveform at

Po=490W

APT9903
6
405 S.W. Columbia Street Parc Cadera Nord - Av. Kennedy BAT B4
Bend, Oregon 97702 USA 33700 Merignac, France
Phone: (541) 382-8028 Phone: 33-556 34 34 71
Fax: (541) 388-0364 Fax: 33-556 47 97 61

Printed -June 1999

7 APT9903