Page 22 - EE Times Europe Magazine | June2020

22 EE|Times EUROPE

WIDE-BANDGAP DEVICES
GaN HEMTs Outperform MOSFETs

in Key Growth Applications

By Maurizio Di Paolo Emilio

ilicon power MOSFETs have not kept cascoded depletion mode (d-GaN). An e-GaN
pace with the evolutionary changes transistor works as a normal MOSFET, even
in the power electronics industry, with a reduced gate-to-source voltage. It
Swhere factors such as efficiency, offers a simpler package and low resistance,
power density, and smaller form factors are with a bidirectional channel and without a
the main demands of the community. Silicon body diode.
MOSFETs have reached their theoretical The d-GaN transistor is normally switched
limits for power electronics, and with board on and needs a negative voltage. You can
space at a premium, power system designers overcome this problem by connecting the
need alternatives. Gallium nitride (GaN) is HEMT transistor in series with a low-voltage
a high-electron-mobility transistor (HEMT) silicon MOSFET, as shown in Figure 1.
semiconductor that is adding real value in In contrast, the e-GaN transistor is nor-
emerging applications. mally off and is turned on with a positive Figure 2: Parasitic capacitance and
voltage applied to the gate. Unlike d-GaN, current flow in a GaN device (Image: EPC)
For a 48-bus system, a e-GaN devices do not need a negative startup
250-kHz GaN solution with bias; with a zero bias on the gate, the device is cause delay during switching transients.
turned off and does not conduct current.
The reverse-conduction characteristics of a
The threshold of an e-GaN FET is lower
double the frequency allows than that of a silicon MOSFET, yielding a switching device are important. In a MOSFET,
the voltage drop of a body diode is low and its
a 35% size reduction, lowers very low gate-to-drain capacitance (C GD ). The reverse recovery is very slow, resulting in high
inductor DCR losses, and cuts low-capacitance structure permits switch- switching losses.
ing hundreds of volts in nanoseconds at
GaN devices do not have a reverse body
system cost by roughly 20%. megahertz frequencies. GaN FETs’ large gate- diode; they are able to conduct in the reverse
to-source capacitance (C GS ) relative to C GD direction because of their physical nature.
gives the devices good dV/dt immunity. In the case of reverse conduction, it will be
GaN transistors are significantly faster and The dV/dt sensitivity of power-switching necessary to have dead time. A d-GaN device
smaller than silicon MOSFETs, enabling effi- devices is caused by the various parasitic in cascade has reverse recovery by means of
ciency gains that have opened the door capacitance and gate-drive circuit impedance the low-voltage silicon device.
to applications not possible with silicon levels. The gate-charge (Q g ) parameter, on In hard-switching converters, the out-
technology. Efficient Power Conversion’s the other hand, indicates the ability of the put charge is dissipated in the FET at each
(EPC’s) eGaN FETs are supplied in low- device to change states quickly, reaching a power-on transition. This loss is proportional
inductance, low-resistance, small, and low- higher dV/dt with minimal switching losses. to the output charge (Q OSS ), bus voltage, and
cost land grid array (LGA) and ball grid array The gate charge of an e-GaN device is 10× switching frequency. GaN FETs have a signifi-
(BGA) packages. The new EPC FETs offer higher than for an equivalent MOSFET, while cantly lower Q OSS than silicon FETs, reducing
designers best-in-class performance com- d-GaN devices have about a 2× to 5× higher the output charge loss per cycle and thus
pared with silicon MOSFETs in both hard- and Q g than MOSFETs. allowing higher frequencies (Figure 2).
soft-switching applications. To determine the dV/dt sensitivity of a
power switch, you can use a figure of merit AUTOMOTIVE AND CONSUMER
GaN FETs called the Miller charge ratio (Q GD /Q GS1 ). A SOLUTIONS
GaN switching devices are available in two Miller charge ratio of <1 will guarantee the Emerging computing applications demand
types: enhancement mode (e-GaN) and theoretical dV/dt immunity. The gate-drive more power in much smaller form factors.
circuit layout is a critical factor in improving In addition to the expanding needs of the
dV/dt immunity. server market, some of the most challenging
The d-GaN transistor has the gate structure applications are multi-user gaming systems,
of a low-voltage silicon MOSFET. Therefore, autonomous cars, and artificial intelligence.
existing commercial MOSFET gate drivers can Automotive systems are increasingly
easily operate d-GaN switches. The downside moving toward 48-V devices, driven by the
is that the addition of the silicon MOSFET increase in electronically controlled
ignition resistance raises the overall ignition energy-hungry functions and the emergence
resistance. The increase can be significant of autonomous vehicles that create additional
for low voltages (<200 V). For higher values demands on the energy distribution system
(600 V), the additional resistance may be only from systems such as LiDAR, radar, cameras,
about 5% of total on-resistance. A d-GaN and ultrasonic sensors. These energy-hungry
transistor also has increased packaging com- processors represent an additional burden for
Figure 1: e-GaN (left) and d-GaN plexity. Parasitic inductance and capacitance traditional 12-V power distribution buses in
configurations (Image: EPC) between the MOSFET and the GaN HEMT may the automotive sector.

JUNE 2020 | www.eetimes.eu