Welcome to our Phase 18 Reliability Report. It is worth highlighting that the Phase 18 Reliability Report
(RR) goes
beyond simply testing more parts. Before initiating the Phase 18 RR, a goal was established to close the
gap between
lab-generated reliability testing results and device lifetimes under various mission profiles. Achieving
this goal
involves extensive, constructive discussions and feedback with our customers. The Phase 18 RR would not
have been
possible without the customers’ productive discussions and, at times, challenging requests. We sincerely
thank all
the customers who contributed to the completion of this document. In addition, to ensure the quality of
the work
presented in this report, most of the content has already been peer-reviewed and published in leading
journals, as
well as published and presented at international power electronics or reliability conferences.
Section 2 highlights the necessity of understanding the fundamental wearout mechanisms applicable to GaN
high-electronmobility-transistors (HEMTs). An executive summary of the primary wearout mechanisms in GaN
HEMTs is provided in Table 1-1.
Section 3 introduces a quantitative approach to estimate overall device lifetime by identifying the dominant
wearout mechanisms
under mission-specific operating conditions in Equation 3-10. In addition, this section presents a more
detailed lifetime modeling
methodology that incorporates multiple stress conditions with different duty cycles, thereby further
extending the applicability
of the proposed approach.
In Section 4, the Phase 18 reliability report follows the same format as the previous Phase 17 report, where
five key wearout
mechanisms in GaN HEMTs are discussed sequentially. It is noted that significant expansions in both
investigation and
understanding have been made and are discussed in detail in the following sections.
-
The wearout mechanism in pGaN gates of GaN HEMTs is discussed in Section 4.1. The Phase 18 Reliability
Report explores
the boundaries of voltage and temperature dependence in leading-edge GaN devices. First-principles
analyses are also
provided to explain the data, and the work has been peer-reviewed and published in journals and in
international conference
proceedings. A new addition is Section 4.1.4, which extends the gate reliability investigation to
dynamic switching conditions
across a wide range of switching frequencies, including the effects of drain-source current during
dynamic gate stress.
-
Section 4.2 discusses the wearout mechanism under drain-source stress. Since the publication of the
Phase 16 RR, the transient
drain overvoltage study has been well received by the customers. In Section 4.2.4 of the Phase 18 RR,
additional drain
overvoltage data are provided on leading-edge 100 VDS-rated and 150 VDS-rated GaN, further demonstrating
the overvoltage
robustness of eGaN technologies.
-
In Section 4.3.3, similar data expansion is presented, based on characterizations of the pulsed current
limits of 100, 150 and
200 VDS-rated GaN devices. Figure 4-29 shows that the pulsed current rating specified on our datasheets
is conservative,
indicating that it may be increased further as more statistical data become available.
-
The thermomechanical wearout section (Section 4.4) has been enhanced in Phase 18 RR. It now features the
investigations
of chip-scale-packaged (CSP) devices and quad-flat no-lead (QFN) packaged devices. As QFN-packaged GaN
devices are
gaining traction in the market, Section 4.4.3 is dedicated to the investigation of thermomechanical
wearout mechanisms
in QFN devices. Temperature cycling, thermal shock, and power cycling reliability are all thoroughly
discussed in this
section. Further progress is underway. Stay tuned!
Lastly, significant efforts have been made to study motor drive specific reliability in GaN. Battery-powered
motor drives are
becoming crucial in applications such as e-mobility systems, humanoid robots, and drones. Devices in these
applications
are subjected to distinct mission profiles, which involve rapid current transients during startup,
acceleration, stall events,
also referred to as the high-power period, followed by a low-power normal operating period characterized by
lower and
more steady current. We present a customized testing methodology that emulates these mission-specific stress
conditions.
Preliminary results, as shown in Section 5.4, demonstrate that EPC’s GaN technology provides a robust
solution for this
type of motor drive system.