Qorvo
Cadence is a pivotal leader in electronic design and computational expertise, using its Intelligent
System Design strategy to turn design concepts into reality. Cadence customers are the world's
most creative and innovative companies, delivering extraordinary electronic products from chips
to boards to systems for the most dynamic market applications. www.cadence.com
© 2020 Cadence Design Systems, Inc. All rights reserved worldwide. Cadence, the Cadence logo, and the other Cadence marks
found at www.cadence.com/go/trademarks are trademarks or registered trademarks of Cadence Design Systems, Inc. All other
trademarks are the property of their respective owners. 14481 10/20 DB/SA/SS-X-QRV/PDF
The power and speed of
AXIEM 3D planar EM
software made it possible
to accurately and efficiently
simulate the entire
structure of this very
complex NDPA MMIC.
Chuck Campbell, Qorvo
Summary
The power and speed of the AWR AXIEM simulator made it
possible to accurately and efficiently simulate the entire
structure of the very complex Qorvo NDPA MMIC device.
The simulated and measured results, as shown in Figure 3,
were in good agreement, and, in the 1.5–17GHz band, exper-
imental results of 9-15W saturated output power with an
associated PAE were typically above 20%. To Campbell's
knowledge, these results are among the highest reported
for a monolithic solid-state power amplifier covering this
frequency range.
Figure 3: 30V small signal gain and return loss (simulated small
signal gain is the broken line)
The Solution
Prior to EM, Qorvo had never attempted to simulate the
entire MMIC circuit of their NDPA, which includes more
than 32 ports and, for a gridded planar EM tool, upwards of
30,000 unknowns.
Nevertheless, Campbell decided to put the AXIEM
simulator to the test. The result: the entire structure from
DC to 120GH was solved in under two minutes per
frequency using a quad-core desktop PC with 4G RAM
(32-bit OS) (Figure 2).
Figure 2: 32-port AXIEM layout
What's more, the software's shape pre-processor and
hybrid adaptive meshing algorithms meant that the final
mesh size was little more than 6,000 unknowns, which
was highly efficient.
