Abstract
Abstract
In this article, indium phosphide heterojunction bipolar transistors measured at two different measurement setups are characterized by systematic comparison of the 16-term corrected and conventional 8-term multiline Thru-Reflect-Line (mTRL) corrected scattering parameters. The on-wafer calibration standards and down-scaled transistors with emitter widths of
left bracket 850 right bracket nm
$[850]{nm}$
[
850
]
n
m
and
left bracket 500 right bracket nm
$[500]{nm}$
[
500
]
n
m
are measured up to
left bracket 67 right bracket GHz
$[67]{GHz}$
[
67
]
G
H
z
and up to
left bracket 110 right bracket GHz
$[110]{GHz}$
[
110
]
G
H
z
. A detailed methodology for defining the actual scattering parameters of the calibration standards, which are essential for the successful implementation of on-wafer 16-term calibration, is presented. A direct comparison of Mason’s unilateral power gain determined using a 16-term calibration and a conventional 8-term mTRL calibration demonstrates that the 16-term calibration approach yields a reasonably stable estimate of transistor figures of merit such as maximum oscillation frequency
fmax
$f_{\mathrm{max}}$
f
m
a
x
in a wide range of measurement frequencies, bias conditions, and technology nodes. A statistical analysis of the extracted
fmax
$f_{\mathrm{max}}$
f
m
a
x
values highlights the necessity of employing a 16-term calibration procedure for accurate characterization of transistors with smaller emitter widths.