Experimental Verification of Role of Ballast Resistors in
SiGe Power HBTs
by
Yuanyuan Yang
A project report submitted in partial fulfillment of
the requirements for the degree of
Master of Science
(Electrical Engineering)
at the
UNIVERSITY OF WISCONSIN–MADISON
2011
i
ACKNOWLEDGEMENTS
I would like to express my great gratitude to my advisor Prof. Zhenqiang (Jack)
Ma. It is a precious experience to receive guidance from him. He gave me a lot of
support, encouragement and valuable advice during my two-year graduate school
study.
I would like to thank my research-mentor Namki Cho. He taught me a lot,
including all the use of facilities and research tools. As a mentor, he showed me to
learn how to do research work and how to solve problems. He also contributes a lot to
this thesis.
I would like to thank all the other group members and friends: Gui Gui,
Jung-Hun Seo, Kan Zhang, Hongyi Mi, Tzu-Hsuan Chang, Munho Kim and Han
Zhou. They gave me a lot of help and advice, and their support makes my graduate
life more enjoyable.
I am most grateful to my parents who support my every decision in my life, not
only financially but also spiritually. They gave me all the courage and care I need to
finish my academic life. Their proud and smile are the greatest reward for me.
ii
ABSTRACT
To prevent the dramatic degradation of the device performance of SiGe power
heterojunction bipolar transistors (HBTs) and maintain the power efficiency,
emitter-ballast resistors and base-ballast resistors have been employed in high power
bipolar transistors. The microwave performance of emitter-ballast and base-ballast
HBTs has been compared and discussed empirically in the previous work. However, it
has not been clearly pointed out that which ballasting scheme should be judiciously
employed over the entire useful frequency range and under different operation
configurations, in order to extract the maximum RF power performance from power
devices. This thesis analyzes the role of ballast resistors in SiGe power HBTs
theoretically and experimentally. No-ballast resistors, emitter-ballast resistors and
base-ballast resistor in both of common-emitter (CE) and common-base (CB) all have
been evaluated separately.
iii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ............................................................................................ i
ABSTRACT ...................................................................................................................ii
Chapter I Introduction of SiGe HBT power amplifiers ................................................. 1
1.
History of SiGe ................................................................................................ 1
2.
Principles of SiGe HBTs .................................................................................. 5
Chapter II Previous work on ballast implementations on power performance of SiGe
power HBTs ................................................................................................................... 9
1.
Introduction ...................................................................................................... 9
2.
Influence of ballast resistors on small-signal RF performance of
common-emitter SiGe HBTs................................................................................. 11
3.
Influence of ballast resistors on common-base SiGe HBTs ........................... 15
Chapter III Simulations and measurements of ballast resistor implementation under
common-base configuration......................................................................................... 17
1.
Expression for power gain of CB SiGe HBTs in different frequency ranges 17
2.
Simulation for power gain of CB SiGe HBTs ............................................... 26
3.
Measurement for power gain of CB SiGe HBTs ........................................... 29
Chapter IV Conclusion ................................................................................................ 31
Bibliography ................................................................................................................ 32
1
Chapter I
Introduction of SiGe HBT power amplifiers
1. History of SiGe
At the 1950s, Shockley and Kroemer have raised the theory of hetero-junction
bipolar transistors (HBTs) and band-gap engineering first [1]. However, the
fabrication of HBTs was stagnated for 30 years until the epitaxial growth of
hetero-structures was obtained. At that time, HBTs were limited to expensive III-V
material systems, until the late 1980s the epitaxial growth technique of SiGe/Si
structures was used to enable the device fabrication of Si-based HBTs, including SiGe
HBTs, which also present the first practical bandgap-engineered transistors in the Si
material system.
The first SiGe HBT was developed by IBM in December of 1987 [2], with the
molecular beam epitaxy (MBE) growth technique, shortly followed by CVD growth
technique. Since then, other groups developed more SiGe HBT demonstrations. In
June of 1990, the demonstration of a non-self-aligned SiGe HBT, which is grown by
ultra-high vacuum/chemical vapor deposition (UHV/CVD) and also with a peak
cut-off frequency of 75 GHz, has caught the world’s attention [3, 4]. The performance
of this SiGe HBT was almost twice as much as that of Si BJTs. The first SiGe
BiCMOS technology was developed in December of 1992 [5], where the new SiGe
HBTs were integrated with CMOS process. Considering the low cost and high
2
integration of Si CMOS and the high speed performance of SiGe HBTs, SiGe
BiCMOS meet the demands of many different circuit and system applications, such as
digital, analog, mixed-signal and RF/microwave applications.
Many researchers have been developing SiGe HBT technology facilitating the
design and fabrication for high performance RF/microwave circuits, and enable the
fast development of SiGe HBT technology. Currently, the performance of high-speed
low-power SiGe HBTs is very competitive to that of III-V devices, especially due to
aggressive lateral and vertical downscaling of device dimension. Since 2004, IBM
researchers first published the SiGe HBT with both peak cutoff frequency (fT) and
peak maximum oscillation frequency (fmax) higher than 300 GHz [6, 7], and the
technology has been commercialized. Historical events in the development of SiGe
HBTs have been listed in Table 1.1, and historical trends in the peak fT and fmax for
integrated Si bipolar junction transistor (BJT), SiGe HBT and SiGeC HBT
technologies are plotted in Fig. 1.1, Fig. 1.2, and Fig. 1.3, respectively.
3
Historical event
Year
Fundamental HBT patent
1951
Drift-base HBT concept
1954
Basic HBT theory
1957
First growth of SiGe strained layers
1975
First growth of SiGe epitaxy by MBE
1985
First growth of SiGe epitaxy by UHV/CVD
1986
First SiGe HBT
1987
First ideal SiGe HBT grown by CVD
1989
First high-performance SiGe HBT
1990
First self-aligned SiGe HBT
1990
First SiGe HBT ECL ring oscillator
1990
First pnp SiGe HBT
1990
First SiGe HBT BiCMOS technology
1992
First LSI SiGe HBT Integrated Circuit
1993
First SiGe HBT with peak fT above 100 GHz
1993
First SiGe HBT technology in 200-mm manufacturing
1994
First SiGe HBT technology optimized for 77K
1994
First SiGeC HBT
1996
First high power SiGe HBT
1996
First sub-10 psec SiGe HBT ECL circuit
1997
First SiGe HBT with peak fT above 200 GHz
2001
First SiGe HBT with both peak fT and peak fmax above 300 GHz
2004
Table 1.1 Key steps in the evolution of SiGe HBT technology. [5, 6, 7]
4
Fig 1.1 Historical trends in peak cutoff frequency fT for integrated self-aligned SiGe
HBT and SiGeC HBT techenologies. [5]
Fig 1.2 Historical trends in peak maximum oscillation frequency fmax for integrated
self-aligned SiGe HBT and SiGeC HBT technologies. [5]
5
Fig 1.3 Maximum oscillation frequency fmax versus cutoff frequency fT for different
SiGe HBT technology generations. [5]
2. Principles of SiGe HBTs
As with the base CMOS technologies, for higher speed and reduced power
consumption, each HBT generation has moved toward smaller sizes. In this section
we will take the IBM SiGe HBT as an example, discusses importance of employing
SiGe HBT. The improvement in device performance can be shown by comparing the
energy band diagram of a SiGe HBT to that of a conventional Si BJT [8], as shown in
Fig. 1.5.
6
Fig. 1.4 Cross-sectional schematic of a first-generation SiGe HBT. [9]
Fig. 1.5 Energy band diagram of a graded-base SiGe HBT compared to a Si BJT. [9]
7
There is significant importance to use SiGe heterostructure. Firstly, because the
band-gap offset between emitter and base is intentionally introduced, the energy
barrier height for electrons injecting from emitter to base is lowered. Secondly, since
bandgap offset reduces the intrinsic carrier density in base region, the collector
current density is increased. Consequently, the current gain β is exponentially
increased comparing with the homojunction counterpart. While, by using a graded Ge
profile, it can improve the output conductance and early voltage VA. Therefore, for a
SiGe HBT, the β-VA product, which is an important figure of merit for analog
applications, is significantly improved over a comparable Si BJT. What’s more,
because of high β, the extremely high base doping concentration is allowed to reduce
the base resistance RB, therefore the frequency response which is heavily dependent
on RB can be improved. 1/f and board-band noise is also reduced because of the high
base doping concentration. With the improvement of the epitaxial growth technique
and device structure optimization, the performance of SiGe HBTs has been highly
improved.
For RF performance, another important figure of merit is the cutoff frequency fT.
The base transit time τb limits the maximum fT of a transistor. The built-in electric
field, which is introduced by the graded Ge profile across the neutral base region,
helps the transport of electron from emitter to collector and effectively decreases τb.
Additionally, the emitter transit time τe is reciprocally related to β of the transistor, and
therefore greatly reduce τe of SiGe HBTs. The result is the peak fT of a SiGe HBT will
improve dramatically over a comparable Si BJT.
8
Although the structure of the majority of the commercial SiGe HBT is as similar
as shown above, the other type of device structure with low-to-high emitter-to-base
doping profile is also available. This type of SiGe HBTs takes advantage of low base
resistance due to the high base doping concentration. It can easily offer high fmax.
High percentage box-shape Ge profile is usually employed in this type of SiGe HBTs
to counter balance the negative effect of high base doping on current gain, and thus to
have enough current gain. Therefore, the two types of SiGe HBTs have different
device structures in terms of doping profile and Ge profile.
9
Chapter II
Previous work on ballast implementations on power
performance of SiGe power HBTs
1. Introduction
When devices are working at high power densities, thermal stability becomes an
inevitable issue, which may set the ultimate limits on the performance of power HBTs.
Both of the electrothermal properties and the optimal thermal design of high power
bipolar transistors have been studied theoretically and experimentally [10-26]. An
important thermal issue associated with multi-finger high power HBTs is the current
gain collapse [10, 17]. This phenomenon is caused by the thermally induced
non-uniform distribution of the collector current among all fingers. When the
collector current is increased over a critical point for a certain collector voltage, one
finger draws most of the collector current, and then causes a sudden decrease of the
current gain [10]. Although this phenomenon is not destructive to the device, it should
be avoided to prevent the dramatic degradation of the device performance. To achieve
this, emitter-ballast resistors have been employed in high power bipolar transistors to
ensure thermal stability at high power densities [12, 13, 15, 22, 23]. The disadvantage
of emitter ballasting is the reduced power efficiency due to the significant DC power
consumption in the emitter-ballast resistors. Thermal shunt is another effective
method to maintain thermal stability [16, 21]. The disadvantage of thermal shunt is
10
requiring a thick metal bridge connecting all the emitter fingers, which is not
compatible with conventional fabrication processes. To address the efficiency problem
encountered in emitter ballasting scheme, the use of base-ballast resistors has been
demonstrated [11] and employed in SiGe PA designs [27]. Because of the small base
current through base-ballast resistors, the DC power consumption is much smaller
than emitter-ballast resistors. To achieve the effective degree of thermal stability (i.e.,
to achieve the same VBE drop), a much larger value for base-ballast resistors is needed
because the base current is much smaller than the emitter current. To maintain high
RF performance, capacitors paralleled with base-ballast resistors are necessary to
bypass high-frequency signal. In this way, base-ballast transistors can achieve thermal
stability, good RF performance and high efficiency at the same time [11]. The
disadvantage of base ballasting is the requirement of bypass capacitors, which are
large enough to minimize the negative effect of base-ballast resistors on device RF
performance. In [11], the microwave performance of emitter-ballast and base-ballast
HBTs has been compared and discussed empirically. However, it has not been clearly
pointed out that which ballasting scheme should be judicious employed over the entire
useful frequency range and under different operation configurations, in order to
extract the maximum RF power performance from power devices. Especially, the
experimental verification data on the effects of using base-ballast resistors is lacking.
Additionally, GaAs based HBTs were studied in [11], which have a good
semi-insulating substrate. Unfortunately, SiGe HBTs are fabricated on low resistivity
Si substrates, which have the well-known issue of substrate parasitics [28]. The issue
11
of poor substrate properties on the implementation of base ballasting has not been
evaluated. Some of the related work on the implementation of ballast resistors,
including the calculation and simplification of equations, simulation and measurement
analysis of CE SiGe power HBTs, as well as previous simulation of CB SiGe power
HBTs are done by Dr. Ningyue Jiang. For completeness of this work, several details
are included in this thesis. For more information, please refer to reference [29-31].
Some of the important aspects of Dr. Jiang’s work that are related to this thesis are
briefly described in the following section to ease the reading of this thesis.
2. Influence of ballast resistors on small-signal RF performance of
common-emitter SiGe HBTs
The power HBTs used in this work are high-voltage devices (34 GHz fT, 60 GHz
fmax, 6V BVCEO). Three different types of SiGe power HBTs have been implemented:
no-ballast HBT, emitter-ballast HBT and base-ballast HBT. They share same device
structure but the implementation of the ballast resistors. All three types of devices
have 36 subcell units, and in each unit, there are 4 emitter fingers with 0.9x10.16μm
2
emitter area for each finger. No intentional ballast resistors are placed in no-ballast
HBTs. In emitter-ballast HBTs, a 7.2Ω resistor is used in series with the emitter in
each subcell unit. In base-ballast HBTs, a 144Ω resistor is inserted in series with the
base in each subcell unit. Both emitter and base-ballast SiGe power HBTs are
thermally stable under high power operations, which indicate the effectiveness of both
12
emitter and base-ballast resistors.
The power gain (Gmax) of a CE HBT in different frequency ranges can be
expressed, in terms of equivalent circuit parameters, as follows:
A. Low-frequency unconditionally stable region: the maximum available power
gain (MAG) is flat, which is
β
β
(2.1)
B. Intermediate-frequency potentially unstable region: the maximum stable gain
(MSG) rolls off with the slope of -10 dB/decade, which is
μ
(2.2)
C. High-frequency unconditionally stable region: MAG generally rolls off with
the slope of -20 dB/decade (The stability factor is ignored here. For an accurate
expression, please refer to [29].), which is
(2.3)
where, rb is total base resistance including pinch resistance, link resistance,
extrinsic resistance and resistance of the added base ballast resistor, rex is external
emitter resistance including added emitter ballast resistance, Cμ is BC junction
capacitance, C is BE junction capacitance including depletion and diffusion
components, ro is the output resistance, and β is the current gain. In addition,
β
and
,
is the transconductance. Since the device structures are the same
except for the ballast resistors, the additional emitter and base-ballast resistors will
only change the values of
and
without affecting other parameters.
The typical power gain (Gmax) curve spanning the entire frequency range for a
13
CE HBT is showed in Fig. 2.1, including the effects of emitter or base-ballast resistors.
Because the frequency of region I is usually very low, it is not of much interest for RF
and microwave applications, so it is not sketched in the figure. If the operation
frequency is within region II, there is an emitter resistance dependence of MSG for
CE HBTs. Thus, MSG will degrade if emitter-ballast resistor is applied, according to
(2.2). At the same time, base resistance has no much effect on MSG of CE HBTs,
indicating that base ballast resistor should be a reasonable choice in this region,
because the power gain characteristics can still be maintained even without
high-frequency bypass capacitors. If the operation frequency falls into region III, as
shown in (2.3), MAG is dependent on base resistance rb, so the value of rb should be
as small as possible.
Therefore, use of base ballast resistors should be avoided in
this MAG range, while, emitter ballast should be used in this stable high-frequency
region.
Fig. 2.1 Schematic of the power gain as a function of frequency for no-ballast,
emitter-ballast and base-ballast CE SiGe HBTs. [30]
14
The measured MAG/MSG curves of power SiGe HBTs of different types, which
is no-ballast, emitter-ballast and base-ballast separately, are shown in Fig. 2.2. For all
types of devices, the bias conditions are the same: Ic=180mA and VCE=5V. From the
curves of no-ballast and emitter-ballast HBTs, it can be seen that power gain
difference is only within the potentially unstable (MSG) region: MSG of
emitter-ballast HBT is about 3 dB lower than no-ballast HBT. While, these two MAG
curves overlap almost exactly, which shows that emitter ballast resistors do not affect
MAG, but MSG, which is consistent with (2.2). Between no-ballast and base-ballast
HBTs, the MAG curve of base-ballast HBT is lower than that of no-ballast HBT, but
their MSG curves are close enough to overlap, which is also consistent with (2.3).
From the figure, we also can see that the MAG/MSG curves of base-ballast and
emitter ballast HBTs merge at 0.34 GHz. Based on Fig. 2.2, base ballast resistor
should be used below 0.34 GHz, and emitter ballast should be used above 0.34 GHz
in order to maximize the power gain values.
15
Fig. 2.2 Measured power gain (MSG/MAG) as a function of frequency for no-ballast,
emitter-ballast and base-ballast CE SiGe HBTs. [30]
3. Influence of ballast resistors on common-base SiGe HBTs
The power gain (Gmax) of a CB HBT in different frequency ranges can be
expressed as the following [29]:
A. Low-frequency unconditionally stable region: the maximum available power
gain (MAG) is flat, which is
β
(2.4)
B. Intermediate-frequency potentially unstable region: the maximum stable gain
(MSG) rolls off with the slope of 10 dB/decade, which is
μ
(2.5)
C. High-frequency unconditionally stable region: MAG generally rolls off with
the slope of 20 dB/decade (The stability factor is ignored here. For an accurate
expression, please refer to [29].), which is
16
μ
(2.6)
The simulations of power performance using JazzSemi 0.18μm SiGe HBT
large-signal models were conducted. The simulated small-signal MAG/MSG curves
of CB SiGe HBTs are shown in Fig. 2.3. According to (2.5) and (2.6), emitter
resistance has no significant effect on high frequency power gain of CB HBTs, which
is consistent with the simulation results. As shown in Fig. 2.3, emitter-ballast and
no-ballast CB HBTs have the same MSG tendency, and emitter ballast resistors only
cause a slight degradation in MAG. However, base-ballast resistors significantly
degrade over the entire frequency range, which strongly indicates that emitter ballast
is the only choice for CB HBTs. It is worth to note that the theoretical prediction on
the ballast resistor implementation in common-base configuration by Dr. Jiang did not
consider the stability effects of the SiGe HBTs.
Fig. 2.3 Simulated power gain (MSG/MAG) as a function of frequency for no-ballast,
emitter-ballast and base-ballast CB SiGe HBTs. [30]
17
Chapter III
Simulations and measurements of ballast resistor
implementation under common-base configuration
1. Expression for power gain of CB SiGe HBTs in different frequency
ranges
The SiGe power HBTs used in this work are high-voltage devices (35 GHz PA,
100 mW CB, 6V BVCEO). Three types of SiGe power HBTs have been implemented,
including no-ballast HBT (Fig. 3.1), emitter-ballast HBT (Fig. 3.2) and base-ballast
HBT (Fig. 3.3).
Fig. 3.1 No-ballast CB SiGe HBT
18
Fig. 3.2 Emitter-ballast CB SiGe HBT
Fig. 3.3 Base-ballast CB SiGe HBT
19
The device structures of these three types of HBTs are the same except for the
implementation of the ballast resistors. All of them consist of 16 subcell units. In each
one, there are 4 emitter fingers with 0.2×10.16μm2 emitter area for each finger. No
ballast resistors have been used in no-ballast HBTs. A 7.2Ω resistor has been
implanted in series with the emitter in emitter-ballast HBTs in each subcell unit. A
108Ω resistor has been placed in series with the base in base-ballast HBTs in each
subcell unit. With these described ballast schemes, both emitter and base-ballast SiGe
power HBTs are thermally stable under high power operations. Device layouts (Fig.
3.4-Fig. 3.9) and simulation circuits (Fig. 3.11-Fig. 3.13) for three different types can
be found below.
Fig. 3.4 Layout of no-ballast CB SiGe HBT
20
Fig. 3.5 Layout of no-ballast CB SiGe HBT
Fig. 3.6 Layout of emitter-ballast CB SiGe HBT
21
Fig. 3.7 Layout of emitter-ballast CB SiGe HBT
Fig. 3.8 Layout of base-ballast CB SiGe HBT
22
Fig. 3.9 Layout of base-ballast CB SiGe HBT
The parasitic emitter resistance (rex) and collector resistance (rc) are included for
analysis in the circuit of small signal T model for CB SiGe HBTs, as showed in Fig.
3.10. They are ignored in the simplified equivalent circuits of SiGe HBTs that have
lower base doping concentrations than the emitter doping concentration, because they
are negligible comparing to total base resistance (rb). For SiGe HBTs with much
higher base doping concentrations than the emitter doping concentration, however, the
total base resistance is reduced to a comparable value to the parasitic emitter
resistance and collector resistance. Then both of them have to be included in the
circuits.
23
Fig. 3.10 Small signal T model for CB SiGe HBTs
The H-parameters for the CB configuration equivalent circuit can be derived as
the following (i, input port; o, output port; r, reverse transmission; f, forward
transmission; b, common-base )
(3.1)
(3.2)
(3.3)
(3.4)
Where
and
. We make reasonable
approximations (justified by actual values) to simplify the above H-parameters and
the power gain expressions in low, intermediate and high frequency regions.
A. Low-frequency Region
In the low-frequency region, the H-parameters in the CB configuration can be
simplified as
(3.5)
24
(3.6)
(3.7)
(3.8)
and the power gain MAGCBl can be simplified as
β
(3.9)
B. Intermediate-Frequency Region
Most of the SiGe HBTs are operated in this region, so it is considered as the
maximum stable power gain (MSG) in this range. MSG can be expressed as
(3.10)
So the MSG for the CB configuration is
(3.11)
C. High-Frequency Region
In this region, devices are unconditionally stable. MAG can be expressed as
(3.12)
Where K is the Rollett’s stability factor (the K-factor). For the CB configuration, the
K factor can be derived as
(3.13)
Where,
and
. MAG can be
roughly simplified as
μ
(3.14)
25
Fig. 3.11 Simulation circuit for no-ballast CB SiGe HBT by ADS
Fig. 3.12 Simulation circuit for emitter-ballast CB SiGe HBT by ADS
26
Fig. 3.13 Simulation circuit for base-ballast CB SiGe HBT by ADS
2. Simulation for power gain of CB SiGe HBTs
As shown in the equation (3.13), emitter resistors do affect the K-factor, which
can be also seen from the simulation results below. According to the equation of
K-factor (3.13) above, for no-ballast CB SiGe HBTs,
(3.15)
For emitter-ballast CB SiGe HBTs,
(3.16)
For base-ballast CB SiGe HBTs,
(3.17)
27
In this device,
and
. When K=1,
femitter-ballast>fbase-ballast>fno-ballast. The simulated results shown in Fig. 3.14 follow the
sequence, which discussed above exactly. The frequency point (femitter-ballast) at which
the K-factor of the emitter-ballast CB SiGe HBTs reaches unity (breakpoint of
MSG/MAG) is larger than the corresponding frequency point (fbase-ballast) of the
base-ballast CB SiGe HBTs, which is also larger than that (fno-ballast) of no-ballast CB
SiGe HBTs.
Fig. 3.14 Simulated power gain (MSG/MAG) as a function of frequency for no-ballast,
emitter-ballast and base-ballast CB SiGe HBTs (VCB=2.2V, VBE=0.9V).
28
(a) VCB=2.2V, VBE=0.9V
(b) VCB=2.5V, VBE=0.9V
29
(c) VCB=2.8V, VBE=0.9V
Fig. 3.15 Measured power gain (MSG/MAG) under different bias as a function of
frequency for no-ballast, emitter-ballast and base-ballast CB SiGe HBTs, (a)
VCB=2.2V, VBE=0.9V; (b) VCB=2.5V, VBE=0.9V; (c) VCB=2.8V, VBE=0.9V.
3. Measurement for power gain of CB SiGe HBTs
The measured small-signal MAG/MSG curves of three types of CB SiGe HBTs
are shown in Fig. 3.15, which fits both of the calculation and simulation results. All
the three frequency points (fK=1, emitter-ballast, fK=1, base-ballast and fK=1, no-ballast) arrange in
the same order. According to the simulated and measurement results, emitter
resistance has no significant effect on high-frequency power gain region. As the figure
shown above, emitter-ballast and no-ballast CB HBTs show almost the same MAG. In
30
MSG region, both power gain curves of base and emitter-ballast are degraded. This
indicates that emitter-ballast would be a good choice for CB HBTs over the entire
frequency range, but not so ideal in MSG range.
31
Chapter IV
Conclusion
In this thesis, experimental work has been done to clarify role of ballast resistors
of SiGe power HBTs in common-base configuration. For the 35 GHz HBTs, as a
general example, used in this thesis, according to the simulated and measurement
results, emitter resistance has no significant effect on high-frequency power gain
region. Emitter-ballast and no-ballast CB HBTs show almost the same MAG. In MSG
region, both power gain curves of base and emitter-ballast are degraded. This
indicates that emitter-ballast would be a good choice for CB HBTs over the entire
frequency range, but not so ideal in MSG range. Due to the effects of stability factor,
previous theoretical predictions without considering those effects for common-base
SiGe HBTs seem to be over simplified. By combining the previous conclusions
obtained for the common-emitter HBTs, the theoretical analysis, and the simulated
and measured results exhibited in this thesis, the use of ballast resistors need to be
wisely considered. This study will be significant for the practical use of ballast
resistors, although specific to SiGe power HBTs, is also applicable to other type of
HBTs where thermal effects are not avoidable.
32
Bibliography
[1] H. Kroemer, “Theory of a wide-gap emitter for transistors,” Proc. IRE, vol. 45, pp.
1535-1537, 1957.
[2] S. Iyer, G. Patton, S. Delage, S. Tiwari, and J. Stork, “Silicon-germanium base
heterojunction bipolar transistors by molecular beam epitaxy,” Tech. Dig. Int.
Electron Devices Meeting, 1987, pp. 874-875.
[3] G. Patton, J. Comfort, B. Meyerson, E. Crabbe, E. Fresart, J. Stork, J. Sun, D.
Harame and J. Burghartz, “63-75 GHz fT SiGe-base heterojunction-bipolar
technology, ” Tech. Dig. IEEE Symposium on VLSI Tech., 1990, pp. 49-50.
[4] G. Patton, J. Comfort, B. Meyerson, E. Crabbe, G. Scilla, E. Fresart, J. Stork, J.
Sun, D. Harame and J. Burghartz, “75 GHz fT SiGe-base heterojunction bipolar
transistors, ” IEEE Trans. Electron Devices, vol. 11, no. 4, pp. 171-183, Apr.
1990.
[5] J. Cressler and G. Niu, Silicon Germanium Heterojunction Bipolar Transistors,
Boston, MA: Artech House, 2003.
[6] J. Rieh, B. Jagannathan, H. Chen, K. Schonenberg, D. Angell, A. Chinthakindi, J.
Florkey, F. Golan, D. Greenberg, S. Jeng, M. Khater, F. Pagette, C. Schnabel, P.
Smith, A. Strieker, K. Vaed, R. Volant, D. Ahlgren, G. Freeman, K. Stein, and S.
Subbanna, “SiGe HBTs with cut-off frequency of 350 GHz,” Tech. Dig. Int.
Electron Devices Meeting, 2002, pp. 771-774.
[7] J. S. Rieh, D. Greenberg, A. Strieker, and G. Freeman, “Scaling of SiGe heteroj
33
unction bipolar transistors,” Proceedings of the IEEE, vol. 93, no. 9, pp.
1522-1538, 2005.
[8] D. Harame, J. Comfort, J. Cressler, E. Crabbe, J. Sun, B. Meyerson and T. Tice,
“Si/SiGe epitaxial-base transistors-Part I: materials, physics, and circuits,” IEEE
Trans. Electron Devices, vol. 12, no. 3, pp. 455-468, Mar. 1995.
[9] J. D. Cressler, “SiGe HBT technology: A new contender for Si-based RF and
microwave circuit applications,” IEEE Trans. Microw. Theory Tech., vol. 46, no.
5, pp. 572-589, May 1998.
[10] W. Liu, S. Nelson, D. G. Hill, and A. Khatibzadeh, “Current gain collapse in
microwave multifinger heterojunction bipolar transistors operated at very high
power densities,” IEEE Trans. Electron Devices, vol. 40, no. 11, pp. 1917-1927,
Nov. 1993.
[11] W. Liu, A. Khatibzadeh, J. Sweder, and H.-F. Chau, “The use of base ballasting to
prevent the collapse of current gain in AlGaAs/GaAs heterojunction bipolar
transistors,” IEEE Trans. Electron Devices, vol. 43, no. 2, pp. 245-251, Feb.
1996.
[12] R. H. Winkler, “Thermal properties of high-power transistors,” IEEE Trans.
Electron Devices, vol. 14, no. 5, pp. 260-263, May 1967.
[13] R. P. Arnold and D. S. Zoroglu, “A quantitative study of emitter ballasting,” IEEE
Trans. Electron Devices, vol. 21, no. 7, pp. 385-391, Jul. 1974.
[14] G.-B. Gao, M.-Z. Wang, X. Gui, and H. Morkoç, “Thermal design studies of
high-power heterojunction bipolar transistors,” IEEE Trans. Electron Devices, vol.
34
36, no. 5, pp. 854-863, May 1989.
[15] G.-B. Gao, M. S. Ünlü, H. Morkoç, and D. L. Blackburn, “Emitter ballasting
resistor design for, and current handling capability of AlGaAs/GaAs power
heterojunction bipolar transistors,” IEEE Trans. Electron Devices, vol. 38, no. 2,
pp. 185-196, Feb. 1991.
[16] B. Bayraktaroglu, J. Barrette, L. Kehias, C. I. Huang, R. Fitch, R. Neikhard, and
R. Scherer, “Very high-power-density CW operation of GaAs/AlGaAs microwave
heterojunction bipolar transistors,” IEEE Electron Device Lett., vol. 14, no. 10, pp.
493-495, Oct. 1993.
[17] W. Liu and A. Khatibzadeh, “The collapse of current gain in multi-finger
heterojunction bipolar transistors: Its substrate temperature dependence,
instability criteria, and modeling,” IEEE Trans. Electron Devices, vol. 41, no. 10,
pp. 1698-1707, Oct. 1994.
[18] L. L. Liou and B. Bayraktaroglu, “Thermal stability analysis of AlGaAs/GaAs
heterojunction bipolar transistors with multiple emitter fingers,” IEEE Trans.
Electron Devices, vol. 41, no. 5, pp. 629-636, May 1994.
[19] W. Liu, “Thermal coupling in 2-finger heterojunction bipolar transistors,” IEEE
Trans. Electron Devices, vol. 42, no. 6, pp. 1033-1038, Jun. 1995.
[20] K. Lu and C. M. Snowden, “Analysis of thermal instability in multi-finger power
AlGaAs/GaAs HBT’s,” IEEE Trans. Electron Devices, vol. 43, no. 11, pp.
1799-1805, Nov. 1996.
[21] J. Sewell, L. L. Liou, D. Barlage, J. Barrette, C. Bozada, R. Dettmer, R. Fitch, T.
35
Jenkins, R. Lee, M. Mack, G. Trombley, and P. Watson, “Thermal
characterization of thermally-shunted heterojunction bipolar transistors,” IEEE
Electron Device Lett., vol. 17, no. 1, pp. 19-21, Jan. 1996.
[22] M. G. Adlerstein, “Thermal stability of emitter ballasted HBT’s,” IEEE Trans.
Electron Devices, vol. 45, no. 8, pp. 1653-1655, Aug. 1998.
[23] C. Liao, C. Lee, N. L. Wang, and B. Lin, “Optimum design for a thermally stable
multifinger power transistors,” IEEE Trans. Electron Devices, vol. 49, no. 5, pp.
902-908, May 2002.
[24] N. Nenadović, L. K. Nanver, and J. W. Slotboom, “Electrothermal limitations on
the current density of high-frequency bipolar transistors,” IEEE Trans. Electron
Devices, vol. 51, no. 12, pp. 2175-2180, Dec. 2004.
[25] N. Rinaldi and V. d’ Alessandro, “Theory of electrothermal behavior of bipolar
transistors: Part I – Single-finger devices,” IEEE Trans. Electron Devices, vol. 52,
no. 9, pp. 2009-2021, Sep. 2005.
[26] N. Rinaldi and V. d’ Alessandro, “Theory of electrothermal behavior of bipolar
transistors: Part II – Two-finger devices,” IEEE Trans. Electron Devices, vol. 52,
no. 9, pp. 2022-2033, Sep. 2005.
[27] P. Tseng, L. Zhang, G. Gao, and M. Chang, “A 3-V monolithic SiGe HBT power
amplifier for dual-mode (CDMA/AMPS) cellular handset applications,” IEEE J.
Solid-State Circuits, vol. 35, no. 9, pp. 1338-1344, Sep. 2000.
[28] M. Pfost, H.-M. Rein, and T. Holzwarth, “Modeling substrate effects in the design
of high-speed Si-bipolar IC’s,” IEEE J. Solid-State Circuits, vol. 31, no. 10, pp.
36
1493-1501, Oct. 1996.
[29] Z. Ma and N. Jiang, “On the operation configuration of SiGe HBTs based on
power gain analysis,” IEEE Trans. Electron Devices, vol. 52, no. 2, pp. 248-255,
Feb. 2005.
[30] N. Jiang, Z. Ma, P. Ma, V. Reddy and M. Racanelli, “Impact of ballast resistor
implementations on power performance of SiGe power HBTs,” IEEE BCTM,
2005.
[31] N. Jiang, “Novel SiGe BiCMOS device physics phenomena and their application
to power amplifiers”, Dissertation for the Degree of Philosophy, University of
Wisconsin-Madison, 2006.
[32] G. Qin, N. Jiang, J. Seo, N. Cho, G. Ponchak, D. Weide, P. Ma, S. Stetson, M.
Racanelli and Z. Ma, “Cryogenic operation of a 24 GHz MMIC SiGe HBT
medium power amplifier”, Semicond. Sci. Technol., 25, 2010.
[33] H. Page, C. Becker, A. Robertson, G. Glastre, V. Ortiz, and C. Sirtori, Appl. Phys.
Lett. 78, 3529 (2001)
[34] D. Harame, D. Ahlgren, D. Coolbaugh, J. Dunn, G. Freeman, J. Gillis, R. Groves,
G. Hendersen, R. Johnson, A. Joseph, S. Subbanna, A. Victor, K. Watson, C.
Webster, and P. Zampardi, “Current status and future trends of SiGe BiCMOS
technology,” IEEE Trans. Electron Devices, vol. 48, no. 11, pp. 2575-2594, Nov.
2001.
[35] R. Götzfried, F. Beisswanger, and S. Gerlach, “Design of RF integrated circuits
using SiGe bipolar technology,” IEEE J. Solid-State Circuits, vol. 33, no. 9, pp.
37
1417-1422, Sep. 1998.
[36] Z. Ma, S. Mohammadi, P. Bhattacharya, L. P. B. Katehi, S. A. Alterovitz, G. E.
Ponchak, “High power X-band (8.4GHz) SiGe/Si heterojunction bipolar
transistor,” Electron. Lett., vol. 37, no. 12, pp. 790-791, Jun. 2001.
[37] Z. Ma, S. Mohammadi, P. Bhattacharya, L. P. B. Katehi, S. A. Alterovitz, G. E.
Ponchak, K. M. Strohm and J.-F. Luy, ”Ku-band (12.6 GHz) SiGe/Si high-power
heterojunction bipolar transistors,” Electron. Lett., vol. 37, no. 18, pp. 1140-1142,
Aug. 2001.
[38] Z. Ma, S. Mohammadi, P. Bhattacharya, L. P. B. Katehi, S. A. Alterovitz and G. E.
Ponchak, ”A high power and high gain X-band Si/SiGe/Si heterojunction bipolar
transistor,” IEEE Trans. on Microw. Theory Tech., vol. 50, no. 4, pp. 1101-1108,
Apr. 2002.
[39] Z. Ma, N. Jiang, G. Wang and S. A. Alterovitz, “An 18 GHz, 300 mW SiGe
power HBT,” IEEE Electron Device Lett., vol. 26, no. 6, pp. 381-383, Jun. 2005.
[40] M. Racanelli, and P. Kempf, “SiGe BiCMOS technology for RF circuit
applications,” IEEE Trans. Electron Devices, vol. 52, no. 7, pp. 1259-1270, Jul.
2005.
[41] H. Bennett, R. Brederlow, J. Costa, P. Cottrell, W. Huang, A. Immorlica, Jr., J.
Mueller, M. Racanelli, H. Shichijo, C. Weitzel, and B. Zhao, “Device and
technology evolution for Si-based RF integrated circuits,” IEEE Trans. Electron
Devices, vol. 52, no. 7, pp. 1235-1258, Jul. 2005.
[42] R. Ludwig and P. Bretchko, RF circuit design - theory and applications, Upper
38
Saddle River, NJ: Prentice Hall, 2000.
[43] F. Raab, P. Asbeck, S. Cripps, P. Kenington, Z. Popovic, N. Pothecary, J. Sevic
and Nathan O. Sokal, “RF and microwave power amplifier and transmitter
technologies-Part 1”, High Frequency Electronics, pp. 22-54, 2003.
[44] M. Racanelli, and P. Kempf, “SiGe BiCMOS technology for communication
products,” IEEE 2003 Custom Integrated Circuits Conference, pp. 331-334,
2003.