Embodiments of the present disclosure relate generally to the field of circuits, and more particularly to field effect transistors (FETs).
A power amplifier (PA) such as a complementary metal-oxide-semiconductor (CMOS)-based PA may offer cost benefits, and lend itself well for monolithic integration of analog and digital circuits. Specifically, competing heterojunction bipolar transistor (HBT) and gallium arsenide (GaAs) technologies may cost more to implement per given area than a CMOS-based PA, and may not be well suited for digital and complex analog control circuits. Specifically, in HBT and GaAs PA technologies, CMOS die are often used only for digital and analog control. However, the introduction of the CMOS die to the HBT or GaAs PA may end up increasing the cost of the PA that includes the CMOS die due to extra assembly steps required to make the PA. Additionally, this mixed technology may provide practical limits on input and outputs (I/Os) and may limit the complexity of control functions. Additionally, the overall linearity and power-added efficiency (PAE) of CMOS PAs may not match that of GaAs HBT PAs.
Embodiments are illustrated by way of example and not by way of limitation in the figures of the accompanying drawings, in which like references indicate similar elements and in which:
Embodiments include an apparatus, system, and method related to a body driven FET. In embodiments, a body terminal of the FET may be electrically coupled with a source terminal of a second FET, and a drain terminal of the FET may be electrically coupled with a drain terminal of the second FET. Other embodiments may be described.
In embodiments, the disclosure may increase the linear power and efficiency of the FET. A typical 2 decibel (dB) increase in linear power and 6% improvement in power-added efficiency may be achieved in embodiments of this disclosure.
The circuit described herein may operate by making corrections as a function of output power to a body voltage of the FET to improve AM-AM and AM-PM distortion as described in detail below. Dynamic bias changes may be made to the RF transistor bias of the FET to increase current, which may avoid compression of the FET as RF power is increased by controlling the body terminal of the FET.
In embodiments of the disclosure, the second FET may be used to lower the saturation voltage or knee region of the voltage between the drain terminal and source terminal (VDS) of the first FET as the RF output power increases, while also making dynamic bias adjustments. Generally, the threshold voltage (VT) of the first FET may be decreased as RF power to the first FET is increased. This change in VT may simultaneously increase current and lower the saturation voltage or knee region of VDS when higher output powers are used. Additionally, embodiments of the disclosure may change the body voltage of the FET to achieve bias control and VT modulation to lower the minimum allowable VDS excursions.
Various aspects of the illustrative embodiments will be described using terms commonly employed by those skilled in the art to convey the substance of their work to others skilled in the art. However, it will be apparent to those skilled in the art that alternate embodiments may be practiced with only some of the described aspects. For purposes of explanation, specific devices and configurations are set forth in order to provide a thorough understanding of the illustrative embodiments. However, it will be apparent to one skilled in the art that alternate embodiments may be practiced without the specific details. In other instances, well-known features are omitted or simplified in order not to obscure the illustrative embodiments.
Further, various operations will be described as multiple discrete operations, in turn, in a manner that is most helpful in understanding the present disclosure; however, the order of description should not be construed as to imply that these operations are necessarily order dependent. In particular, these operations need not be performed in the order of presentation.
The phrase “in one embodiment” is used repeatedly. The phrase generally does not refer to the same embodiment; however, it may. The terms “comprising,” “having,” and “including” are synonymous, unless the context dictates otherwise.
In providing some clarifying context to language that may be used in connection with various embodiments, the phrases “NB” and “A and/or B” mean (A), (B), or (A and B); and the phrase “A, B, and/or C” means (A), (B), (C), (A and B), (A and C), (B and C) or (A, B and C).
The term “coupled with,” along with its derivatives, may be used herein. “Coupled” may mean one or more of the following. “Coupled” may mean that two or more elements are in direct physical or electrical contact. However, “coupled” may also mean that two or more elements indirectly contact each other, but yet still cooperate or interact with each other, and may mean that one or more other elements are coupled or connected between the elements that are said to be coupled with each other.
As used herein, “electrically coupled” may mean that two elements are able to transfer one or more electrical signals between one another. However, unless indicated otherwise, two elements may be electrically coupled without being directly physically coupled to one another. In some embodiments, one or more additional electrical circuit components (i.e., a resistor, a capacitor, or some other component) may be electrically between two components that are considered to be “electrically coupled.”
Generally, for modulated signals, PA linearity may correlate to amplitude/amplitude (AM-AM) and amplitude/phase (AM-PM) characteristics of the PA. AM-AM characteristics and AM-PM characteristics may refer to the transformation of input amplitude variations of a FET into variations of the output amplitude and phase of the FET, respectively. For an ideal PA, input variations will result in a constant or near-constant (amplification factor) output variation and a constant or near-constant phase delay of the FET, regardless of input power to the FET. Specifically, AM-AM distortion may refer to the PA characteristic wherein PA output power deviates from the expected constant amplification factor as PA input power increases. AM-AM distortion may generally be attributed to or related to a phenomenon which may be known as PA saturation or compression.
AM-PM distortion may refer to the change of output phase of the FET as the PA input power to the FET is increased. AM-PM distortion may be related to a phenomenon wherein a PA phase response of a FET may be non-constant due to internal device capacitance change as power is changed. In some cases, reducing AM-AM and AM-PM distortion of a PA may improve linearity performance of the PA. This disclosure relates to dynamically biasing the body of a silicon on insulator (SOI) field effect transistor (FET), and specifically an SOI metal-oxide-semiconductor FET (MOSFET) to minimize AM-AM and AM-PM distortion as RF power of the FET is increased.
In embodiments, the body bias of the RF power FET may change as a function of output power to significantly correct for nonlinearities while maintaining or improving efficiency as compared to a similar RF power set that is not using the same technique. Specifically, as output power is increased, the average DC voltage of the body terminal of the RF power FET may increase with respect to the source terminal of the RF power FET, which in turn may increase the current and operating point of the RF power FET to prevent compression. Generally, the drain current of a FET may be defined by:
Where λ may refer to an early effect or channel length modulation coefficient of the FET; VTB may refer to a threshold voltage of the FET at a body to source bias point; VBS may refer to a body to source voltage of the FET; VDS may refer to a drain to source voltage of the FET; VAS may refer to a gate to source voltage of the FET; VTO may refer to a threshold voltage of the FET at zero source body voltage; γ may refer to a body effect parameter of the FET; 2φ may refer to a surface body potential of the FET at zero source body voltage; and ∝ may be equal to
Further, Cox may refer to a gate oxide process parameter of the FET, W may refer to width of the gate of the FET, and L may refer to length of the gate of the FET.
As can be seen from the above equations, as VBS is increased, VTB may be decreased. From equation 1, it can also be noted that with all the other parameters constant except for VTB, drain current (ID) may increase as VTB decreases. Therefore, modulating the body of the FET to the source of the FET may dynamically increase or decrease current going through the FET as needed. Specifically, at higher power levels, VTB may decrease, which may result in extra current and power going through the FET. Conversely, at low power levels, VTB may increase, which may result in less current and/or power flowing through the FET.
At low power levels, efficiency of the RF power FET may be maintained, because the average body voltage (Vbody) of the FET, which may affect the power consumed by the FET, may be low. However, at higher power levels, the body voltage of that RF power FET may be increased. A circuit that quickly responds to RF power changes of the FET to maintain linearity in non-constant envelope conditions may therefore be desirable. An example of such a circuit 100 is depicted in
The circuit 100 may include a main RF power FET 103. The FET 103 may include a drain terminal 106, body terminal 109, source terminal 112, and gate terminal 115. In embodiments, the source terminal 112 may be coupled with ground. The gate terminal 115 of the FET 103 may be coupled with an RF input 118. In some embodiments, an input matching network (IMN) 121 may be positioned between the RF input 118 and gate terminal 115. In embodiments, the gate terminal 115 may further be coupled with a gate bias 124.
The circuit 100 may further include FET 127. The FET 127 may include a drain terminal 133, a source terminal 130, and gate terminal 136. In embodiments, the body terminal 109 of the FET 103 may be coupled with the source terminal 130 of FET 127. In some embodiments, a resistor 139 may be placed between the body terminal 109 and the source terminal 130. The resistor 139 may be on the order of 10 to 10000 ohms. Additionally, as shown in
In some embodiments, the source terminal 130 and the body terminal 109 may be coupled with ground 142, and in some embodiments a resistor 148 may be placed between ground 142 and the source terminal 130/body terminal 109. In embodiments, the resistor 148 may be on the order of 10 to 10000 ohms.
Additionally, in some embodiments, the source terminal 130 and the body terminal 109 may be coupled with ground 175, and in some embodiments a capacitor 145 may be placed between ground 175 and the source terminal 130/body terminal 109. In embodiments, the capacitor 145 may be on the order of 1 pico to 1 micro Farad.
In some embodiments, the gate terminal 136 may be coupled with drain terminals 133 and 106, as shown in
In some embodiments, the gate terminal 136, drain terminal 133, and drain terminal 106 may further be coupled with a voltage source 151. In some embodiments, a resistor 154 may be positioned between the voltage source 151, and the gate terminal 136/drain terminal 133/drain terminal 106 as shown in
In embodiments, the values of resistors 139, 148, 154 may be dependent on one or more factors that are related to device size and/or application. For example, the value of resistor 139 may be dependent on a physical size of FET 103, and/or the frequency or frequency range of the RF signal being propagated through circuit 100.
Similarly, the values of capacitors 145 and/or 157 may be generally related to or based on one or more factors such as device size and/or application. For example, if the circuit 100 is being used in a PA that is being used primarily for relatively low frequency RF applications, then the value of capacitor 145 may be on the order of approximately 1 micro Farad, and the value of capacitor 157 may be on the order of approximately 100 pico Farads. By contrast, if the circuit 100 is being used in a PA that is being used primarily for relatively high frequency RF applications, then the value of capacitor 145 may be on the order of approximately 1 pico Farad, and the value of capacitor 157 may be on the order of approximately 1 pico Farad.
In some embodiments, the circuit 100 may optionally include additional RF power FETs such as FETs 160 and 163. In some embodiments, the FETs 160 and 163 may further be coupled with the gate bias 124. In some embodiments, the body terminal 172 of the additional FETs may be coupled with the source terminal 169 of the additional FETs, such as shown with FET 160. In some embodiments, the gate bias may be different for one or more of FETs 103, 163, and 160 than another one of FETs 103, 163, and 160.
In embodiments, the RF output of the circuit 100 may be or be coupled with the drain terminal of one or more of FETs 103, 163, or 160. Specifically, as shown in
It will be understood the circuit 100 is depicted merely as one example of an embodiment of the present disclosure. In some embodiments, circuit 100 may have additional or omitted items than shown in
As shown in
The circuit 100 may operate according to one or more of the following principles. Specifically, the increasing body voltage of FET 103 with the increasing power may dynamically lower the saturation region of VDS, and enable the FET 103 to operate with lower VDS headroom without shutting off as RF power is increased. This lowered saturation region of VDS may allow for extended range of gain compensation correction to avoid AM-AM nonlinearity. Specifically, the saturation region for FET 103 may be defined as VGS>VT, and VDS (VGS−VT), where VGS is the voltage between the gate terminal 115 and the source terminal 112 of the FET 103, VDS is the voltage between the drain terminal 106 and the source terminal 112 of the FET 103, and VT is a threshold voltage. In embodiments, by dynamically lowering VT through body modulation, the FET 103 may operate at a lower VDS without entering the triode region, as depicted in greater detail with respect to
Cgs of the RF power FET 103 may contribute to AM-PM variation, and changes significantly as the FET 103 goes from the saturation mode of operation to the triode or linear region. Specifically, Cgs may be defined as approximately half of Cox*W*L for the triode mode (Equation 4) and approximately two thirds of Cox*W*L for the saturated mode (Equation 5), wherein Cox, W, and L are as described above. Although the FET 103 may work in the saturation region mainly at peak power excursions VDS may drop and the triode region may be approached or entered, which may reduce Cgs.
The equations 4 and 5 indicate approximate values for Cgs in different modes of operation that may approximate the extremes of the signal swing due to operation of circuit 100. Generally, Cgs may not change spontaneously between the two cases, and there may be Cgs reduction even before the triode or linear region is fully reached. Again, dynamically changing the body voltage of the FET 103 to increase as power to the FET 103 is increased may minimize Cgs change due to the FET 103 entering or approaching the linear region and AM-PM changes extend to higher RF drive power levels. With the delayed onset of AM-PM variation, improved linearity may be achieved via extended output power levels.
In embodiments, the correction voltage is applied to the body terminal 109 of the FET 103 as a function of output power that may be empirically extracted, and a third order polynomial function of the form Vbody=kb+abx+bbx2+cbx3 (Equation 6) may be considered a reasonable fit for the correction voltage. Specifically,
In equation 6, the term kb may be a static constant; ab, bb, and cb may be coefficients of the equation; and x may be the output power level. Generally, it may be desirable for the body voltage to follow the third order polynomial for improved performance of the FET 103. The third order polynomial function may increase the body voltage at body terminal 109 adequately as a function of output power to avoid compression of the FET 103 without overcompensation.
The third order polynomial equation could be implemented with a power detector and log amps that are able to respond quickly to changes in the circuit; however, doing so could add complexity and cost and increase power consumption of the circuit 100. By contrast, the circuit 100 depicts an efficient implementation that utilizes the drain current of the FET 103 to generate the body driving voltage needed, thereby eliminating the overhead of detectors and log amps. To improve efficiency, FETs such as FETs 103, 160, and 163 may be biased in class AB. A nonlinearized FET's current bias in class AB may be modeled by a third order polynomial such as Id=kd+adx+bdx2+cdx3 (Equation 7) where Id may be the drain current as described above; kd may be static constant; ad, bd, and cd may be coefficients; and x may be the output power level of the FET 103.
As described above, the third order polynomial behavior of a representative nonlinearized FET drain current may be modeled by a third order polynomial as shown in the example depicted in
From equation 6 and equation 7, it may be readily seen that the term kb may be easily implemented from kd with a DC offset as a correction. The coefficient terms of equations 6 and 7 may be produced based on the drain current.
Returning to
In some embodiments, FET 127 may be scaled to be smaller than FET 103. For example, FET 127 may be at least an order of magnitude smaller than FET 103 so as to not consume as much power or area as FET 103.
In embodiments, the capacitor 157 may serve to couple RF energy to the gate of FET 127 to aid the third order polynomial fit that is a function of power. Specifically, capacitor 157 may serve to match the inflection point in the third order polynomial. In some embodiments, as described above, capacitor 157 may be omitted from circuit 100. In some embodiments, resistor 139 may limit current entering the body terminal 109 of FET 103, and in some embodiments may be omitted without affecting the example third order polynomial fit described above. In embodiments, resistor 148 may serve as a negative feedback mechanism to ensure the current of FET 127 self-limits; hence the voltage applied to the body terminal 109 similarly self-limits. When the circuit 100 is operating at low output power levels, the FET 127 may operate in a Class A-like fashion, and its output current may have little to no DC content as shown in
However, as output power increases, FET 127 may not be able to conduct current when FET 103 is strongly conducting current, and will produce an asymmetrical current pulse correlating to the current of FET 103. As the waveform is asymmetrical, it may have at least some DC content, and may be used to drive the body of FET 103.
Capacitor 145 may be used to filter the voltage generated to drive the body of FET 103, but in some embodiments capacitor 145 may be omitted because an internal body to source capacitance may exist in FET 103. Furthermore, in some embodiments, omitting capacitor 145 and allowing body voltage variation of the circuit 100 during an RF cycle may not be detrimental, and may be used as a design tool to change dynamic characteristics of the FET 103.
Generally, the efficiency and output power of a FET such as FET 103 may be limited by the presence of the knee voltage. With a smaller knee voltage, the FET may handle a larger voltage swing and peak current than a FET with a larger knee voltage. The knee voltage of the FET may be reduced when the power load is increased, thus enabling extended power and improved efficiency in our claim.
In mathematical form, the efficiency reduction factor due to knee voltage may be considered to be approximately equal to 1/(1+VK/VDS) where VDS indicates the DC voltage from the drain to the source of the FET and VK is the knee voltage. It can be seen from the above equation that with no knee voltage (VK=0), the efficiency of the FET may be only minimally impacted, or not impacted at all. Furthermore, it can be seen that at low supply voltages (VDS), the knee voltage may have a larger impact on efficiency. In case of SOI CMOS FETs, VDS may be kept low due to breakdown issues related to the FETs, and the knee voltage may have a significant impact.
The circuit 100 may be incorporated into a variety of systems. A block diagram of an example system 1000 is illustrated in
The PA module 1002 may receive an RF input signal, RFin, from the transceiver 1004. The PA module 1002 may amplify the RF input signal, RFin, to provide the RF output signal, RFout. The RF input signal, RFin, and the RF output signal, RFout, may both be part of a transmit chain, respectively noted by Tx-RFin and Tx-RFout in
The amplified RF output signal, RFout, may be provided to the ASM 1006, which effectuates an over-the-air (OTA) transmission of the RF output signal, RFout, via an antenna structure 1008. The ASM 1006 may also receive RF signals via the antenna structure 1008 and couple the received RF signals, Rx, to the transceiver 1004 along a receive chain.
In various embodiments, the antenna structure 1008 may include one or more directional and/or omnidirectional antennas, including, e.g., a dipole antenna, a monopole antenna, a patch antenna, a loop antenna, a microstrip antenna or any other type of antenna suitable for OTA transmission/reception of RF signals.
The system 1000 may be suitable for any one or more of terrestrial and satellite communications, radar systems, and possibly in various industrial and medical applications. More specifically, in various embodiments, the system 1000 may be a selected one of a radar device, a satellite communication device, a mobile computing device (e.g., a phone, a tablet, a laptop, etc.), a base station, a broadcast radio, or a television amplifier system.
Although the present disclosure has been described in terms of the above-illustrated embodiments, it will be appreciated by those of ordinary skill in the art that a wide variety of alternate and/or equivalent implementations calculated to achieve the same purposes may be substituted for the specific embodiments shown and described without departing from the scope of the present disclosure. Those with skill in the art will readily appreciate that the teachings of the present disclosure may be implemented in a wide variety of embodiments. This description is intended to be regarded as illustrative instead of restrictive.
Number | Name | Date | Kind |
---|---|---|---|
3166719 | Wiencek | Jan 1965 | A |
6229405 | Hashimoto | May 2001 | B1 |
20090251213 | Mangudi | Oct 2009 | A1 |
20120056673 | Boor | Mar 2012 | A1 |
20120064852 | Lee | Mar 2012 | A1 |
Entry |
---|
Sowlati, Tirdad, “A 2.4-GHz 0.18-um CMOS Self-Biased Cascode Power Amplifier”, IEEE Journal of Solid-State Circuits, vol. 38, No. 8, Aug. 2003. |