Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
TRANSIENT STABILIZED SOI FETS
Document Type and Number:
WIPO Patent Application WO/2018/212975
Kind Code:
A1
Abstract:
Integrated circuits (ICs) that avoid or mitigate creation of changes in accumulated charge in a silicon-on-insulator (SOI) substrate, particularly an SOI substrate having a trap rich layer. In one embodiment, a FET is configured such that, in a standby mode, the FET is turned OFF while maintaining essentially the same VDS as during an active mode. In another embodiment, a FET is configured such that, in a standby mode, current flow through the FET is interrupted while maintaining essentially the same VGS as during the active mode. In another embodiment, a FET is configured such that, in a standby mode, the FET is switched into a very low current state (a "trickle current" state) that keeps both VGS and VDS close to their respective active mode operational voltages. Optionally, S-contacts may be formed in an IC substrate to create protected areas that encompass FETs that are sensitive to accumulated charge effects.

Inventors:
ENGLEKIRK, Robert Mark (13 Winter Cress, Littleton, Colorado, 80127, US)
BARGROFF, Keith (2616 Snowsdrop Street, San Diego, California, 92105, US)
MURPHY, Christopher C. (840 Warwick Lane, Lake Zurich, Illinois, 60047, US)
RANTA, Tero Tapio (11988 Halfoak Terrace, San Diego, California, 92128, US)
WILLARD, Simon Edward (168 West Yale Loop, Irvine, California, 92604, US)
Application Number:
US2018/030179
Publication Date:
November 22, 2018
Filing Date:
April 30, 2018
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
PSEMI CORPORATION (9380 Carroll Park Drive, San Diego, California, 92121, US)
International Classes:
H03K17/14; H01L21/762; H01L29/00; H01L29/786; H03K17/0412; H03K17/0416; H03K17/042
Foreign References:
US20060011807A12006-01-19
US20040077327A12004-04-22
US20160035899A12016-02-04
US201615268229A2016-09-16
US7910993B22011-03-22
US201514964412A2015-12-09
Attorney, Agent or Firm:
CASH, Brian et al. (Steinfl + Bruno LLP, 155 N. Lake Avenue Suite 70, Pasadena California, 91101, US)
Download PDF:
Claims:
CLAIMS

WHAT IS CLAIMED IS:

1 . An integrated circuit susceptible to accumulated charge and fabricated on a silicon-on- insulator (SOI) substrate, including at least one field effect transistor (FET) having a VDS characteristic and configured such that, in a standby mode, the FET is turned OFF while maintaining essentially the same VDS characteristic as during an active mode, thereby eliminating or reducing changes in accumulated charge.

2. A circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, the circuit including:

(a) at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, and a signal path through the FET between the drain and the source;

(b) a first switch coupled to the gate of the FET, configured to selectively couple the gate to one of a bias voltage or a standby voltage source; and

(c) a second switch coupled to the signal path of the FET, configured to selectively couple the signal path to one of a load or a pseudo-load voltage source;

wherein, in an active mode, the first switch couples the gate of the FET to the bias voltage and the second switch couples the signal path of the FET to the load; and

wherein, in a standby mode, the first switch couples the gate of the FET to the standby voltage source and the second switch couples the signal path of the FET to the pseudo-load voltage source, thereby maintaining essentially the same VDS characteristic as during the active mode, and thereby eliminating or reducing changes in accumulated charge.

3. The invention of claim 2, wherein the standby mode pseudo-load voltage source outputs a voltage approximately equal to the voltage present on the drain of the FET during active mode operation.

4. An integrated circuit susceptible to accumulated charge and fabricated on a silicon-on- insulator (SOI) substrate, including at least one field effect transistor (FET) having a VQS characteristic and configured such that, in a standby mode, current flow through the FET is interrupted while maintaining essentially the same VGS characteristic as during an active mode, thereby eliminating or reducing changes in accumulated charge.

A circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, the circuit including, the circuit including:

(a) at least one field effect transistor (FET) having a drain, a source, a gate, a VGS characteristic, and a signal path through the FET between the drain and the source; and

(b) a switch coupled to the signal path of the FET, configured to selectively couple the signal path to a load or interrupt current flow through the signal path of the FET;

wherein, in an active mode, the switch couples the signal path of the FET to the load; and wherein, in a standby mode, the switch interrupts current flow through the signal path of the FET, thereby maintaining essentially the same VGS characteristic as during the active mode, and thereby eliminating or reducing changes in accumulated charge.

An integrated circuit susceptible to accumulated charge and fabricated on a silicon-on- insulator (SOI) substrate, including at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, a VGS characteristic, and a signal path through the FET between the drain and the source, and configured such that, in a standby mode, the FET is switched into a very low current state with respect to current flow through the signal path that keeps both the VGS characteristic and the VDS characteristic of the FET close to respective active mode operational voltages for the VQS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge.

7. A circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, the circuit including:

(a) at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, a VQS characteristic, and a signal path through the FET between the drain and the source; and

(b) a switch coupled to the signal path of the FET, configured to selectively couple the signal path to a normal current flow path or to a trickle current path;

wherein, in an active mode, the switch couples the signal path of the FET to the normal current flow path; and

wherein, in a standby mode, the switch couples the signal path of the FET to the trickle current path such that both the VQS characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the VQS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge.

8. The invention of claim 7, wherein the trickle current path has a high resistance relative to the normal current flow path.

9. The invention of claim 7, wherein the trickle current path includes a regulated current source that allows only a trickle of current to flow through the FET relative to the normal current flow path.

10. A circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, the circuit including:

(a) at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, a VQS characteristic, and a signal path through the FET between the drain and the source; and

(b) a switch coupled to the signal path of the FET, configured to selectively couple the signal path to a load resistance or to a high resistance;

wherein, in an active mode, the switch couples the signal path of the FET to the load resistance; and

wherein, in a standby mode, the switch couples the signal path of the FET to the high resistance, such that both the VQS characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the VQS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge.

1 1. The invention of claim 10, wherein the high resistance is at least about 100 times greater than the load resistance.

12. A circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, the circuit including:

(a) at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, a VQS characteristic, and a signal path through the FET between the drain and the source; and

(b) a switch coupled to the signal path of the FET, configured to selectively couple the signal path to a first current source or to a second current source, the second current source providing a lower current than the first current source;

wherein, in an active mode, the switch couples the signal path of the FET to the first current source; and

wherein, in a standby mode, the switch couples the signal path of the FET to the second current source, such that both the VQS characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the VQS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge.

13. An integrated circuit susceptible to accumulated charge and fabricated on a silicon-on- insulator (SOI) substrate, including:

(a) at least one field effect transistor (FET) having a VDS characteristic and a VQS characteristic; and

(b) at least one switch coupled to the FET and configured to be set to a standby mode in which no significant current flows through the at least one FET while maintaining essentially the same VDS characteristic and VQS characteristic for the at least one FET as during an active mode, thereby eliminating or reducing changes in accumulated charge.

14. The invention of claims 1 , 2, 4, 5, 6, 7, 10, 12, or 13, wherein the SOI substrate includes a trap rich layer susceptible to accumulated charge in or near such trap rich layer.

15. The invention of claims 1 , 2, 4, 5, 6, 7, 10, 12, or 13, further including at least one substrate contact near at least one FET.

16. The invention of claims 1 , 2, 4, 5, 6, 7, 10, 12, or 13, further including at least a partial ring of substrate contacts around at least one FET.

17. A method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on a silicon-on-insulator (SOI) substrate, including:

(a) providing at least one field effect transistor (FET) having a VDS characteristic; and

(b) configuring the at least one FET such that, in a standby mode, the FET is turned OFF while maintaining essentially the same VDS characteristic as during an active mode, thereby eliminating or reducing changes in accumulated charge.

18. A method for eliminating or reducing changes in accumulated charge in a circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, including:

(a) providing at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, and a signal path through the FET between the drain and the source;

(b) coupling a first switch to the gate of the FET and configuring the first switch to selectively couple the gate to one of a bias voltage or a voltage source;

(c) coupling a second switch to the signal path of the FET, and configuring the second switch to selectively couple the signal path to one of a load or a pseudo-load voltage source;

(d) in an active mode, setting the first switch to couple the gate of the FET to the bias voltage and setting the second switch to couple the signal path of the FET to the load; and

(e) in a standby mode, setting the first switch to couple the gate of the FET to the voltage source and setting the second switch to couple the signal path of the FET to the pseudo- load voltage source, thereby maintaining essentially the same VDS characteristic as during the active mode, and thereby eliminating or reducing changes in accumulated charge.

19. The method of claim 18, wherein the standby mode pseudo-load voltage source outputs a voltage approximately equal to the voltage present on the drain of the FET during active mode operation.

20. A method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on a silicon-on-insulator (SOI) substrate, including:

(a) providing at least one field effect transistor (FET) having a VGS characteristic; and

(b) configuring the at least one FET such that, in a standby mode, current flow through the FET is interrupted while maintaining essentially the same VGS characteristic as during an active mode, thereby eliminating or reducing changes in accumulated charge.

21. A method for eliminating or reducing changes in accumulated charge in a circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, including:

(a) providing at least one field effect transistor (FET) having a drain, a source, a gate, a VQS characteristic, and a signal path through the FET between the drain and the source;

(b) coupling a switch to the signal path of the FET and configuring the switch to selectively couple the signal path to a load or interrupt current flow through the signal path of the FET;

(c) in an active mode, setting the switch to couple the signal path of the FET to the load; and

(d) in a standby mode, setting the switch to interrupt current flow through the signal path of the FET, thereby maintaining essentially the same VQS characteristic as during the active mode, and thereby eliminating or reducing changes in accumulated charge.

22. A method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on a silicon-on-insulator (SOI) substrate, including:

(a) providing at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, a VQS characteristic, and a signal path through the FET between the drain and the source; and

(b) configuring the at least one FET such that, in a standby mode, the FET is switched into a very low current state with respect to current flow through the signal path that keeps both the VQS characteristic and the VDS characteristic of the FET close to respective active mode operational voltages for the VQS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge.

23. A method for eliminating or reducing changes in accumulated charge in a circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, including:

(a) providing at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, a VGS characteristic, and a signal path through the FET between the drain and the source;

(b) coupling a switch to the signal path of the FET and configuring the switch to selectively couple the signal path to a normal current flow path or to a trickle current path;

(c) in an active mode, setting the switch to couple the signal path of the FET to the normal current flow path; and

(d) in a standby mode, setting the switch to couple the signal path of the FET to the trickle current path such that both the VGS characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the VGS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge.

24. The method of claim 23, wherein the trickle current path has a high resistance relative to the normal current flow path.

25. The method of claim 23, further including providing a regulated current source coupled to the trickle current path that allows only a trickle of current to flow through the FET relative to the normal current flow path.

26. A method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, including:

(a) providing at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, a VQS characteristic, and a signal path through the FET between the drain and the source;

(b) coupling a switch to the signal path of the FET, configured to selectively couple the signal path to a load resistance or to a high resistance;

(c) in an active mode, setting the switch to couple the signal path of the FET to the load resistance; and (d) in a standby mode, setting the switch to couple the signal path of the FET to the high resistance, such that both the VGS characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the VGS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge.

27. The method of claim 26, wherein the high resistance is at least about 100 times greater than the load resistance.

28. A method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on a silicon-on-insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge, including:

(a) providing at least one field effect transistor (FET) having a drain, a source, a gate, a VDS characteristic, a VGS characteristic, and a signal path through the FET between the drain and the source;

(b) coupling a switch coupled to the signal path of the FET, configured to selectively couple the signal path to a first current source or to a second current source, the second current source providing a lower current than the first current source;

(c) in an active mode, setting the switch to couple the signal path of the FET to the first current source; and

(d) in a standby mode, setting the switch to couple the signal path of the FET to the second current source, such that both the VQS characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the VQS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge.

29. A method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on a silicon-on-insulator (SOI) substrate, including:

(a) providing at least one field effect transistor (FET) having a VDS characteristic and a VQS characteristic; and

(b) configuring the at least one FET such that, in a standby mode, no significant current flows through the at least one FET while maintaining essentially the same VDS characteristic and VQS characteristic for the at least one FET as during an active mode, thereby eliminating or reducing changes in accumulated charge.

30. The method of claims 17, 18, 20, 21 , 22, 23, 26, 28, or 29, wherein the SOI substrate includes a trap rich layer susceptible to accumulated charge in or near such trap rich layer.

31. The method of claims 17, 18, 20, 21 , 22, 23, 26, 28, or 29, further including providing at least one substrate contact near at least one FET.

32. The method of claims 17, 18, 20, 21 , 22, 23, 26, 28, or 29, further including providing at least a partial ring of substrate contacts around at least one FET.

Description:
Transient Stabilized SOI FETs

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims priority to US Patent Application 15/600,579, filed on May 19, 2017 (Attorney Docket No. PER-231-PAP), the disclosure of which is incorporated herein by reference in its entirety. This application may be related to U.S. Patent Application No. 15/600,588, filed May 19, 2017, entitled "Managed Substrate Effects For Stabilized SOI FETs" (Attorney Docket No. PER-232-PAP), to U.S. Patent Application No. 14/964,412, filed December 9, 2015, entitled "S-Contact for SOI", to U.S. Patent Application No. 15/385,618, filed December 20, 2016, entitled "Systems, Methods and Apparatus for Enabling High Voltage Circuits", and to U.S. Patent Application No. 15/268,297, filed September 16, 2016, entitled "Standby Voltage Condition for Fast RF Amplifier Bias Recovery", all assigned to the assignee of the present invention and hereby incorporated by reference.

BACKGROUND

(1) Technical Field

[0002] The invention relates to electronic circuits, and more particularly to radio frequency circuitry fabricated with silicon-on-insulator technology.

(2) Background

[0003] Modern radio frequency (RF) transceivers have become ubiquitous, with manifestations found in cellular telephones, wireless networks (e.g. , WiFi, Bluetooth), laptop computers, tablets, computer peripheral devices (e.g. , printers, keyboards, mice), etc. That ubiquity has increased demand for smaller and lower power RF transceiver circuit designs.

[0004] FIG. 1 is a block diagram of a typical prior art transceiver 100 such as the type that might be used in a cellular telephone. As illustrated, the transceiver 100 includes a mix of RF analog circuitry for directly conveying and/or transforming signals on an RF signal path, non-RF analog circuity for operational needs outside of the RF signal path (e.g. , for bias voltages and switching signals), and digital circuitry for control and user interface requirements. In this example, a receiver path Rx includes RF Front End, IF Block, Back-End, and Baseband sections (noting that in some implementations, the lines between sections may be blurred).

[0005] The receiver path Rx receives over-the-air RF signals through an antenna 102 and a switching unit 104, which may be implemented with active switching devices (e.g. , field effect transistors or FETs), or with passive devices that implement frequency-domain multiplexing, such as a diplexer or duplexer. An RF filter 106 passes desired received RF signals to a low noise amplifier (LNA) 108, the output of which is combined in a mixer 1 10 with the output of a first local oscillator 1 12 to produce an intermediate frequency (IF) signal. The IF signal may be amplified by an IF amplifier 1 14 and subjected to an IF filter 1 16 before being applied to a demodulator 1 18, which may be coupled to a second local oscillator 120. The output of the demodulator 1 18 is transformed to a digital signal by an analog-to-digital converter 122 and provided to one or more system components 124 (e.g. , a video graphics circuit, a sound circuit, memory devices, etc.). The converted digital signal may represent, for example, images (video or still), sounds, or symbols, such as text or other characters.

[0006] In the illustrated example, a transmitter path Tx includes Baseband, Back-End, IF Block, and RF Front End sections (again, in some implementations, the lines between sections may be blurred). Digital data from one or more system components 124 is transformed to an analog signal by a digital-to-analog converter 126, the output of which is applied to a modulator 128, which also may be coupled to the second local oscillator 120. The output of the modulator 128 may be subjected to an IF filter 130 before being amplified by an IF amplifier 132. The output of the IF amplifier 132 is then combined in a mixer 134 with the output of the first local oscillator 1 12 to produce an RF signal. The RF signal may be amplified by a driver 136, the output of which is applied to a power amplifier (PA) 138. The amplified RF signal may be coupled to an RF filter 140, the output of which is coupled to the antenna 102 through the switching unit 104.

[0007] The operation of the transceiver 100 is controlled by a microprocessor 142 in known fashion, which interacts with system control components (e.g. , user interfaces, memory/storage devices, application programs, operating system software, power control, etc.). In addition, the transceiver 100 will generally include other circuitry, such as bias circuitry 146 (which may be distributed throughout the transceiver 100 in proximity to transistor devices), electro-static discharge (ESD) protection circuits (not shown), testing circuits (not shown), factory programming interfaces (not shown), etc.

[0008] As should be apparent to one of ordinary skill in the art, some components of the transceiver 100 may be positioned in a different order (e.g. , filters), and some of the components may be implemented as distinct integrated circuits (ICs). For example, the RF front end may be separately embodied from the rest of the circuitry (although there is a trend to try to integrate as much of the transceiver functionality as possible in a single IC). Other components can be (and usually are) added (e.g. , additional filters, impedance matching networks, variable phase shifters/attenuators, power dividers, etc.). Importantly, in modern transceivers, there are often more than one receiver path Rx and transmitter path Tx - for example, to accommodate multiple frequencies and/or signaling modalities - so switching and control becomes more complex.

[0009] A challenge with electronic circuit design in general is that ideal components do not exist, particularly when dealing with RF signals. For example, the operational characteristics of many passive and active components in an RF signal path are frequency dependent. As another example, a significant problem with RF circuit design is eliminating or controlling unwanted cross-effects ("cross-talk") and self-effects, such as parasitic capacitances and inductances, undesired signal coupling, performance changes due to environment temperature changes as well as self-heating, and others. The problems of RF design become more acute when embodying RF circuits as ICs, where components materials, circuit juxtaposition, and power constraints add to the difficulties of optimizing operational parameters for all components. As one example, FETs are inherently designed to operate with fields, but fields do not have distinct edges and often cause cross-effects and self-effects. As another example, FETs have operating parameters that are subject to process, voltage, and temperature (PVT) variations. Accordingly, RF circuit designs embodied as ICs generally require optimizations of some circuit parameters at the expense of other parameters.

[0010] As an example of such trade-offs, as noted above, a transceiver 100 will generally include bias circuitry. FIG. 2 is a block diagram of a prior art bias voltage generation circuit 200. In this example, a stable reference voltage source 202 provides a reference voltage VBG- The reference voltage source 202 may be, for example, a band-gap voltage reference, which is a temperature-independent voltage reference circuit widely used in ICs that produces an essentially constant voltage regardless of power supply variations, temperature changes, and load. The reference voltage VBG is applied to a voltage-to-current converter circuit 204, which essentially divides the reference voltage VBG by a resistance R to generate a reference current IREF- The reference current IREF may be applied to a bias generator circuit 206 which generates a specific bias voltage VBIAS suitable for a particular amplifier 208. The bias generator circuit 206 may be simple or complex, depending on the needs of the overall circuit; examples of a variety of bias generator circuits are described in U.S. Patent Application No. 15/268,229, filed September 16, 2016, entitled "Cascode Amplifier Bias Circuits", assigned to the assignee of the present invention and hereby incorporated by reference.

[0011] A notable characteristic of RF circuits in general is that different components may require different optimizations. For example, the amplifier 208 is in the RF signal path and is generally an analog circuit optimized for performance at RF frequencies. Conversely, while the components of the bias voltage generation circuit 200 are also generally analog circuitry, they generally do not operate at RF frequencies and they may need optimizations that differ from RF signal path components. A distinct challenge of RF circuit design is that optimization for some circuitry may adversely affect optimization of other circuitry.

[0012] In general, for most transceivers, the RF signal path components are the most important circuitry to optimize. It was recognized some time ago that semiconductor-on-insulator (SOI) IC technology is particularly useful for such optimization. An important aspect of SOI technology is that the semiconductor region in which circuitry is formed is isolated from the bulk substrate by an electrically insulating layer. An advantage of isolating circuitry from the bulk substrate is a dramatic decrease in cross-talk and parasitic capacitance, which significantly improves speed and power characteristics for RF components.

[0013] FIG. 3 is block diagram showing a typical prior art SOI IC structure 300 for a single FET. The SOI structure 300 includes a substrate 302, an insulator layer 304, and an active layer 306. The substrate 302 is typically a semiconductor material such as silicon. The insulator layer 304 is a dielectric which is often silicon dioxide formed through the oxidation of a silicon substrate 302; the insulator layer 304 is often referred to as a buried oxide (or "BOX") layer. The active layer 306 may include some combination of implants and/or layers that include dopants, dielectrics, polysilicon, metal wiring, passivation, and other materials to form active and/or passive electronic components and/or mechanical structures. For example, in the illustrated embodiment, a FET is shown comprising a source S, a drain D, and a gate G atop an insulating layer 308. Additional elements, not shown, may include contacts, conductive interconnections to other components and/or external connections, and protective coatings.

[0014] While the basic SOI IC structure 300 of FIG. 3 works reasonably well up to a point for RF circuitry such as transceiver components, as frequency increases and power specifications are reduced, cross-talk and parasitic elements again begin to affect such parameters as linearity and switching speed. Accordingly, other optimizations may be required. For example, FIG. 4 is block diagram showing an improved prior art SOI IC structure 400 for a single FET. The SOI structure 400 includes a high resistivity substrate 402, a trap rich layer 404, a BOX insulator layer 406, and an active layer 408 that includes a single example FET 410.

[0015] The high resistivity of the substrate 402 makes the impedance of the parasitic paths through the substrate 402 higher. Materials used for the high resistivity substrate 402 typically include very lightly doped silicon such that the high resistivity substrate 402 takes on some of the characteristics of an insulator. The use of high resistivity substrates alone has proven capable of extending the benefit of SOI structures for RF communication circuits by roughly two orders of frequency magnitude.

[0016] The trap rich layer 404 is a solution to another problem. Although high resistivity substrates alone are capable of reducing substrate loss when used in SOI IC structures, they are highly susceptible to another phenomenon called parasitic surface conduction. The problem of parasitic surface conduction occurs because the high resistivity substrate 402 is capable of terminating field lines, but a thin surface region of the high resistivity substrate 402 can be formed into an inversion or accumulation region as charge carriers are affected by changes in signal voltages in the active layer 408. The degree to which charge carriers in the thin surface region are displaced is directly altered by signals in the active layer 408. As a result, without the trap layer 404, the capacitance of the junction between the high resistivity substrate 402 and the active layer 408, as seen by the active layer 408, depends on the voltage applied, resulting in nonlinearity and a loss of signal purity In addition, an applied voltage can invert this interface on the side of the high resistivity substrate 402 and create a channel-like layer within the thin surface region where charge can move very easily in a lateral direction despite the fact that the substrate layer 402 is highly resistive. Therefore, this effect can also lead to signal- degrading cross-talk in RF communication circuits.

[0017] Forming a trap rich layer 404 on top of the substrate layer 402 substantially mitigates parasitic surface conduction. The trap rich layer 404 is typically formed as a layer of amorphous or polycrystalline silicon on a top surface of the substrate 402, and significantly degrades the mobility of the charge carriers in the thin surface region of the substrate 402. Since the carriers cannot travel far, the effective resistance of the substrate 402 is preserved and the capacitance as seen by the active layer 408 is not as dependent upon the signals in the active layer 408. The improved RF performance of SOI IC substrates with a trap rich layer 404 is so marked that wafers having that configuration are commercially available.

[0018] Thus, SOI based IC technology improves RF performance, to a point; high resistivity SOI substrates further improve RF performance, to a point; and high resistivity SOI substrates with trap rich layers further still improves RF performance. However, the improved performance of SOI-based RF circuitry formed in the active layer 408 above the trap rich layer 404 can adversely affect desired operational characteristics of non-RF circuitry formed in the active layer 408 above the trap rich layer 404, such as switches and bias circuits.

[0019] For example, for some components (e.g. , a power amplifier for use in a WiFi transceiver, such as one conforming to the IEEE 802.11 ax or 802.1 l ac standards), the component is active for only short bursts of time (e.g. , ~4 mS in a WiFi power amplifier) and in a low power mode (e.g. , "sleep" or "standby" mode) for most of the time to save power. In the WiFi example, the standby power consumption for a power amplifier may be very low (e.g. , < 10 μΑ), but the amplifier may be required to have a very quick sleep-to-active transition time (e.g. , < 1 μ8) and to achieve a very stable gain very soon (e.g. , < 30 μ8) after becoming active. The low power standby mode may be achieved by essentially switching some of the active circuitry OFF. More generally, many radio-based systems, particularly battery-operated systems (e.g. , cellular telephones and WiFi radios) make significant use of standby modes to conserve power usage and/or to preserve battery life. Transitioning from a standby mode to an active mode introduces transient effects, especially in bias circuits, which can take a long time to settle out. However, many RF and analog performance requirements typically do not allow a lot of recovery ("buffer") time. In general, such radio systems are designed to start transmitting a maximum amount of data as soon as possible. This requires that the radio system waken from standby mode back to active mode and be ready for full performance without significant transient effects.

[0020] Compounding the challenges of IC design is the fact that an SOI FET can develop an accumulated charge over time. For example, one type of accumulated charge depends upon the potentials at the source S, drain D, and gate G of the FET 410. More particularly, when used in certain SOI circuit implementations, conventional FETs may operate in an accumulated charge regime. This phenomenon is also referred to as a "floating body effect" or "kink effect", and more generally is the effect of dependence of the body potential of an SOI FET on the history of its biasing and carrier recombination processes. The body of the FET forms a capacitor with respect to the insulated substrate, and accordingly charge accumulates on this capacitor. The accumulated charge may cause adverse effects, such as opening of parasitic transistors in the structure and causing OFF-state leakages (resulting in higher current consumption), and also causes a history effect, which is the dependence of the threshold voltage VTH of a FET on the previous states of the FET.

[0021 ] Another aspect of accumulated charge is related to the OFF state of a FET and the resultant effect on device linearity, and is described in U.S. Patent No. 7,910,993 B2, issued March 22, 101 1 , entitled "Method and Apparatus for use in Improving Linearity of MOSFET'S using an Accumulated Charge Sink" and assigned to the assignee of the present invention.

[0022] A trap rich layer 404 on an SOI substrate 402 can be a separate source of accumulated charge. For example, when FETs in the active layer switch between ON and OFF states, the changing electrical fields accompanying such change in state can cause charge to accumulate between the trap rich layer 404 and the BOX insulator layer 406. The high impedance nature of the trap rich layer 404 hinders dissipation of that trapped accumulated charge. [0023] Regardless of source, accumulated charge can take a long time to equilibrate (e.g. , > 4mS), and can affect the intrinsic properties of a FET device (e.g. , the threshold voltage of the device, VTH)- Thus, turning most or all switch FETs OFF to achieve a standby mode power level specification - as in the WiFi example above - results in accumulated charge in or near the trap rich layer 404 which takes time to dissipate when the FETs are turned back ON to the point of stable gain; that time may exceed a required sleep-to-active transition specification.

[0024] Anything that affects the operating point of a FET (e.g. , VTH, biasing, current IDS, etc.) generally adversely impacts the performance of circuits based on such devices, such as the gain stability of an amplifier circuit. Because of the accumulated charge phenomenon, SOI FETs essentially have "memory": that is, changing the operating point of a FET involves also settling the body of the FET and the substrate to a new potential. Further, neighboring FETs and charged nodes can affect the intrinsic properties of a FET. Under such conditions, turning a FET completely OFF and then getting it to turn ON and be completely stable immediately is quite challenging.

[0025] Even providing substrate contacts (S-contacts) to mitigate accumulated charges of various types, as taught in U.S. Patent Application No. 14/964,412 referenced above, does not fully mitigate the problem during operation (even if the contacts are ohmic), due to the high resistivity of the substrate 402 restricting the flow of charge. Moreover, anything coupling to the substrate 402 may change its potential, thereby affecting nearby FET devices.

[0026] Prior attempts to mitigate the accumulated charge problem for FETs fabricated on SOI substrates having a trap rich layer include improving the effectiveness of body ties (such as by interspersing ties on the source side of a FET), using shorter width and/or longer length FETs to lessen accumulated charge (noting that longer length FETs generally do not work for power amplifiers), or adopting process techniques to reduce body-tie resistance. Attempts have also been made at the circuit level to compensate for the effects of the accumulated charge problem (rather than actually mitigate the problem), such as by creating replica circuits to try to track out gain variation due to accumulated charge. However, such circuit specific solutions are generally inadequate as a general solution. [0027] Accordingly, there is a need for methods and circuits that eliminate or reduce the formation of accumulated charge in SOI substrates having a trap rich layer, or which mitigate the effects of accumulated charge on the performance of FETs fabricated on SOI substrates having a trap rich layer. The present invention addresses these and other needs.

SUMMARY OF THE INVENTION

[0028] The invention encompasses several types of radio frequency (RF) integrated circuits (ICs) that avoid or mitigate creation of accumulated charge in a silicon-on-insulator (SOI) substrate, particularly an SOI substrate having a trap rich layer, by avoiding a fully OFF state for at least some field effect transistors (FETs) within the circuit. By keeping the standby operating conditions of such critical FETs at the active state or as close to the active state as possible, accumulated charge is stabilized at a near-constant level. An important insight into the functioning of such FETs was the realization that the less that certain node voltages of FETs within an SOI IC change, the more stable the charge that may accumulate as a result of circuit activity in the active layer of an IC.

[0029] In a first general embodiment, a FET is configured such that, in a standby mode, the FET is turned OFF while maintaining essentially the same VDS as during an active mode, thereby eliminating or reducing changes in accumulated charge.

[0030] In a second general embodiment, a FET is configured such that, in a standby mode, current flow through the FET is interrupted while maintaining essentially the same V G S as during the active mode, thereby eliminating or reducing changes in accumulated charge.

[0031 ] In a third general embodiment, a FET is configured such that, in a standby mode, the FET is switched into a very low current state (a "trickle current" state) that keeps both V G S and VDS close to their respective active mode operational voltages, thereby reducing changes in accumulated charge.

[0032] In some embodiments, S-contacts are formed in an IC substrate to create protected areas that encompass FETs that are sensitive to accumulated charge effects. More specifically, S- contacts substantially surround each circuit to be protected, essentially creating corresponding "wells" surrounded at least in part by S-contact "rings". The rings of S-contacts reduce substrate impedance and thus settling time of the substrate voltage under the circuits, help in shielding the circuits from electrical interference, help in draining accumulated charge from certain layers of the IC, and help to improve impedance matching for the circuits within the wells by preventing uneven substrate potential between circuits. [0033] Notably, embodiments of the invention, particularly the "trickle current" approach described below, help resolve a number of accumulated charge effects of SOI substrates (particularly SOI substrates having a trap rich layer), including floating body effects.

[0034] The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.

DESCRIPTION OF THE DRAWINGS

[0035] FIG. 1 is a block diagram of a typical prior art transceiver such as the type that might be used in a cellular telephone.

[0036] FIG. 2 is a block diagram of a prior art bias voltage generation circuit.

[0037] FIG. 3 is block diagram showing a typical prior art SOI IC structure for a single FET.

[0038] FIG. 4 is block diagram showing an improved prior art SOI IC structure for a single FET.

[0039] FIG. 5A is a schematic diagram of a first embodiment of the present invention that stabilizes VDS of a FET when in a standby mode.

[0040] FIG. 5B is a schematic of one circuit for generating VL O AD for the circuit of FIG. 5A.

[0041] FIG. 5C is a schematic diagram of the circuit of FIG. 5 A configured in an active (non-standby) mode.

[0042] FIG. 5D is a schematic diagram of the circuit of FIG. 5A configured in a standby mode.

[0043] FIG. 6A is a schematic diagram of a second embodiment of the present invention that stabilizes VQS of a FET when in a standby mode.

[0044] FIG. 6B is a schematic diagram of the circuit of FIG. 6A, but configured in a standby mode.

[0045] FIG. 7A is a schematic diagram of a third embodiment of the present invention that provides a trickle current state for a PFET when in a standby mode.

[0046] FIG. 7B is a schematic diagram of a fourth embodiment of the present invention that provides a trickle current state for an NFET when in a standby mode. [0047] FIG. 7C is a schematic diagram of a fifth embodiment of the present invention that provides a trickle current state for an NFET when in a standby mode, utilizing fixed current sources.

[0048] FIG. 7D is a schematic diagram of a fourth embodiment of the present invention that provides a trickle current state for a PFET when in a standby mode.

[0049] FIG. 8 is a schematic diagram of a current mirror circuit for providing a reference current VBG/R, such as to the bias generator circuit of FIG. 2.

[0050] FIG. 9A is a schematic diagram of a current mirror circuit based on the PFET circuit of FIG. 7A.

[0051 ] FIG. 9B is a schematic diagram of a current mirror circuit based on the NFET circuit of FIG. 7B.

[0052] FIG. 10 is a schematic diagram of a current mirror circuit based on the PFET circuit of FIG. 7A having matching drain characteristics.

[0053] FIG. 1 1 is a schematic diagram of a bias circuit based on the NFET circuit of FIG. 7B used in conjunction with a variable bias source-follower circuit.

[0054] FIG. 12 is a schematic diagram of a bias circuit based on the NFET circuit of FIG. 7C used in conjunction with a current mirror having a variable current source.

[0055] FIG. 13 is block diagram showing an SOI IC structure with a trap rich layer, a BOX insulator layer, and substrate contacts for a single FET.

[0056] FIG. 14 is a top plan view of an area of a stylized IC that includes twelve example circuits (e.g. , current mirrors for bias circuits of a power amplifier).

[0057] FIG. 15 is a top plan view of an area of a stylized IC that includes twelve example circuits surrounded by a plurality of S -contacts. [0058] FIG. 16 is a process flow diagram showing a first method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on a silicon-on-insulator (SOI) substrate.

[0059] FIG. 17 is a process flow diagram showing a second method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on an SOI substrate as part of an integrated circuit susceptible to accumulated charge.

[0060] FIG. 18 is a process flow diagram showing a third method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on an SOI substrate.

[0061] FIG. 19 is a process flow diagram showing a fourth method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on an SOI substrate as part of an integrated circuit susceptible to accumulated charge.

[0062] FIG. 20 is a process flow diagram showing a fifth method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on an SOI substrate.

[0063] FIG. 21 is a process flow diagram showing a sixth method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on an SOI substrate as part of an integrated circuit susceptible to accumulated charge.

[0064] FIG. 22 is a process flow diagram showing a seventh method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on an SOI substrate as part of an integrated circuit susceptible to accumulated charge.

[0065] FIG. 23 is a process flow diagram showing an eighth method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on a silicon-on- insulator (SOI) substrate as part of an integrated circuit susceptible to accumulated charge. [0066] FIG. 24 is a process flow diagram 2400 showing a ninth method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated as part of an integrated circuit on a silicon-on-insulator (SOI) substrate.

[0067] Like reference numbers and designations in the various drawings indicate like elements.

DETAILED DESCRIPTION OF THE INVENTION

[0068] The invention encompasses several types of radio frequency (RF) integrated circuits (ICs) that avoid or mitigate creation of accumulated charge in a silicon-on-insulator (SOI) substrate, particularly an SOI substrate having a trap rich layer, by avoiding a fully OFF state for at least some field effect transistors (FETs) within the circuit. By keeping the standby operating conditions of such critical FETs at the active state or as close to the active state as possible, accumulated charge is stabilized at a near-constant level. An important insight into the functioning of such FETs was the realization that the less that certain node voltages of FETs within an SOI IC change, the more stable the charge that may accumulate as a result of circuit activity in the active layer of an IC.

[0069] In specific applications, use of one or more embodiments of the present invention lowers standby power consumption of FETs while enabling a very quick sleep-to-active transition time (e.g. , < 1 μ8) and achieving a very stable gain very soon after becoming active (e.g. , < 0.05 dB gain stability in < 30 μβ).

[0070] Notably, embodiments of the invention, particularly the "trickle current" approach described below, help resolve a number of accumulated charge effects of SOI substrates (particularly SOI substrates having a trap rich layer), including floating body effects.

[0071 ] Fixed VDS Embodiment

[0072] FIG. 5A is a schematic diagram of a first embodiment of the present invention that stabilizes VDS of a FET when in a standby mode. Shown is a P-type FET Ml coupled between a supply voltage VDD and a controlled through-path switch SI . The gate of Ml is coupled to either a bias voltage VBIAS or to a standby voltage VSB (which may be VDD) through a controlled gate switch S2. As an example, FET Ml may be a component of a power amplifier bias circuit.

[0073] The through-path switch SI may be switched to connect Ml to either a Load or a "pseudo" load VL O AD- VL O AD may be provided by a voltage supply that is approximately equal to the voltage present on the drain of Ml during active mode operation. FIG. 5B is a schematic of one circuit for generating VL O AD for the circuit of FIG. 5A. In this example, VL O AD is generated by a simple resistive divider circuit comprising resistances Rl and R2 series coupled between VDD and circuit ground. The voltage for VLOAD depends on VDD and the ratio of Rl to R2. Other circuitry may be coupled between Ml and VDD and/or between Ml and the Load. An NFET version of the circuit of FIG. 5 A would look similar, but "upside down": SI , Load, and VLOAD would be connected between VDD and Ml , with the source of the NFET version of Ml coupled to circuit ground; the gate of Ml would be switched between VBIAS and circuit ground, and the drain of Ml would be switched between Load and VLOAD by SI .

[0074] As should be apparent to one of ordinary skill in the art, SI and S2 may each be implemented as FETs coupled in conventional fashion to function as a single-pole, double throw (SPDT) switch. The difference between Ml and S1/S2 is that Ml is generally modulated by an applied input signal (not shown) and essentially behaves as a variable resistor, while SI and S2 have two binary states, connecting a common terminal to either a first or a second node (in FIG. 5A, SI and S2 are drawn as being in a neutral position but both are generally binary).

[0075] FIG. 5C is a schematic diagram of the circuit of FIG. 5A configured in an active (non-standby) mode. In this configuration, the state of SI is set to couple Ml to the Load and the state of S2 is set to couple the gate of Ml to VBIAS- Accordingly, all currents to the gate of Ml and through Ml (i.e. , source-drain current IDS) flow normally, and Ml is in an active ON mode.

[0076] FIG. 5D is a schematic diagram of the circuit of FIG. 5A configured in a standby mode. In this configuration, the state of SI is set to couple Ml to VLOAD and the state of S2 is set to couple the gate of Ml to VSB, where VSB is set at a value that will turn all currents OFF. For example, by applying VSB = VDD to the gate of Ml , the gate-to-source voltage VQS of Ml becomes 0 V, turning Ml OFF. Since VLOAD is approximately equal to the voltage present on the drain of Ml during active mode operation, then VDS across Ml is essentially the same as during the active (non-standby) mode, thus reducing operating point changes to VDS- Since VDS of Ml remains essentially unchanged when shifting from the active mode to the standby mode, little or no additional charge can accumulate that would adversely affect the switching characteristics of Ml . The result is that Ml can be rapidly returned to the active mode by toggling SI and S2 back to the configuration shown in FIG. 5C.

[0077] Control signals (not shown) for switches SI and S2 may be provided, for example, from a microprocessor 142, or from dedicated power control circuitry or external circuitry (for example, if Ml is part of a power amplifier IC that does not include all of the circuitry for a complete transceiver).

[0078] While FIGS. 5A, 5C, and 5D depict a P-type FET, the same principle applies to an N- type FET: in a standby mode, turn the FET OFF while maintaining essentially the same VDS as during the active mode, thereby eliminating or reducing changes in accumulated charge.

[0079] In summary, this aspect of the invention encompasses an integrated circuit susceptible to accumulated charge and fabricated on an SOI substrate (particularly an SOI substrate having a trap rich layer), including at least one FET having a VDS characteristic and configured such that, in a standby mode, the FET is turned OFF while maintaining essentially the same VDS characteristic as during an active mode, thereby eliminating or reducing changes in accumulated charge in or near the SOI substrate and/or any trap rich layer and/or elsewhere in the FET (e.g. , at the gate, drain, or source of the FET).

[0080] This aspect of the invention further encompasses a circuit fabricated on an SOI substrate (particularly an SOI substrate having a trap rich layer) as part of an integrated circuit susceptible to accumulated charge, the circuit including: at least one FET having a drain, a source, a gate, a VDS characteristic, and a signal path through the FET between the drain and the source; a first switch coupled to the gate of the FET, configured to switchably couple the gate to one of a bias voltage or a standby voltage source; and a second switch coupled to the signal path of the FET, configured to switchably couple the signal path to one of a load or a pseudo-load voltage source; wherein, in an active mode, the first switch couples the gate of the FET to the bias voltage and the second switch couples the signal path of the FET to the load; and wherein, in a standby mode, the first switch couples the gate of the FET to the standby voltage source and the second switch couples the signal path of the FET to the pseudo-load voltage source, thereby maintaining essentially the same VDS characteristic as during the active mode, and thereby eliminating or reducing changes in accumulated charge in or near the SOI substrate and/or any trap rich layer and/or elsewhere in the FET (e.g. , at the gate, drain, or source of the FET). [0081 ] Fixed V G S Embodiment

[0082] FIG. 6A is a schematic diagram of a second embodiment of the present invention that stabilizes VGS of a FET when in a standby mode. Shown is a P-type FET Ml coupled between a supply voltage VDD and a controlled through-path switch SI , which in turn is coupled to a Load. The gate of Ml is coupled to a bias voltage VBIAS- SI may be implemented as a FET coupled in conventional fashion to function as a single-pole, single throw (SPST) switch. Accordingly, SI has two binary states, a CLOSED state (as shown in FIG. 6A) and an OPEN state (as shown in FIG. 6B). As an example, FET Ml may be a component of a bias circuit for a power amplifier. Further, other circuitry may be coupled between Ml and VDD and/or between Ml and the Load. An NFET version of the circuit of FIG. 6A would look similar, but "upside down": SI and Load would be between VDD and the drain of Ml , with the source of the NFET version of Ml coupled to circuit ground, the gate of Ml connected to VBIAS, and the drain of Ml coupled through SI to Load.

[0083] In FIG. 6A, Ml is configured in an active (non-standby) mode. In this configuration, the state of SI is set to CLOSED, thus coupling Ml to the Load. Accordingly, all currents to the gate of Ml and through Ml {i.e. , source-drain current IDS) flow normally, and Ml is in an active ON mode.

[0084] FIG. 6B is a schematic diagram of the circuit of FIG. 6A, but configured in a standby mode. In this configuration, the state of SI is set to OPEN and hence current cannot flow through Ml to the Load. More particularly, VDS of Ml is 0 V, and VGS of Ml does not change, thus reducing operating point changes to VQS- Since VQS of Ml remains essentially unchanged when shifting from the active mode to the standby mode, little or no additional charge can accumulate that would adversely affect the switching characteristics of Ml . The result is that Ml can be rapidly returned to the active mode by toggling SI back to the configuration shown in FIG. 6A.

[0085] Again, control signals (not shown) for switch SI may be provided, for example, from a microprocessor 142, or from dedicated power control circuitry or external circuitry. While FIGS. 6A and 6B depict a P-type FET, the same principle applies to an N-type FET: in a standby mode, interrupt current flow (IDS) through the FET while maintaining essentially the same VQS as during the active mode, thereby eliminating or reducing changes in accumulated charge. [0086] In summary, this aspect of the invention encompasses an integrated circuit susceptible to accumulated charge and fabricated on an SOI substrate (particularly an SOI substrate having a trap rich layer), including at least one FET having a VQS characteristic and configured such that, in a standby mode, current flow through the FET is interrupted while maintaining essentially the same VQS characteristic as during the active mode, thereby eliminating or reducing changes in accumulated charge in or near the SOI substrate and/or any trap rich layer and/or elsewhere in the FET (e.g. , at the gate, drain, or source of the FET).

[0087] This aspect of the invention further encompasses a circuit fabricated on an SOI substrate (particularly an SOI substrate having a trap rich layer) as part of an integrated circuit susceptible to accumulated charge, the circuit including: at least one FET having a drain, a source, a gate, a VQS characteristic, and a signal path through the FET between the drain and the source; and a switch coupled to the signal path of the FET, configured to switchably couple the signal path to a load or interrupt current flow through the signal path of the FET; wherein, in an active mode, the switch couples the signal path of the FET to the load; and wherein, in a standby mode, the switch interrupts current flow through the signal path of the FET, thereby maintaining essentially the same VQS characteristic as during the active mode, and thereby eliminating or reducing changes in accumulated charge in or near the SOI substrate and/or any trap rich layer and/or elsewhere in the FET (e.g. , at the gate, drain, or source of the FET).

[0088] Trickle Current Embodiment

[0089] In the first and second embodiments described above, either VQS or VDS was set to essentially 0 V to turn current flow through a FET OFF or interrupt such current flow. However, to meet a particular power consumption specification, a FET does not necessarily have to be shut completely OFF or have all current flow interrupted; instead, the FET can be switched into a very low current state (a "trickle current" state) that keeps both VQS and VDS close to their respective active mode operational voltages. Essentially, trickle current embodiments of the invention allow enough current to pass through a FET to maintain an active conduction channel in the FET. In contrast, if a FET is shut OFF completely, the conduction channel is no longer formed and needs to be re-formed when bring the FET out of a standby mode. Effectively, a trickle current mode FET circuit passes just barely enough current to keep the conduction channel active and the FET ready to turn ON when transitioning from the standby mode to the active mode. For example, a trickle current in standby mode of less than 1/1 ,000 of the active mode current will suffice for many applications, while in other applications, a trickle current in standby mode of less than 1/10,000 or even 1/100,000 of the active mode current may be more useful. As another example, in one power amplifier bias circuit, the active mode current consumption was about 10 mA, while the standby mode current consumption was in the range of about 0.1 -1 μΑ.

[0090] FIG. 7A is a schematic diagram of a third embodiment of the present invention that provides a trickle current state for a PFET when in a standby mode. Shown is a P-type FET Ml coupled between a supply voltage VDD and a controlled through-path switch S I . The gate of Ml is coupled to a bias voltage VBIAS- Also coupled to the gate of the Ml PFET is a feedback voltage 702 from the drain of Ml that provides control over the VGS of Ml ; the feedback voltage 702 may be controlled or modified in some applications by optional circuitry 704 (an example of one such control circuit is shown in FIG. 8). Note that the feedback voltage 702 may be VBIAS; that is, no external bias voltage is applied to the gate of Ml , just the feedback voltage 702. As in the above examples, FET Ml may be a component of a power amplifier bias circuit.

[0091 ] The through-path switch S I may be switched to connect Ml to either an active mode resistance RL (which may be provided by a Load) or a standby mode resistance RSB- AS in the embodiments described above, S I may be implemented as FETs coupled in conventional fashion to function as a single-pole, double throw (SPDT)switch such that S I has two binary states, connecting a common terminal to either a first or a second node (S I is drawn as being in a neutral position but is generally binary). Other circuitry may be coupled between Ml and VDD and/or between Ml and RL-

[0092] When the circuit of FIG. 7A is configured in an active (non-standby) mode, the state of S I is set to couple Ml to RL. Accordingly, current through Ml flows normally, and Ml is in an active ON mode. When the circuit of FIG. 7A is configured in a standby mode, the state of S I is set to couple Ml to RSB- Current through Ml is primarily controlled by the gate of Ml . When RL is changed to RSB, IDS must necessarily change, which happens automatically by the concurrent change to the VGS of Ml by means of the feedback voltage 702 (if VBIAS was a fixed value and RL was switched to RSB, VDS would be near 0 V because the normal mode IDS X RSB is much greater than VDD)-

[0093] The resistance of RSB should be substantially greater than the resistance of RL SO as to impede the flow of current through Ml down to a trickle. In general, RSB should have some repeatedly achievable resistance value (taking into account PVT affects) such that the trickle current is above the leakage current of Ml . For example, in some embodiments, the ratio of resistances of RSB to RL ranges from approximately 100: 1 to approximately 1000: 1 . Accordingly, because a small amount of trickle current continues to flow through Ml when RSB is switched into circuit, Ml is still active but at much diminished power and speed levels.

[0094] FIG. 7B is a schematic diagram of a fourth embodiment of the present invention that provides a trickle current state for an NFET when in a standby mode. Shown is an N-type FET Ml coupled between circuit ground and a controlled through-path switch SI (SI is drawn as being in a neutral position but is generally binary). The gate of Ml is coupled to a bias voltage VBIAS- Also coupled to the gate of the Ml NFET is a feedback voltage 702 from the drain of Ml that provides control over the V G S of Ml ; the feedback voltage 702 may be controlled or modified in some applications by optional circuitry 704. Again note that VBIAS may be entirely provided by the feedback voltage 702.

[0095] The through-path switch SI may be switched to connect Ml to VDD through either an active resistance RL (which may be provided by a Load) or a standby mode resistance RSB- When the circuit of FIG. 7B is configured in an active (non-standby) mode, the state of SI is set to couple Ml to RL. Accordingly, current through Ml flows normally, and Ml is in an active ON mode. When the circuit of FIG. 7B is configured in a standby mode, the state of SI is set to couple Ml to VDD through RSB , and the V G S of Ml is changed as a function of the feedback voltage 702, thereby reducing IDS- Again, the resistance of RSB should be substantially greater than the resistance of RL SO as to impede the flow of current through Ml down to a trickle. Accordingly, because a small amount of current continues to flow through Ml when RSB is switched into circuit, Ml is still active but at much diminished power and speed levels.

[0096] FIG. 7C is a schematic diagram of a fifth embodiment of the present invention that provides a trickle current state for an NFET when in a standby mode, utilizing fixed current sources. Shown is an N-type FET Ml coupled between circuit ground and a controlled through- path switch SI (SI is drawn as being in a neutral position but is generally binary). The gate of Ml is coupled to a bias voltage VBIAS- Also coupled to the gate of the Ml NFET is a feedback voltage 702 from the drain of Ml that provides control over the VQS of Ml ; again, the feedback voltage 702 may be controlled or modified in some applications by optional circuitry 704, and VBIAS may be entirely provided by the feedback voltage 702.

[0097] The through-path switch SI may be switched to connect Ml either to an active mode current source IN or to a lower power standby mode current source ISB- When the circuit of FIG. 7C is configured in an active (non-standby) mode, the state of SI is set to couple Ml to IN- Accordingly, current through Ml flows normally, and Ml is in an active ON mode. When the circuit of FIG. 7C is configured in a standby mode, the state of SI is set to couple Ml to ISB, and the VQS of Ml is changed as a function of the feedback voltage 702, thereby reducing IDS- The current provided by ISB should be regulated to allow only a trickle of current to flow through Ml . Thus, because a small amount of current continues to flow through Ml when ISB is switched into circuit, Ml is still active but at much diminished power and speed levels.

[0098] FIG. 7D is a schematic diagram of a fourth embodiment of the present invention that provides a trickle current state for a PFET when in a standby mode. Shown is a P-type FET Ml coupled between a supply voltage VDD and a controlled through-path switch SI . The gate of Ml is selectively coupled by switch S2 to a bias voltage VBIAS or to a trickle voltage VTRI C KLE- AS in the above examples, FET Ml may be a component of a power amplifier bias circuit. The embodiments shown in FIGS. 7B and 7C can be similarly configured with a switch S2 to select a bias voltage VBIAS ° r a trickle voltage VTRI C KLE rather than use a feedback voltage 702.

[0099] The through-path switch SI may be switched to connect Ml to either an active mode resistance RL (which may be provided by a Load) or a standby mode resistance RSB- Switch S2 would be controlled to concurrently select between VBIAS (f° r active mode) or VTRI C KLE (for standby mode). As in the embodiments described above, SI and S2 may each be implemented as FETs coupled in conventional fashion to function as a single-pole, double throw (SPDT) switch such that they have two binary states, connecting a common terminal to either a first or a second node (both S I and S2 are drawn as being in a neutral position but are generally binary). Other circuitry may be coupled between Ml and VDD and/or between Ml and RL-

[01 00] When the circuit of FIG. 7D is configured in an active (non-standby) mode, the state of S I is set to couple Ml to RL and the state of S2 is set to couple VBIAS to the gate of Ml . Accordingly, current through Ml flows normally, and Ml is in an active ON mode. When the circuit of FIG. 7D is configured in a standby mode, the state of S I is set to couple Ml to R$B and the state of S2 is set to couple VTRICKLE to the gate of Ml , thereby reducing IDS-

[01 01 ] A common feature of the embodiments of FIGS. 7A-7D is that the switch S I selectively couples the source-drain signal path to one of an active mode normal current flow path, or to a standby mode trickle current path in which current is either impeded (e.g. , by R$B) or regulated (e.g. , by ISB)-

[01 02] In all of the trickle current embodiments described above, since VDS and VGS of Ml change very little when shifting from the active mode to the standby mode, little additional charge can accumulate in or near the SOI substrate and/or any trap rich layer and/or elsewhere in the FET (e.g. , at the gate, drain, or source of the FET). The result is that Ml can be rapidly returned to the active mode by toggling S I back to the active mode configuration. In all cases, control signals (not shown) for switch S I may be provided, for example, from a microprocessor 142, or from dedicated power control circuitry or external circuitry.

[01 03] Regardless of specific implementation, the same principle applies to all trickle current embodiments: in a standby mode, substantially restrict current flow (IDS) through a FET while keeping both VGS and VDS close to their respective active mode operational voltages, thereby reducing changes in accumulated charge.

[01 04] In summary, this aspect of the invention encompasses an integrated circuit susceptible to accumulated charge and fabricated on an SOI substrate (particularly an SOI substrate having a trap rich layer), including at least one FET having a drain, a source, a gate, a VDS characteristic, a VGS characteristic, and a signal path through the FET between the drain and the source, and configured such that, in a standby mode, the FET is switched into a very low current state with respect to current flow through the signal path that keeps both the VQS characteristic and the VDS characteristic of the FET close to respective active mode operational voltages for the V G S characteristic and the VDS characteristic, thereby reducing changes in accumulated charge in or near SOI substrate and/or any the trap rich layer and/or elsewhere in the FET (e.g. , at the gate, drain, or source of the FET).

[0105] This aspect of the invention further encompasses a circuit fabricated on an SOI substrate (particularly an SOI substrate having a trap rich layer) as part of an integrated circuit susceptible to accumulated charge, the circuit including: at least one FET having a drain, a source, a gate, a VDS characteristic, a V G S characteristic, and a signal path through the FET between the drain and the source; and a switch coupled to the signal path of the FET, configured to switchably couple the signal path to a normal current flow path or to a trickle current path; wherein, in an active mode, the switch couples the signal path of the FET to the normal current flow path; and wherein, in a standby mode, the switch couples the signal path of the FET to the trickle current path such that both the V G S characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the VQS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge in or near SOI substrate and/or any the trap rich layer and/or elsewhere in the FET (e.g. , at the gate, drain, or source of the FET). The trickle current path may have a high resistance relative to the normal current flow path, and/or the trickle current path may include a regulated current source that allows only a trickle of current to flow through the FET relative to the normal current flow path.

[0106] Trickle Current Circuit Applications

[0107] The trickle current circuits shown in FIGS. 7A-7D can be used in a wide variety of applications. For example, FIG. 8 is a schematic diagram of a current mirror circuit 800 for providing a reference current VB G /R, such as to the bias generator circuit 206 of FIG. 2. Based on the PFET circuit of FIG. 7A, a differential amplifier 802 provides an essentially constant bias voltage VBIAS to Ml and to a current mirror PFET, M2, based on comparing a reference voltage VB G to the voltage at the drain of Ml , V SENSE- In both cases, a load (resistance) is switched to a much higher value in the standby mode.

[0108] As in FIG. 2, the reference voltage VB G may be, for example, a band-gap voltage reference. As in the example of FIG. 7A, the current mirror circuit 800 can be switched between an active mode (with Ml coupled to resistance R through switch SI) and a trickle current standby mode (with Ml coupled to higher resistance RSB through switch S I).

[0109] FIG. 9A is a schematic diagram of a current mirror circuit 900 based on the PFET circuit of FIG. 7A. FIG. 9B is a schematic diagram of a current mirror circuit 920 based on the NFET circuit of FIG. 7B. In each configuration, FET Ml is self-biased to function as a diode and provides a bias voltage to current mirror FET M2, which is typically sized to be significantly larger than Ml by a scaling factor m; for example, in some embodiments, m = 100. In each configuration, FET M3 represents an example load, or a replica device for a power amplifier. A usable voltage is provided at a respective node N, which may be coupled to other circuitry downstream (not shown); for example, each configuration may be useful for biasing a power amplifier FET stack or a low-noise amplifier (LNA) from node N.

[0110] In either case, in the low power standby mode {i.e. , RSB is switched into circuit by SI), normal bias current is replaced with a trickle current. The trickle current reduces the voltage change on each FET gate (for example, compared to pulling a gate to ground) when switching between the active mode and the standby mode. The trickle current also reduces voltage changes on FET drains if the normal-mode loads remain connected. Note that many more current mirror stages may be connected to M2, and each will remain in the preferred low power state as well when RSB is switched into circuit by SI .

[0111] FIG. 10 is a schematic diagram of a current mirror circuit having matching drain characteristics. In this example, two SPDT switches, Sla and Sib, are configured to switch in unison (but in opposite directions for the circuit as drawn; both switches are drawn as being in a neutral position but are generally binary). PFET Ml is self-biased to function as a diode. In the normal mode, the drain of PFET Ml is coupled to R0 through Sla, and the drain of current mirror PFET M2 is coupled to a Load through Sib. In the standby mode, the drain of Ml is coupled to RSB through Sla and the drain of M2 is coupled to RSB through Sib, thus shorting the drains together.

[0112] Shorting the drains together in the low power standby mode limits additional SOI substrate coupling effects because the two FETs act as one device, which means their terminal voltages match and current densities are the same (ignoring possible device mismatch). Upon return to the active mode, the FETs will have the same starting conditions for VDS and VGS, which helps active mode current accuracy and settling time (note that the active mode configuration may include a feedback loop, not shown).

[01 1 3] In some configurations, separate resistances (e.g. , RSB^ and RsBb) corresponding to switches S l a and S ib may be used to better match the trickle current requirements of other circuitry (not shown) coupled to Ml and/or M2. As one of ordinary skill in the art would understand, the concepts illustrated in FIG. 10 are extensible to NFETs.

[01 14] FIG. 1 1 is a schematic diagram of a bias circuit based on the NFET circuit of FIG. 7B used in conjunction with a variable bias source-follower circuit. NFET Ml is self-biased to function as a diode and provides a bias voltage to NFET MN1 , which is typically sized to be significantly larger than Ml by a scaling factor m. In this example, the drain of MN1 is controlled by a source-follower NFET MNSF coupled to a Load. Actual implementations may use stacks of two or more source-follower stages instead of one, as suggested by the dotted line connecting the single illustrated NFET MNSF to the Load.

[01 1 5] The function of switch S I is as described above for FIG. 7B, coupling resistance R 0 to Ml in the active mode and coupling much higher resistance RSB to Ml in the standby mode. In the illustrated example, as an added feature, the bias voltage to the gate of MNSF may also be varied by mode by controlling switch S2 in unison with switch S I (both switches are drawn as being in a neutral position but are generally binary). In the active mode, switch S2 would be set to supply a normal bias voltage Vbl to the gate of MNSF- However, in the standby mode, switch S2 would be set to supply a lower bias voltage Vb2 to the gate of MNSF- The lower bias voltage Vb2 would keep the drain voltage of MN1 closer to where it would be in the active mode (e.g. , just above VTH for NFET MN1), thereby reducing changes in accumulated charge.

[01 1 6] FIG. 12 is a schematic diagram of a bias circuit based on the NFET circuit of FIG. 7C used in conjunction with a current mirror having a variable current source. Standby mode current source ISBI and active mode current source INI can be coupled by a switch S l a to a diode D. Similarly, standby mode current source ISB2 and active mode current source IN 2 can be coupled by a switch S ib to the sources of PFETS Ml and M3. Switches S l a, S ib are drawn as being in a neutral position but are generally binary, and both are configured to switch in unison. PFET Ml is biased by the diode D, and is series coupled to NFET M2, which is self-biased to function as a diode and provides a bias voltage to NFET M4. PFET M3 is self-biased to function as a diode and is series coupled to NFET M4. The output AY between M3 and M4 may be used to bias power amplifier circuitry (not shown).

[01 1 7] In the normal mode, switches S l a and S ib couple normal current sources INI and IN 2 in-circuit. In the standby mode, switches S l a and S ib couple low power current sources ¾ΒΙ and ISB2 in-circuit. In the illustrated configuration, to prevent large voltage changes on amplifier input (i. e. , large AV) when switching between the standby and active modes, the input to the diode D remains biased with a trickle current in the low power standby mode, rather than being discharged to ground.

[01 1 8] Summary of Embodiments

[01 1 9] More generally, embodiments of the invention encompass an integrated circuit susceptible to accumulated charge and fabricated on an SOI substrate, including: at least one FET having a VDS characteristic and a VQS characteristic, and at least one switch coupled to the FET and configured to be set to a standby mode in which no more than a trickle current flows through the at least one FET while maintaining essentially the same VDS characteristic and VQS characteristic for the at least one FET as during an active mode, thereby eliminating or reducing changes in accumulated charge. Further, one or more of the embodiments described above may be used in the same IC.

[01 20] Substrate Stabilization

[01 21 ] Additional techniques may optionally be used in conjunction with embodiments of the circuits described above. For example, in some embodiments, it may be useful to create protected areas on an SOI substrate that encompass FETs that are sensitive to accumulated charge effects by surrounding such areas with substrate contacts (S-contacts), such as the type of S-contacts taught in U.S. Patent Application No. 14/964,412 referenced above.

[01 22] An S-contact in the context of an IC structure is a path which provides a resistive conduction path between a contact region at a surface of a layer of the IC structure and a contact region at or near a surface of a high resistivity substrate of the IC structure (high resistivity includes a range of 3,000 to 20,000 or higher ohm-cm; as known to a person skilled in the art, standard SOI process uses substrates with a low resistivity, typically below 1 ,000 ohm-cm).

[0123] For example, FIG. 13 is block diagram showing an SOI IC structure 1300 with a trap rich layer 404, a BOX insulator layer 406, and substrate contacts for a single FET 410. In the illustrated embodiment, which is otherwise similar to FIG. 4, two S-contacts 1302a, 1302b penetrate through corresponding isolation regions 1304 from the active layer 408 to or near the upper surface of the high resistivity substrate 402. The material used for the S-contacts 1302a, 1302b can be any low resistivity conductive material, such as polysilicon and various metals (e.g. , tungsten, copper, etc.). In the case of an SOI device, the isolation regions 1304 can be shallow trench isolation (STI) regions. By virtue of penetrating through the isolation regions 1304 within the active layer 408, the S-contacts 1302a, 1302b remain isolated from direct contact with other active regions of the active layer 408. In common practice, the S-contacts 1302a, 1302b are electrically connected, directly or through other circuit elements, to the source S or the gate G of a FET 410; one possible electrical connection is shown by a dotted line 1306 from the source S of the FET 410 to one S-contact 1302a (other possible contacts are not shown here, but are illustrated in U.S. Patent Application No. 14/964,412 referenced above). However, as discussed below, the S-contacts 1302a, 1302b may electrically connected to circuit ground or to another known potential.

[0124] In the case of SOI substrates having a trap rich layer 404, as shown in FIG. 13, an S- contact 1302a can penetrate through the trap rich layer 404 to make direct contact with the high resistivity substrate 402. Alternatively, since the trap rich layer 404 has some conductivity (and may be as conductive as the high resistivity substrate 402), in some applications an S-contact 1302b can make a resistive contact with the high resistivity substrate 402 by contacting the surface of the trap rich layer 404. In other applications, an S-contact 1302b can penetrate the trap rich layer 404 to a depth sufficient enough to make a resistive contact, through a remaining portion of the thickness of the trap rich layer 404, with the high resistivity substrate 402.

[0125] In addition to the purposes taught in U.S. Patent Application No. 14/964,412 referenced above, S-contacts can be used in conjunction with embodiments of the invention (such as the embodiments described above) to create protected areas on an IC substrate that encompass FETs that are sensitive to accumulated charge effects. For example, FIG. 14 is a top plan view of an area 1400 of a stylized IC that includes twelve example circuits 1402 of a type susceptible to accumulated charge resulting from the interaction of an unmodified trap rich layer 404 and transient changes of state of FETs comprising such circuits (e.g. , current mirrors for bias circuits of a power amplifier). FIG. 15 is a top plan view of an area 1500 of a stylized IC that includes the twelve example circuits 1402 of FIG. 14 surrounded by a plurality of S-contacts 1502. As illustrated, S-contacts 1502 substantially surround each circuit 1402, essentially creating corresponding "wells" surrounded by S-contact "rings" (even though not circular). The rings of S-contacts 1502 around the wells reduce substrate impedance and thus settling time of the substrate voltage under the circuits 1402, help in shielding the circuits 1402 from electrical interference (from each other and from other circuits outside the area 1500), help in draining accumulated charge from the high resistivity substrate 402 and/or trap rich layer 404, and help to improve impedance matching for the circuits 1402 within the wells by preventing uneven substrate potential between circuits. However, even a single S-contact near a circuit 1402 may provide a benefit.

[0126] Each of the S-contacts 1502 may be electrically connected, directly or through other circuit elements, to the source S or the gate G of a FET. However, when used with embodiments of the present invention, it may be quite beneficial to connect the S-contacts 1502 to circuit ground or to another known potential (even the IC supply voltage, VDD), to avoid imposing signals on the S-contacts 1502. Such imposed signals may create accumulated charge in the high resistivity substrate 402 and/or in or near the trap layer 404 and/or elsewhere in the FET (e.g. , at the gate, drain, or source of the FET), and may arise, e.g. , due to varying voltages applied to active layer 408 elements, such as the source S or gate G of a FET. While a static potential may be most beneficial in some applications, in other applications it may be useful to dynamically change the potential applied to the S-contacts 1502, such as by raising or lowering an applied voltage to counteract accumulated charge that arises during some operational phases (e.g. , bursts of signal transmissions in an active mode versus essentially quiescent periods during a standby mode). In some applications, it may be useful to purposefully inject charge into the high resistivity substrate 402 and/or the trap layer 404 by biasing the S-contacts 1502 with a suitable voltage signal. When a potential other than circuit ground is desired, it may be useful use a charge pump or similar means to inject offsetting charge, or apply a negative potential, or apply a positive potential that exceeds the voltage of the IC power supply (e.g. , greater than VDD)-

[0127] The size, number, and spacing of the S-contacts 1502 generally is a matter of design choice. However, to improve transient effects, wells defined by the S-contacts 1502 should be small enough such that there are essentially no gradients under large circuits 1402 that might necessitate additional impedance matching. Accordingly, the size of the S-contact rings should be similar in size to the wells of potential formed by the S-contacts. Note that complete encirclement of each circuit 1402 may not be necessary in all applications, and that a partial ring of S-contacts may suffice. For example, S-contacts may be omitted in some applications for edges of circuits 1402 not shared with other close-by circuits 1402, such as the S-contacts shown within the dotted oval 1506 of FIG. 15. Moreover, while individual "island" type S-contacts 1502 are illustrated in FIG. 15, S-contacts can be formed as trenches, in known fashion.

[0128] If the S-contacts 1502 are biased in some manner, it may be useful to form a guard ring 1508 of S-contacts around the area 1500 to protect other circuitry; S-contact trenches would work particularly well for such a guard ring 1508, which typically would be grounded.

[0129] In addition to the methods taught in U.S. Patent Application No. 14/964,412 referenced above, a person skilled in the art will know of many fabrication methods to provide S- contacts suitable for the purposes described in this disclosure.

[0130] Methods

[0131 ] Another aspect of the invention includes methods for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on a silicon-on-insulator (SOI) substrate (particularly an SOI substrate having a trap rich layer), and for eliminating or reducing changes in accumulated charge in a circuit fabricated on an SOI substrate (particularly an SOI substrate having a trap rich layer) as part of an integrated circuit susceptible to accumulated charge. Following are a number of examples of such methods.

[0132] FIG. 16 is a process flow diagram 1600 showing a first method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on a silicon-on-insulator (SOI) substrate, including: providing at least one field effect transistor (FET) having a VDS characteristic (STEP 1602); and configuring the at least one FET such that, in a standby mode, the FET is turned OFF while maintaining essentially the same VDS characteristic as during an active mode, thereby eliminating or reducing changes in accumulated charge (STEP 1604).

[0133] FIG. 17 is a process flow diagram 1700 showing a second method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on an SOI substrate as part of an integrated circuit susceptible to accumulated charge, including: providing at least one FET having a drain, a source, a gate, a VDS characteristic, and a signal path through the FET between the drain and the source (STEP 1702); coupling a first switch to the gate of the FET and configuring the first switch to selectively couple the gate to one of a bias voltage or a standby voltage source (STEP 1704); coupling a second switch to the signal path of the FET, and configuring the second switch to selectively couple the signal path to one of a load or a pseudo- load voltage source (STEP 1706); in an active mode, setting the first switch to couple the gate of the FET to the bias voltage and setting the second switch to couple the signal path of the FET to the load (STEP 1708); and in a standby mode, setting the first switch to couple the gate of the FET to the standby voltage source and setting the second switch to couple the signal path of the FET to the pseudo-load voltage source, thereby maintaining essentially the same VDS characteristic as during the active mode, and thereby eliminating or reducing changes in accumulated charge (STEP 1710).

[0134] FIG. 18 is a process flow diagram 1800 showing a third method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on an SOI substrate, including: providing at least one FET having a VQS characteristic (STEP 1802); and configuring the at least one FET such that, in a standby mode, current flow through the FET is interrupted while maintaining essentially the same VQS characteristic as during an active mode, thereby eliminating or reducing changes in accumulated charge (STEP 1804).

[0135] FIG. 19 is a process flow diagram 1900 showing a fourth method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on an SOI substrate as part of an integrated circuit susceptible to accumulated charge, including: providing at least one FET having a drain, a source, a gate, a VGS characteristic, and a signal path through the FET between the drain and the source (STEP 1902); coupling a switch to the signal path of the FET and configuring the switch to selectively couple the signal path to a load or interrupt current flow through the signal path of the FET (STEP 1904); in an active mode, setting the switch to couple the signal path of the FET to the load (STEP 1906); and in a standby mode, setting the switch to interrupt current flow through the signal path of the FET, thereby maintaining essentially the same VGS characteristic as during the active mode, and thereby eliminating or reducing changes in accumulated charge (STEP 1908).

[01 36] FIG. 20 is a process flow diagram 2000 showing a fifth method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on an SOI substrate, including: providing at least one FET having a drain, a source, a gate, a VDS characteristic, a VGS characteristic, and a signal path through the FET between the drain and the source (STEP 2002); and configuring the at least one FET such that, in a standby mode, the FET is switched into a very low current state with respect to current flow through the signal path that keeps both the VGS characteristic and the VDS characteristic of the FET close to respective active mode operational voltages for the VGS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge (STEP 2004).

[01 37] FIG. 21 is a process flow diagram 2100 showing a sixth method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on an SOI substrate as part of an integrated circuit susceptible to accumulated charge, including: providing at least one FET having a drain, a source, a gate, a VDS characteristic, a VQS characteristic, and a signal path through the FET between the drain and the source (STEP 2102); coupling a switch to the signal path of the FET and configuring the switch to selectively couple the signal path to a normal current flow path or to a trickle current path (STEP 2104); in an active mode, setting the switch to couple the signal path of the FET to the normal current flow path (STEP 2106); and in a standby mode, setting the switch to couple the signal path of the FET to the trickle current path such that both the VQS characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the VGS characteristic and the VDS characteristic, thereby reducing changes in accumulated charge (STEP 2108). [0138] FIG. 22 is a process flow diagram 2200 showing a seventh method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on an SOI substrate as part of an integrated circuit susceptible to accumulated charge, including: providing at least one FET having a drain, a source, a gate, a VDS characteristic, a V G S characteristic, and a signal path through the FET between the drain and the source (STEP 2202); coupling a switch to the signal path of the FET, configured to selectively couple the signal path to a load resistance or to a high resistance (STEP 2204); in an active mode, setting the switch to couple the signal path of the FET to the load resistance (STEP 2206); and in a standby mode, setting the switch to couple the signal path of the FET to the high resistance, such that both the V G S characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the V G S characteristic and the VDS characteristic, thereby reducing changes in accumulated charge (STEP 2208).

[0139] FIG. 23 is a process flow diagram 2300 showing an eighth method for eliminating or reducing changes in accumulated charge in an integrated circuit fabricated on a silicon-on- insulator (SOI) as part of an integrated circuit substrate susceptible to accumulated charge, including: providing at least one FET having a drain, a source, a gate, a VDS characteristic, a V G S characteristic, and a signal path through the FET between the drain and the source (STEP 2302); coupling a switch coupled to the signal path of the FET, configured to selectively couple the signal path to a first current source or to a second current source, the second current source providing a lower current than the first current source (STEP 2302); in an active mode, setting the switch to couple the signal path of the FET to the first current source (STEP 2302); and in a standby mode, setting the switch to couple the signal path of the FET to the second current source, such that both the V G S characteristic and the VDS characteristic of the FET are close to respective active mode operational voltages for the V G S characteristic and the VDS characteristic, thereby reducing changes in accumulated charge (STEP 2302).

[0140] FIG. 24 is a process flow diagram 2400 showing a ninth method for eliminating or reducing changes in accumulated charge in an integrated circuit susceptible to accumulated charge and fabricated on a silicon-on-insulator (SOI) substrate, including: providing at least one field effect transistor (FET) having a VDS characteristic and a VQS characteristic (STEP 2402); and configuring the at least one FET such that, in a standby mode, no significant current flows through the FET while maintaining essentially the same VDS characteristic and V G S characteristic as during an active mode, thereby eliminating or reducing changes in accumulated charge (STEP 2404).

[0141 ] Other aspects of the above methods may include one or more of the following: the standby mode pseudo-load voltage source outputting a voltage approximately equal to the voltage present on the drain of the FET during active mode operation; the trickle current path having a high resistance relative to the normal current flow path; providing a regulated current source coupled to the trickle current path that allows only a trickle of current to flow through the FET relative to the normal current flow path; the high resistance being at least about 100 times greater than the load resistance; the SOI substrate including a trap rich layer susceptible to accumulated charge in or near such trap rich layer; and/or providing one or more (e.g. , at least a partial ring) substrate contacts (S-contacts) near or around at least one FET.

[0142] Fabrication Technologies and Options

[0143] The term "MOSFET", as used in this disclosure, means any field effect transistor (FET) with an insulated gate and comprising a metal or metal-like, insulator, and semiconductor structure. The terms "metal" or "metal-like" include at least one electrically conductive material (such as aluminum, copper, or other metal, or highly doped polysilicon, graphene, or other electrical conductor), "insulator" includes at least one insulating material (such as silicon oxide or other dielectric material), and "semiconductor" includes at least one semiconductor material.

[0144] As should be readily apparent to one of ordinary skill in the art, various embodiments of the invention can be implemented to meet a wide variety of specifications. Unless otherwise noted above, selection of suitable component values is a matter of design choice. Various embodiments of the invention may be implemented in any suitable IC technology (including but not limited to MOSFET and IGFET structures) that exhibits accumulated charge, including (but not limited to) silicon-on-insulator (SOI) and silicon-on-sapphire (SOS).

[0145] Voltage levels may be adjusted or voltage and/or logic signal polarities reversed depending on a particular specification and/or implementing technology (e.g. , NMOS, PMOS, or CMOS, and enhancement mode or depletion mode transistor devices). Component voltage, current, and power handling capabilities may be adapted as needed, for example, by adjusting device sizes, serially "stacking" components (particularly FETs) to withstand greater voltages, and/or using multiple components in parallel to handle greater currents. Additional circuit components may be added to enhance the capabilities of the disclosed circuits and/or to provide additional functional without significantly altering the functionality of the disclosed circuits.

[0146] The term "circuit ground" includes a reference potential, and is not limited to an earth ground or other "hard" ground.

[0147] A number of embodiments of the invention have been described. It is to be understood that various modifications may be made without departing from the spirit and scope of the invention. For example, some of the steps described above may be order independent, and thus can be performed in an order different from that described. Further, some of the steps described above may be optional. Various activities described with respect to the methods identified above can be executed in repetitive, serial, or parallel fashion.

[0148] It is to be understood that the foregoing description is intended to illustrate and not to limit the scope of the invention, which is defined by the scope of the following claims, and that other embodiments are within the scope of the claims. (Note that the parenthetical labels for claim elements are for ease of referring to such elements, and do not in themselves indicate a particular required ordering or enumeration of elements; further, such labels may be reused in dependent claims as references to additional elements without being regarded as starting a conflicting labeling sequence).