The present invention relates generally to reference voltage circuits used in integrated circuits, and, in particular, to low power reference voltage circuits.
In general, in the descriptions that follow, we will italicize the first occurrence of each special term of art that should be familiar to those skilled in the art of integrated circuits (“ICs”) and systems. In addition, when we first introduce a term that we believe to be new or that we will use in a context that we believe to be new, we will bold the term and provide the definition that we intend to apply to that term. In addition, throughout this description, we will sometimes use the terms assert and negate when referring to the rendering of a signal, signal flag, status bit, or similar apparatus into its logically true or logically false state, respectively, and the term toggle to indicate the logical inversion of a signal from one logical state to the other. Alternatively, we may refer to the mutually exclusive boolean states as logic—0 and logic—1. Of course, as is well known, consistent system operation can be obtained by reversing the logic sense of all such signals, such that signals described herein as logically true become logically false and vice versa. Furthermore, it is of no relevance in such systems which specific voltage levels are selected to represent each of the logic states.
Generally, the high power consumption of current integrated circuit technology has become a critical problem for mobile electronics that must run for days, months, or even years on a single battery charge. As is known, a reference voltage generator is a key element in the overall design of integrated circuits. Such a reference voltage generator is typically used to provide a stable reference voltage to the analog modules of the integrated circuit.
Shown in
Shown in
As is known, semiconductor IC designs are generally sensitive, during normal operation, to variations in process, voltage and temperature (“PVT”). Such variations are expected, though not desired, and are a natural product of semiconductor manufacturing. During an IC design phase, conventional computer-aided design (“CAD”), computer-aided engineering (“CAE”), and computer-aided manufacturing (“CAM”) software programs enable an IC development team to design an IC while also helping to predict circuit variations and circuit sensitivities. However, if the development team and associated software programs fail to predict all sensitivities, the manufacturing process will fabricate a certain number of ICs that do not meet data sheet specifications. Off-chip test equipment will generally determine such sensitivities, but the development team must subsequently spend valuable time to debug and modify the IC. As is known, when the manufacturing process fabricates ICs below a predicted yield the result is lost revenue opportunity.
Due to sub-micron feature sizes, in addition to strict power consumption and operation requirements, reference voltage generator designs may be especially susceptible to PVT. As explained in, for example, the text book of David A. Johns and Ken Martin, entitled “Analog Integrated Circuit Design”, Wiley, pg. 360, 1997, a copy of which is submitted herewith and incorporated herein in its entirety by reference, and as illustrated in
As discussed in our Parent Provisional, we submit what is needed is an improved method and apparatus for a reference voltage generator that provides optimized power consumption, resulting in extended battery life, reduced battery size, and reduced cost. In particular, we submit such a method and apparatus for a reference voltage generator should provide an improved solution for low power supply requirements, while also providing an improved topology for trimming out PVT sensitivities. Also, the characteristics of the reference voltage generator should be controllable and observable in a manufacturing and system test environment.
In accordance with the preferred embodiment of our invention, we provide a trimmable reference voltage generator comprising a reference voltage generator adapted to develop a first reference voltage VOUT; and a variable gain amplifier, coupled to the reference voltage generator, adapted to receive VOUT and to develop a second reference voltage VREF as a predetermined function of VOUT.
We also provide a method we prefer for trimming a reference voltage VREF. Generally, the steps comprise developing a first reference voltage VOUT; and developing a second reference voltage VREF as a predetermined function of VOUT.
Our invention may be more fully understood by a description of certain preferred embodiments in conjunction with the attached drawings in which:
In the drawings, similar elements will be similarly numbered whenever possible. However, this practice is simply for convenience of reference and to avoid unnecessary proliferation of numbers, and is not intended to imply or suggest that our invention requires identity in either function or structure in the several embodiments.
Shown in
|VGS|<|Vth| [Eq. 1]
Also, series-coupled MOSFET transistors, M3 and M4, of a reference voltage generator 12G (“VR 12G”) are each biased in the subthreshold region (step 24) to develop a second output voltage (“V2”) on an electrode common to M3 and M4 (step 26). A difference amplifier 22A develops an output reference voltage, VREF, as a function of the voltage difference between V1 and V2 (step 28). The electrical parameters of VR 12F and VR 12G may be tuned such that VREF is proportional-to-absolute temperature (“PTAT”), or complementary-to-absolute temperature (“CTAT”).
In alternate embodiments, the series-coupled MOSFET transistor circuits of VR 12F and VR 12G may each be designed using various transistor types and gate, drain, source, and bulk electrode connections. In general, it is known that a variety of bulk electrode connections are possible, for example, when a transistor is biased in the subthreshold region. As examples, the bulk electrode may be coupled to the source electrode, the drain electrode, ground, a supply voltage, or a circuit node. Since typical integrated circuit design uses a common bulk electrode for same type transistors, we have chosen to simplify our disclosure by not further discussing the variety of known possible bulk electrode connections.
In general, MOSFETs M1, M2, M3, and M4 may each be individually manufactured as an N-channel MOSFET or a P-channel MOSFET. Those skilled in the art will appreciate that gate electrode (“G”), drain electrode (“D”), and source electrode (“S”) for each transistor of reference voltage generator 12E may each be coupled, for example, to a bias voltage, a positive supply voltage, or ground, providing the respective transistor is biased in the subthreshold region. If desired, different Vth values may be manufactured for M1, M2, M3, and M4 through various approaches, for example, by using a predetermined combination of doping implants, gate width sizes, gate length sizes, and bias voltages of the substrate and wells. Various alternate topologies that apply to VR 12F and VR 12G are discussed, for example, in the Related Application. Also, those skilled in the art of integrated circuits will recognize that difference amplifier 22A may be implemented using various hardware circuits and software algorithms. As discussed in our various embodiments, a difference amplifier is not restricted to any specific circuit structure.
Shown in
By way of example, N-channel transistors M2 and M4 are each biased in the subthreshold region by coupling the gate electrode of each to the respective drain electrode. Also, P-channel transistors M1 and M3 are each biased in the subthreshold region by applying a predetermined VBIAS to the gate electrode. Using a current mirror principle, the current I1 of VR 12I and the current I2 of VR 12J are constrained to be proportional to one another. Note that the values of I1 and I2 are not constrained to equal one another, and can be skewed by a predetermined multiple. For example, I1 may be two times larger than I2. Typically, M2 and M4 will be manufactured with different Vth values. As previously discussed, different Vth values may be achieved for M2 and M4 using various manufacturing and transistor geometry techniques. Note that the use of different doping implants typically results in significant post-manufacturing variation since the doping levels of different implants will vary in an uncorrelated manner. This variation may be minimized by manufacturing the transistors with substantially identical implants. Finally, although V1 and V2 may each be used as independent reference voltages, using VREF as the reference voltage for the integrated system is generally preferred since, as we will explain later, VREF has improved stability when compared to either V1 or V2.
As is well known, the current, I1, through M2 and the current, I2, through M4, when both are biased in the subthreshold region, and when the drain to source voltage (“VDS”) is greater than 3 times the thermal voltage (“νT”), is given by equations 2 and 3, respectively:
where:
If the associated P-channel MOSFETs M1 and M3 of VR 12I and VR 12J, are manufactured substantially similar, I1 and I2 may be set equal, as shown in Equation 4. As is known, I1 and I2 may also be multiples of one another for various sizes of M1 and M3. For example, if the width of M1 is 3 times larger than the width of M3, it follows that, in general, I1 will be 3 times larger than I2.
Referring to equation 6, it can be seen that VREF is a function of
If we further approximate that the subthreshold slope factors, m2 and m4, are equivalent, then the relationship simplifies to:
VREF, which is proportional to V2-V1, may be made substantially temperature-insensitive, PTAT, or CTAT by tuning parameters K2 and K4. Also, the temperature dependency of K2 and K4 are substantially the same and consequently these terms will cancel.
As is known, in alternate embodiments, the series-coupled MOSFET transistor circuit of VR 12I and VR 12J may each be designed using various transistor types and gate, drain, source, and bulk electrode connections, resulting in slightly modified equations.
The difference topology of reference voltage generator 12H is advantageous. For example, M2 and M4 may be designed using transistors with substantially the same doping implants. Such a topology achieves a significantly reduced sensitivity to PVT variations when compared to the circuit topologies 12A, 12B, and 12C of
Shown in
Referring again to reference voltage generator 12K, transistors M2 and M4 are each biased in the subthreshold region by coupling the gate electrode of each to the respective drain electrode. Also, transistors M1 and M3 are each biased in the subthreshold region by coupling the gate electrode of each to the respective source electrode. VR 12L and VR 12M operate such that V1 and V2 achieve a predetermined difference in voltage such that difference amplifier 22C may develop VREF as a function of the voltage difference between V1 and V2. Finally, although V1 and V2 may each be used as independent reference voltages, using VREF as the reference voltage for the integrated system 10 is generally preferred since, as we will explain later, VREF has improved stability when compared to either V1 or V2.
By way of example, for reference voltage generator 12K, M1 and M2 are each biased in the subthreshold region. As is known, the current I1, associated with the series-coupled M1 and M2 is given by equations 8 and 9, respectively:
Current I1, as represented by the right side of each equation, may be set equal to each other:
Once again, by taking the logarithm of both sides of equation 10, equation 13 may be derived as:
Subsequently, for I2, the same approach may be used to derive V2:
It can be seen, that the voltage difference between V1 and V2, which is proportional to VREF, may be temperature-insensitive, PTAT, or CTAT by tuning parameters K1, K2, K3, and K4:
As is known, in alternate embodiments, the series-coupled MOSFET transistor circuit of VR 12L and VR 12M may each be designed using various transistor types and gate, drain, source, and bulk electrode connections, resulting in slightly modified equations.
The difference topology of reference voltage generator 12K is advantageous, for example, where the same associated transistor types are manufactured for VR 12L and VR 12M. Such a topology achieves a significantly reduced sensitivity to PVT variations when compared to the circuit topologies of 12A, 12B, and 12C of
Shown in
During manufacturing and system test, driver 42 may be used to receive the unbuffered VREF, the buffered xVREF, the unbuffered IREF, and the buffered xIREF, and may subsequently provide signals as a function of VREF and IREF for use by off-chip equipment (step 44). Providing VREF to external functions, such as test equipment, provides a way to measure and subsequently determine preferred trim settings to be used by internal trimmer circuitry of reference voltage generator 12 (see, e.g. related co-application).
In one embodiment, buffer 34 may provide a buffered VREF using a unity gain amplifier.
In another embodiment, a common source amplifier and a current mirror may be included with reference voltage generator 12 to output a reference current (“IREF”) for distribution throughout the integrated system.
In yet another embodiment, driver 42 may include a multiplexing function so that a pad used to communicate a digital signal, an analog signal, or a power signal, during normal mode, may be used to measure the driven VREF or the driven IREF during a manufacturing or system test mode. In yet another embodiment, driver 42 may comprise a voltage-to-current converter or a current-to-voltage converter to accommodate test equipment which requires a current input, or a voltage input, respectively. For this example, VREF and IREF may be provided subsequent to a current-to-voltage conversion or a voltage-to-current conversion.
In yet another embodiment, during manufacture or system test, VREF or IREF may be read externally and adjusted iteratively until VREF or IREF have been trimmed to a predetermined value.
In still another embodiment, configuration memory 38 may comprise banks of registers, or alternate storage embodiments, comprising a plurality of configuration bits for storing the desired trim settings for the reference voltage generator 12. Configuration memory 38 may typically be implemented using non-volatile memory, although any known memory type, such as RAM, flash memory, and one-time programmable memory may be used.
Shown in
Also shown is an example topology of a variable gain amplifier 50 adapted selectively to adjust and trim an absolute value of a second reference voltage, VREF. An operational amplifier 52, coupled to receive VREF, develops a voltage difference, VDIFF, to drive the gate electrode of a P-channel transistor M3; a voltage divider, comprising a variable divider 54 and a fixed divider 56 develops VREF as a function of the resulting current through M3.
In alternate embodiments, VREF may be developed, for example, from a selected voltage node of variable voltage divider 54, a selected voltage node of fixed voltage divider 56, or from the drain electrode of transistor M3. Also, the topology of variable voltage divider 54 and fixed voltage divider 56 are shown by way of example. According to the invention, variable voltage divider 54 and fixed voltage divider 56 may be swapped, interleaved, or implemented in any topology that develops VREF as a selected fraction of the voltage developed at the source electrode of transistor M3.
During initialization (and, perhaps, at selected times thereafter), sensitivity trimmer 48 is adjusted to selectively trim VOUT to adjust for a predetermined sensitivity, such as a PVT sensitivity, of VOUT (step 58). Such a trim setting may be stored in configuration memory 38. The voltage regulator then develops a regulated voltage at the source electrode of transistor M3 as a function of a voltage difference between VOUT and VREF. In general, the circuit topology of variable voltage divider 54 and fixed voltage divider 56 provides a variable gain feedback network that develops VREF as a function of the voltage developed at the source of transistor M3, so that VREF is adjusted to a value which is a function of VOUT (step 60). Selective trimming of the adjusted absolute value of VREF may be accomplished by programming, for example, a transistor configuration of the variable divider 54 (step 62). The trim setting for variable divider 54 may be stored in configuration memory 38. Using known techniques, either or both of the trim settings may be made available to resources external to the integrated system.
In alternate embodiments, VOUT may be provided using, for example, any selected one of the various reference voltage generator 12 topologies illustrated in
In another embodiment, the one or more n_M1 transistors may use an alternate, non-binary weighting scheme.
In other embodiments, alternate configurations of trimming topologies may be used to trim VREF, and to compensate for sensitivities.
Shown in
Thus it is apparent that we have provided an improved method and apparatus for our reference voltage generator, and, in particular, we submit that our method and apparatus provides optimized power consumption, resulting in extended battery life, reduced battery size, and reduced cost. In particular, we have provided an improved solution for low power supply requirements, while also providing an improved topology for trimming out PVT sensitivities. Also, we have provided for the characteristics of our reference voltage generator to be controllable and observable in a manufacturing and system test environment. Therefore, we intend that our invention encompass all such variations and modifications as fall within the scope of the appended claims.
This application claims priority to U.S. Provisional Application Ser. No. 61/504,223 filed 3 Jul. 2011 (“Parent Provisional”), and hereby claims benefit of the filing dates thereof pursuant to 37 CFR §1.78(a)(4). The subject matter of the Parent Provisional, in its entirety, is expressly incorporated herein by reference. This application is also related to the pending U.S. application Ser. No. 12/823,160, filed on 25 Jun. 2010 by the Regents of the University of Michigan, and to the pending U.S. application Ser. No. 13/472,870, filed on 16 May 2012 by the Regents of the University of Michigan (collectively, “Related Applications”). The subject matter of the Related Applications, in their entirety, is expressly incorporated herein by reference. This application is related to application Ser. No. ______/______, filed simultaneously herewith (“Related Co-application”) [Attorney Docket No. JAM004]. The subject matter of the Related Co-application, in its entirety, is expressly incorporated herein by reference.
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/US12/45113 | 6/29/2012 | WO | 00 | 2/28/2014 |
Number | Date | Country | |
---|---|---|---|
61504223 | Jul 2011 | US |