Dynamic supply control for line driver

[0002] The present invention relates to digital subscriber line (DSL) technologies. More particularly, the present invention relates to a dynamic supply control for a line driver for use with digital subscriber line (DSL) and asymmetrical digital subscriber line (ADSL) systems.

[0003] FIG. 1 schematically illustrates a conventional line driver system 1 with a single-ended line driver (LD) 3. The line driver 3 is supplied with a fixed supply voltage V_ccand drives a transmission line 5 with a load R_Lthrough a transformer 7 having a turns ratio of 1:n. In the traditional line driver system 1, in order to prevent clipping of an output signal, the fixed supply voltage V_ccis chosen high enough to handle the highest signal peak of an input signal. Discrete Multitone (DMT) signals are commonly used in DSL and ADSL systems, and the associated waveforms exhibit a high Peak-to-Average Ratio (PAR).

[0004] FIG. 2 illustrates a typical waveform of a DMT signal. As shown in FIG. 2, however, the DMT signal amplitude is relatively small for most of the time, and very high peaks occur only occasionally. However, in the conventional line driver system 1, the line driver 3 must be always supplied with a high supply voltage higher than the highest peak expected. Since the average power consumption of the line driver 3 is substantially given by the average of (fixed supply voltage)×(line driver current), a high supply voltage increases the power consumption of the line driver system 1. In other words, a fixed high supply voltage, which is actually needed only during occasional high peaks, results in wasting the valuable system power.

[0005] Using a Class-G amplifier for a line driver is one possible solution, which provides a dynamic supply control scheme so as to reduce the line driver power consumption. A Class-G amplifier basically employs switching between a low supply voltage and a high supply voltage based on the output signal strength. However, the Class-G scheme requires an extra voltage supply, and switching between two supply voltages introduces glitches and creates high frequency harmonics. Such glitches and high frequency harmonics are difficult eliminate, and/or require additional circuitry to reduce the undesirable effects. Therefore, this Class-G approach is less attractive in view of both performance and cost.

[0006] A Class-H high-output voltage level line driver can provide an output voltage swing larger than the supply voltage. FIG. 3 schematically illustrates a supply voltage 9 and an output signal 11 of a conventional Class-H high-output voltage line driver for audio signals (up to 20 kHz). The conventional Class-H line driver typically operates as a non-inverting amplifier with a single-ended output. As shown in FIG. 3, in order to provide a higher output voltage, the supply voltage 9 is lifted as soon as the line driver output voltage 11 increases above-the DC output level (common mode voltage level CM). However, although the conventional high-output voltage level line driver allows reducing the gain of the amplifier and may increase the signal-to-noise ratio, lifting the supply voltage whenever the input audio signal crosses the common mode voltage level is not optimum to save the power of the line driver.

[0007] Accordingly, it would be desirable to provide a line driver circuit for DSL/ADSL applications with reduced power consumption without increasing the cost or complexity of the circuitry.

[0008] A dynamic supply control circuit is provided for a line driver. The line driver has an amplification factor G, receives an input signal voltage, and drives a transmission line having a load R_Lvia a transformer having a turns ratio of 1:n. The circuit includes an input node for receiving an input signal voltage, a supply node for supplying a driving voltage to the line driver, a lift diode coupled between a fixed supply voltage and the supply node, a lift capacitor coupled between the supply node and an output node, and a lift amplifier having the amplification factor G coupled between the input node and the output node. The lift amplifier drives the lift capacitor when the input signal voltage is greater than an input threshold voltage, the input threshold voltage having a value greater than a common mode voltage of the input signal voltage.

[0009] The accompanying drawings, which are incorporated into and constitute a part of this specification, illustrate one or more embodiments of the present invention and, together with the detailed description, serve to explain the principles and implementations of the invention.

[0010] In the drawings:

[0011] FIG. 1 is a schematic electrical diagram illustrating a conventional line driver system including a single-ended line driver.

[0012] FIG. 2 is a diagram schematically illustrating a typical waveform of a DMT signal having a high Peak to Average Ratio.

[0013] FIG. 3 is a diagram schematically illustrating waveforms of a supply voltage and an output signal voltage of a line driver in a conventional high-output voltage line driver system.

[0014] FIG. 4 is an electrical diagram schematically illustrating a dynamic supply control circuit for a line driver in accordance with an embodiment of the present invention.

[0015] FIG. 5A is a diagram schematically illustrating the waveforms of a line driver output (OUT), a driving voltage V_cclift, and a lift amplifier output (OL) in accordance with one embodiment of the present invention.

[0016] FIG. 5B is a diagram schematically illustrating a current provided from the lift amplifier and the lift diode corresponding to the driving voltage V_ccliftshown in FIG. 5A.

[0017] FIG. 6 is an electrical diagram schematically illustrating an example of a charger circuit including a clamp circuit and an enable circuit in accordance with one embodiment of the present invention.

[0018] FIG. 7 is an electrical diagram schematically illustrating a charger circuit in accordance with another embodiment of the present invention.

[0019] FIG. 8 is a diagram schematically illustrating an example of the simulation result indicating an optimum lifting point minimizing the system power consumption in accordance with one embodiment of the present invention.

[0020] FIG. 9 is an electrical diagram schematically illustrating a line driver circuit having a dynamic supply control circuit with a dual-supply system in accordance with one embodiment of the present invention.

[0021] FIG. 10 is a diagram schematically illustrating the waveforms of a line driver output (OUT), a positive driving voltage V_ddlift, a first lift amplifier output (OLP), a negative driving voltage V_sslift, and a second lift amplifier output (OLM).

[0022] FIG. 11 is a diagram schematically illustrating a conventional IC chip structure in which an output device (n-type transistor) of the line driver is formed on a substrate.

[0023] FIG. 12A is a diagram schematically illustrating a semiconductor structure having a deep N-well in accordance with one embodiment of the present invention.

[0024] FIG. 12B is an equivalent circuit diagram of the semiconductor structure shown in FIG. 12A.

[0025] FIG. 13 is an electrical diagram schematically illustrating an output stage of the line driver shown in FIG. 9.

[0026] FIG. 14 is a schematic electrical diagram illustrating the negative side of the output stage shown in FIG. 13.

[0027] FIG. 15 is an electrical diagram schematically illustrating one example of a differential implementation of the line driver circuit in accordance with one embodiment of the present invention.

[0028] FIG. 16 is an electrical diagram schematically illustrating another example of a differential implementation of the line driver circuit in accordance with one embodiment of the present invention.

[0029] FIG. 17A is a diagram schematically illustrating the waveforms of the positive and negative driving voltages V_ddliftand V_ssliftwith respect to the first and second line driver outputs (OUTP and OUTM), shown in FIG. 16.

[0030] FIG. 17B is a diagram schematically illustrating the waveforms of the output (OLP) of the first lift amplifier and the output (OLM) of the second lift amplifier with respect to the first and second input signal voltages (INP and INM), shown in FIG. 16.

[0031] FIG. 18A is a schematic electrical diagram illustrating a lift amplifier having an amplification factor G with a dual supply system.

[0032] FIG. 18B is an electrical diagram schematically illustrating an example of a non-inverting type implementation of the lift amplifier shown in FIG. 18A.

[0033] FIG. 18C is an electrical diagram schematically illustrating an example of an inverting type implementation of the lift amplifier shown in FIG. 18A.

[0034] FIG. 19A is a schematic electrical diagram illustrating a lift amplifier and an adder.

[0035] FIG. 19B is an electrical diagram schematically illustrating an example of a non-inverting type implementation of the combination of the lift amplifier and the adder shown in FIG. 19A.

[0036] FIG. 19C is an electrical diagram schematically illustrating an example of an inverting type implementation of the combination of the lift amplifier and the adder shown in FIG. 19A.

[0037] FIG. 20 is a process flow diagram schematically illustrating a method for dynamically controlling a supply voltage for a line driver in accordance with one embodiment of the present invention.

[0038] FIG. 21 is a process flow diagram schematically illustrating a method for dynamically controlling a dual-supply voltage for a line driver in accordance with one embodiment of the present invention.

[0039] FIG. 22 is a process flow diagram schematically illustrating a method for biasing a specific well structure of an output stage of the line driver in accordance with one embodiment of the present invention.

[0040] Embodiments of the present invention are described herein in the context of a dynamic supply control for a line driver. Those of ordinary skilled in the art will realize that the following detailed description of the present invention is illustrative only and is not intended to be in any way limiting. Other embodiments of the present invention will readily suggest themselves to such skilled persons having the benefit of this disclosure. Reference will now be made in detail to implementations of the present invention as illustrated in the accompanying drawings. The same reference indicators will be used throughout the drawings and the following detailed description refer to the same or like parts.

[0041] In the interest of clarity, not all of the routine features of the implementations described herein are shown and described. It will, of course, be appreciated that in the development of any such actual implementation, numerous implementation-specific decisions must be made in order to achieve the developer's specific goals, such as compliance with application- and business-related constraints, and that these specific goals will vary from one implementation to another and from one developer to another. Moreover, it will be appreciated that such a development effort might be complex and time-consuming, but would nevertheless be a routine undertaking of engineering for those of ordinary skilled in the art having the benefit of this disclosure.

[0042] FIG. 4 schematically illustrates a dynamic supply control circuit 20 for a line driver in accordance with an embodiment of the present invention. The line driver (LD) 21 has an amplification factor (gain) G and drives a transmission line 23 having a load R_Lvia a transformer 25 having a turns ratio of 1:n. The dynamic supply control circuit 20 includes an input node 27 for receiving an input signal voltage (IN), a supply node 29 for supplying a driving voltage (i.e., line driver supply voltage) V_ccliftto the line driver 21, a lift diode (DLIFT) 31, a lift capacitor (CLIFT) 33, and a lift amplifier (LA) 35. The line driver 21 is also coupled to the input node 27 and receives the input signal voltage.

[0043] The lift diode 31 is coupled between a fixed supply voltage V_ccand the supply node 29, and the lift capacitor 33 is coupled between the supply-node 29 and an output node 37. The lift amplifier 35 is coupled between the input node 27 and the output node 37 (through an optional limiter circuit 55). The lift amplifier 35 is configured as a gain stage with an amplification factor (gain) G, which is the same as that of the line driver 21. The lift amplifier 35 is also designed to have the same delay as the line driver 21. The lift amplifier 35 is supplied with a lift amplifier supply voltage V_ccawhich is typically equal to the fixed voltage V_cc. However, since the output of the lift amplifier 35 only needs to provide an extra voltage for a portion of the line driver output, namely, (Amax×(1−R)), the lift amplifier supply voltage V_ccamay be chosen to be less than the fixed supply voltage V_cc. This saves some power of the system though an extra power supply is required.

[0044] The lift amplifier 35 drives the lift capacitor 33 when the input signal voltage is greater than an input threshold voltage LFT_in. A comparator or enable circuit (not shown in FIG. 4) may be used to detect the input signal to activate the lift amplifier 35. The input threshold voltage LFT_inis set to have a value greater than a common mode voltage (CM) of the input signal voltage so as to minimize the power consumption of the system. It should be noted that since the gain G of the line driver 21 is a given parameter, an output threshold voltage LFT_outwith respect to the output of the line driver 21 can be expressed in terms of the input threshold voltage LFT_in, i.e, LFT_out=G×LFT_in, and vice versa. Since the line driver supply (driving voltage) V_ccliftmust be greater than the output signal voltage (OUT) of the line driver 21, the operation of the circuit 20 is explained with respect to the output signal voltage (OUT) and the corresponding output threshold voltage (LFT_out).

[0045] FIG. 5A schematically illustrates the waveforms of the output signal voltage (OUT) of the line driver 21 (line 41), the driving voltage (dynamically controlled supply voltage) V_cclift(line 43), and an output voltage (OL) of the lift amplifier 35 at the output node 37 (line 45). When the output signal voltage (OUT) of the line driver 21 is below the output threshold voltage (LFT_out), the driving voltage V_ccliftis supplied from the fixed supply voltage V_ccthrough the lift diode 31, i. e., the driving current is supplied through the lift diode 31 to the line driver 21 (FIG. 4). As shown in FIG. 5A, the driving voltage V_ccliftis a diode drop (V_diode) below the fixed supply voltage V_cc. FIG. 5B schematically illustrates the current provided from the lift amplifier 35 and the lift diode 31 which corresponds to the driving voltage V_cclift.

[0046] The lift amplifier 35 is biased such that, when the output signal voltage (OUT) is at the output threshold voltage LFT_out, the output (OL) of lift amplifier 35 will be at a certain predetermined voltage (LAMIN). This predetermined voltage (LAMIN) is typically chosen to allow enough headroom for the lift amplifier 35 (and for a clamp circuit (CL) 39, as described below).

[0047] When the output signal voltage (OUT) of the line driver 21 goes above the output threshold voltage LFT_out(i.e., the input signal voltage goes above the input threshold voltage LFT_in), the output (OL) of the lift amplifier 35 goes above the predetermined voltage LAMIN. Since the bottom plate of the lift capacitor 33 is connected to the output node 37, the lift capacitor 33 momentarily preserves its voltage and then its top plate which follows the bottom plate is lifted above the fixed voltage V_ccminus the diode drop V_diode, effectively shutting off the lift diode 31. Thus, in this lifting mode, the driving voltage V_ccliftis supplied from the lift amplifier 35 through the lift capacitor 33.

[0048] Since the lift amplifier 35 has the same gain and delay as the line driver 21, during the lifting mode the dynamically controlled supply voltage (driving voltage) V_cclift(line 43) exactly follows the output of the line river 21 (line 41), resulting in a fixed predetermined headroom (V_head) for the line driver 21 throughout the lift operation, as shown in FIG. 5A. This prevents clipping in the line driver output as long as the dynamic line driver supply voltage V_ccliftcan handle the highest signal peaks. Since during lifting the line driver supply voltage V_ccliftgoes above the fixed supply voltage V_cc, the fixed supply voltage V_cccan be chosen to be smaller than a conventional fixed supply voltage (i.e., the maximum peak voltage plus a headroom) without causing clipping.

[0049] It should be noted that in the dynamic supply control system, all the currents are eventually supplied by the fixed supply voltage V_ccand no current is actually provided by the lifted supply voltage V_cclift. During lifting operation where the lift diode 31 is shut off, the lift amplifier 35 provides the current for the line driver 21 through the lift capacitor 33. That is, the lift capacitor 33 discharges some of its electric charge during the lifting operation. Therefore, the capacitance of the lift capacitor 33 is selected to be large enough so that the lost charge does not result in a significant voltage drop in the lift capacitor 33.

[0050] In accordance with one embodiment of the present invention, the dynamic supply control scheme is so designed that lifting occurs only occasionally based on the input signal pattern and statistics. However, since the lift capacitor 33 discharges some of its electric charge whenever it is driven by the lift amplifier 35, the lift capacitor 33 looses its lift capability little by little unless the lost charge is replaced. Thus, a charger circuit may be used to prevent the lift capability from decreasing. The charger circuit is coupled to the output node 37 and charges the lift capacitor 33 so that the output node 37 is maintained at the predetermined voltage (LAMIN) when the input signal voltage is smaller than the input threshold voltage (LFTin).

[0051] The charger circuit may include a clamp circuit (CL) 39 as shown in FIG. 4. FIG. 6 schematically illustrates an example of the charger circuit 40 including a clamp circuit (CL) 39 and an enable circuit 51. The clamp circuit 39 may be as simple as a voltage-controlled current source. The enable circuit 51 enables the clamp circuit 39 whenever lifting does not occur, typically, when the input signal voltage is smaller than the input threshold voltage LFTin.

[0052] The clamp circuit 39 compares the output (OL) of lift amplifier 35 with a reference voltage which is set at the predetermined voltage LAMIN, and conducts the proper amount of current to the lift capacitor 33 so as to charge the lift capacitor 33. With this closed-loop control, the bottom plate of the lift capacitor 33 (i.e., the output node 37) is maintained at the predetermined voltage level (LAMIN) in a continuous manner when the lift amplifier 35 is not lifting. The timing of activating the clamp circuit 39 may be determined by monitoring the input signal voltage and comparing it to a reference voltage using the comparator 51. The reference voltage is typically set to the input threshold voltage LFT_in=LFT_out/G. The comparator 51 may also be used as an enable circuit for the lift amplifier 35, as mentioned above. The reference voltage of the comparator 51 may be set a small amount lower than the input threshold voltage LFT_inconsidering the response time of the circuit, as is well understand by one of ordinary skilled in the art.

[0053] FIG. 7 schematically illustrates a charger circuit 50 in accordance with another embodiment of the present invention. As shown in FIG. 7, the charger circuit 50 includes a current supply 53 and a switching circuit 55 coupling the current supply 53 to the output node 37. The switching circuit 55 electrically connects the current supply 53 to the output node 37 when a voltage of the output node 37 is greater than the predetermined voltage LAMIN and the input signal voltage is smaller than the input threshold voltage LFT_in. It should be noted the circuit structure of the switching circuit 55 is one example and the switching circuit 50 may have a different structure. In addition, hysteresis in the device response may be considered to provide slightly different threshold voltages for the rising edge and falling edge of the input signal voltage (IN).

[0054] In accordance with one embodiment of the present invention, the lift amplifier 35 is a Class-AB amplifier biased as described above. When the output signal voltage (OUT) goes below the output threshold voltage LFT_out, the Class-AB lift amplifier 35 tries to drive the bottom plate of the lift capacitor 33 below the predetermined voltage LAMIN. In order to prevent this from happening, a limiter circuit (LIM) 55 is provided between the input node 27 and the lift amplifier 35 as shown in FIG. 4. The limiter circuit 55 limits the input voltage signal (IN) input to the lift amplifier 35 such that the output (OL) of the lift amplifier 35 is allowed to go freely above the predetermined voltage LAMIN, but is prevented from going below the predetermined voltage LAMIN. For example, the limiter circuit 55 clips the input signal voltage below the input threshold voltage LFT_in. Alternatively, the limiter circuit may be provided at the output of the lift amplifier 35 so as to clip the output (OUT) of the lift amplifier 35 in a similar manner.

[0055] In accordance with another embodiment of the present invention, the lift amplifier 35 is designed to be of a Class-A type amplifier with a limited current sink capability as opposed to Class-AB type having high sink. In this embodiment, the limiter circuit 55 may be omitted, since the lift capacitor 33 will act so as to limit the output of the lift amplifier 35. As the output of the lift amplifier 35 tries to go below the predetermined voltage LAMIN, the lift capacitor 33 resists to this change by preserving the voltage thereon. The lift amplifier 35 sinks as much current as possible, which is limited. The limited current sink of the lift amplifier 35 will help the clamp circuit 39 charging the lift capacitor 33 outside the lifting operation.

[0056] In accordance with one embodiment of the present invention, the lifting point triggering the lift operation (i.e., the input/output threshold voltage) is determined so as to minimize a total power consumption of the dynamic supply control system with respect to parameters given by the system standards and/or specification. In DSL/ADSL systems, the power to be delivered to the transmission line and the PAR of the output signal voltage are typically specified by the standards, and thus a DSL/ADSL line driver must support a certain signal peak. Typically a transformer is used to trade the maximum voltage swing at the line driver output with the maximum current which must be provided to the transmission line.

[0057] Therefore, in a conventional fixed supply system, the available supply voltage and the minimum headroom requirement of the line driver usually dictates the turns ratio of the transformer. However, in a dynamic supply control scheme in accordance with one embodiment of the present invention, the designer of the system has more freedom to chose the maximum achievable amplitude at the line driver output and hence the transformer turns ratio which must be used. For example, referring to the dynamic supply control system shown in FIG. 4, let us consider the following parameters:

[0058] Assuming that the lift point in the output signal (i.e., the output threshold voltage LFT_out) is given by LFT_out=CM+A_max×R, where R is a constant (0<R<1) depending on the input signal pattern, the following relationships are obtained:

[0059] A_max=V_lmax/n  (1)

[0060] I_lmax=n×V_lmax/R_L  (2)

[0061] CM−A_max=V_head  (3)

[0062] CM+A_max×R=V_cc−V_diode−V_head  (4)

[0063] As is seen from FIG. 5A, Equation (3) expresses the headroom requirement for the line driver on the negative side (i.e., below the common mode voltage CM of the output signal voltage), and Equation (4) expresses the headroom requirement. for the line driver on the positive side (i.e., above the common mode voltage CM of the output signal voltage). From Equations (3) and (4), the peak amplitude A. is obtained as:

[0064] A_max=(V_cc−V_diode−2×V_head)/(1+R)  (5)

[0065] From Equation (5), it can be seen that as the value of the constant R decreases the value of the maximum amplitude A_maxincreases. Since the constant R specifies the lift point in terms of the fraction of the maximum peak amplitude at the line driver output, Equation (5) means that the sooner the lift operation begins, the greater will be the maximum achievable voltage swing at the line driver output. From Equation (1), for a given maximum amplitude V_lmaxat the transmission line (as required by the standards), a larger maximum amplitude A_maxof the line driver output will result in a smaller turns ration. Then from Equation (2), a smaller turns ratio n will result in a smaller line driver current and hence a smaller power dissipation. This is because the average power dissipation is the average of (V_cc×I_linedriver) where V_ccdenotes the fixed supply voltage and I_linedriverdenotes the line driver current.

[0066] However, it should be noted that as the lifting point is lowered for a given signal pattern, lifting occurs more often and thus the lift amplifier operates more often. Every time the lift amplifier 35 is activated, the lift capacitor 33 is discharged to supply the current to the line driver 21 (FIG. 4). The electric charge lost from the lift capacitor 33 during lifting is then compensated by the clamp circuit 39 (or charger circuit) outside of lifting in order to maintain the lift capability, as described above. This means that at steady state the current delivered by the lift amplifier 35 during lifting must be equal to the current supplied by the clamp circuit 39, i.e., an equal amount of power used by the lift amplifier 35 during lifting will be consumed by the charging action outside of lifting. In other words, the lifting operation is “extra expensive,” since it consumes the power twice: first by the lift amplifier 35 discharging the lift capacitor 33; and then by the clamp circuit 39 charging the lift capacitor 33.

[0067] Therefore, in order to reduce the power consumption, the amount of time the lift amplifier 35 stays active should be minimized, which would be achieved by increasing the lift point (i.e., the input/output threshold voltages However, this seems to be contrary to the observation that the lift point must be as low as possible to maximize the voltage swing available from the line driver. Considering the two contradicting requirements, the applicants realized that there must be an optimum value for R providing the optimum lifting point for a given input signal voltage pattern.

[0068] The applicants conducted extensive system simulations using actual DMT signal patterns, and found that an optimum R indeed exists to minimize the power dissipation. The optimum value of R is around 0.3 for DMT signals. That is, the lift point should be chosen such that when the output signal voltage (OUT) reaches about 30% of its maximum peak past the common mode (CM), the lifting action should take place. FIG. 8 schematically illustrates an example of the simulation result for a fixed supply voltage: V_cc=10V, power at the transmission line: P_line=100 mW, headroom: V_head=1.0V, diode drop: V_diode=0.8V, and the peak amplitude at the transmission line: V_lmax=36V. As shown in FIG. 8, the power consumption curve shows a shallow minimum at around 30% (i.e., the value of R is around 0.3). It should be noted that the ratio R is signal-dependent, and when the pattern and statistics of the input signal changes, the optimum value of R will also changes to minimize the system power consumption. The statistics of DMT signals used in ADSL system is fairly constant and system level simulations indicate that the optimality of the lift point indicated above is preserved almost all the time.

[0069] The dynamic supply control system according to the above-discussed embodiments is described as having a single fixed supply voltage V_cc. However, the dynamic supply control in accordance with the present invention is also applicable to a bipolar (dual) supply system, i.e., a supply system with one positive supply voltage and one negative supply voltage of equal magnitudes (V_ddand Vss, where V_ss=−V_dd).

[0070] FIG. 9 schematically illustrates a line driver circuit having a dynamic supply control circuit 60 with a dual-supply system in accordance with one embodiment of the present invention. The line driver circuit includes a line driver (LD) 61 and the dynamic supply control circuit 60 for the line driver 61. The line driver 61 has an amplification factor (gain) G and drives a transmission line 63 having a load R_Lvia a transformer 65 having a turns ratio of 1:n. The common-mode voltage (CM) of the line driver output (OUT) is 0V, i.e. the output signal voltage of the line driver 61 swings around the center-voltage of 0V. The line driver 61 is supplied with two dynamically-controlled supply voltages: a positive supply voltage (first driving voltage) V_ddliftand a negative supply voltage (second driving voltage) V_sslift.

[0071] As shown in FIG. 9, the dynamic supply control circuit 60 includes an input node 67 for receiving an input signal voltage (IN), and a first (positive) portion 60aand a second (negative) portion 60bcoupled between the input node 67 and the line driver 61. The line driver 61 is also coupled to the input node 67 and receives the input signal voltage (IN).

[0072] The first portion 60aincludes a first supply node 71 for supplying the first driving voltage V_ddliftto the line driver 61, a first lift diode (DLIFTP) 73, a first lift capacitor (CLIFTP) 75, and a first (positive) lift amplifier (LAP) 77. The first lift diode 73 is coupled between a first fixed supply voltage V_ddand the first supply node 71, and the first lift capacitor 75 is coupled between the first supply node 71 and a first output node 79. The first lift amplifier 77 is coupled between the input node 67 and the output node 79 (through an optional limiter circuit 95), and supplied with a first (positive) lift amplifier supply voltage V_ddaand a second (negative) lift amplifier supply voltage V_ssa.

[0073] Similarly, the second portion 60bincludes a second supply node 81 for supplying the second driving voltage V_ssliftto the line driver 61, a second lift diode (DLIFTM) 83, a second lift capacitor (CLIFTM) 85, and a second lift amplifier (LAN) 87. The second lift diode 83 is coupled between a second fixed supply voltage V_ssand the second supply node 81, and the second lift capacitor 85 is coupled between the second supply node 81 and a second output node 89. The second lift amplifier 87 is coupled between the input node 67 and the second output node 89 (through an optional limiter circuit 97), and supplied with the first (positive) lift amplifier supply voltage V_ddaand the second (negative) lift amplifier supply voltage V_ssa.

[0074] The first lift amplifier supply voltage V_ddaand the second lift amplifier supply voltage V_ssa(=−V_dda) are typically set equal to the first fixed supply voltage V_ddand the second fixed supply voltage V_ss(=−V_dd), respectively. However, since the output of the first and second lift amplifiers only need to provide the driving voltages for a portion of the line driver output, namely, (Amax×(1−R)), the first lift amplifier supply voltage V_ddacan be chosen to be lower than the first fixed supply voltage V_dd, and the second lift amplifier supply voltage V_ssamay be higher than the second fixed supply voltage V_ss. This saves some power of the system though extra power supplies are required.

[0075] The first lift amplifier 77 is shown in FIG. 9 as a non-inverting gain stage with an amplification factor (gain) of G which is the same as that of the line driver 61. The first lift amplifier 77 also has the same delay as the line driver 61. The first lift amplifier 77 drives the first lift capacitor 75 when the input signal voltage is greater than a first input threshold voltage LFTP_in. The first lift amplifier 77 is biased such that when the output signal voltage (OUT) of the line driver 61 is at a certain positive threshold LFTP_out(=G×LFTP_in), the first lift amplifier output (OLP) is at a certain minimum voltage (LAMINP).

[0076] Similarly, the second lift amplifier 87 is shown in FIG. 9 as a non-inverting gain stage with an amplification factor (gain) of G which is the same as that of the line driver 61. The second lift amplifier 87 also has the same delay as the line driver 61. The second lift amplifier 87 drives the second lift capacitor 85 when the input signal voltage is smaller than a second (negative) input threshold voltage LFTM_in. The second lift amplifier 87 is biased such that when the output signal voltage (OUT) of the line driver 61 is at a certain negative threshold LFTM_out(=G×LFTM_in), the second lift amplifier output (OLM) is at a certain maximum voltage (LAMINM).

[0077] FIG. 10 schematically illustrates the waveforms of the output signal voltage (OUT) of the line driver 61 (line 101), the positive driving voltage V_ddlift(line 103), the first lift amplifier output (OLP) at the first output node 79 (line 105), the negative driving voltage V_sslift(line 107), and the second lift amplifier output (OLM) at the second output node 89 (line 109).

[0078] As shown in FIG. 10, when the output signal voltage (OUT) is below the positive output threshold voltage (LFTP_out), the positive driving voltage V_ddliftis supplied from the first fixed supply voltage V_ddthrough the first lift diode 73, i. e., the positive driving current is supplied through the first lift diode. 73 to the line driver 61 (FIG. 9). The positive driving voltage V_ddliftis a diode drop (V_diode) below the first fixed supply voltage V_dd. Similarly, when the output signal voltage (OUT) is above the negative output threshold voltage (LFTM_out), the negative driving voltage V_ssliftis supplied from the second fixed supply voltage V_ssthrough the second lift diode 83, i. e., the negative driving current is supplied through the second lift diode 83 to the line driver 61 (FIG. 9). The negative driving voltage V_ssliftis a diode drop (V_diode) above the second fixed supply voltage V_ss.

[0079] When the output signal voltage (OUT) goes above the positive output threshold voltage LFTP_out(i.e., the input signal voltage goes above the positive input threshold voltage LFTP_in), the positive output (OLP) of the first lift amplifier 77 goes above the predetermined voltage LAMINP. Since the bottom plate of the first lift capacitor 75 is connected to the first output node 79, the first lift capacitor 75 momentarily preserves its voltage and then its top plate which fellows the bottom plate is lifted above the first fixed voltage V_ddminus the diode drop V_diode, effectively shutting off the first lift diode 73. Thus, in this positive lifting mode, the positive driving voltage V_ddliftis supplied from the first lift amplifier 77 though the first lift capacitor 75.

[0080] Since the first lift amplifier 77 has the same gain and delay as the line driver 61, during the positive lifting mode the positive driving voltage V_ddlift(line 103) exactly follows the output of the line river 61 (line 101), providing a fixed predetermined headroom (V_head) for the positive output of the line driver 61 during the positive lifting operation, as shown in FIG. 10.

[0081] When the output signal voltage (OUT) goes below the negative output threshold voltage LFTM_out(i.e., the input signal voltage goes below the negative input threshold voltage LFTM_in), the negative output (OLM) of the second lift amplifier 87 goes below the predetermined voltage LAMINM. Since the top plate of the second lift capacitor 85 is connected to the second output node 89, the second lift capacitor 85 momentarily preserves its voltage and then its bottom plate which follows the top plate is lifted below the second fixed voltage V_ssplus the diode drop V_diode, effectively shutting off the second lift diode 83. Thus, in this negative lifting mode, the negative driving voltage V_ssliftis supplied from the second lift amplifier 87 though the second lift capacitor 85.

[0082] Since the second lift amplifier 87 has the same gain and delay as the line driver 61, during the negative lifting (pulling down) mode the negative driving voltage V_sslift(line 107) exactly follows the output of the line river 61 (line 101), resulting in a fixed predetermined headroom (V_head) for the negative output of the line driver 61.

[0083] For the same reasons as described above with respect to the single supply system, charger circuits may be used to prevent the first and second lift capacitors 75 and 85 from gradually loosing their lift capability, in accordance with one embodiment of the present invention. As shown in FIG. 9, a first charger circuit (for example, a clamp circuit) 91 is coupled to the first output node 79 and charges the first lift capacitor 75 so that the first output node 79 is maintained at the predetermined positive voltage (LAMINP) when the input signal voltage is smaller than the positive input threshold voltage (LFTP_in). Similarly, a second charger circuit (for example, a clamp circuit) 93 is coupled to the second output node 89 and charges the second lift capacitor 85 so that the second output node 89 is maintained at the predetermined negative voltage (LAMINM) when the input signal voltage is larger than the negative input threshold voltage (LFTM_in). An enable circuit (not shown in FIG. 9) may be used for each of the clamp circuits so as to monitor the input signal voltage and activates the corresponding clamp circuit in the same manner as described above in the previous embodiments.

[0084] Instead of using a clamp circuit which charges the lift capacitor in a continuous manner, each of the charger circuits may employ a switched-current supply circuit having a fixed current supply and a switching circuit such as shown in FIG. 7. It should be noted that the switched-current supply circuit for the second portion 60bshould have the opposite polarity to that for the first portion 60aof the dynamic supply control circuit.

[0085] In accordance with one embodiment of the present invention, the first and second lift amplifiers 77 and 87 are of Class-AB type, and thus limiter circuits are provided in a similar manner as the previous embodiments. As shown in FIG. 9, the dynamic supply control circuit 60 may further include a first limiter circuit (LIMP) 95 coupled between the input node 67 and the first lift amplifier 77, and a second limiter circuit (LIMM) 97 coupled between the input node 67 and the second lift amplifier 87. The first limiter circuit 95 clips the input signal voltage lower than the positive input threshold voltage LFTP_in, and the second limiter circuit clips the input signal voltage higher than the negative input threshold voltage LFTM_in. Alternatively, the limiter circuits may be provided at the output of the lift amplifiers 77 and 87.

[0086] In accordance with another embodiment of the present invention, the first and second lift amplifiers are designed to be of Class-A type having limited current sink and limited source capabilities, respectively. In this case, the limiter circuits 95 and 97 can be omitted.

[0087] In the same manner as the previous embodiments, the optimum lifting point, e.g., the positive and negative input threshold voltages LFTP_inand LFTM_incan be determined so as to minimize a total power consumption of the system. Through extensive system level simulations, it can be shown that the same optimum R as the one found for a single supply system, holds for a dual supply system as well.

[0088] Using the parameters defined above, the design constraints for the dual supply system are expressed as:

[0089] A_max×R=V_dd−V_diode−V_head  (6)

[0090] A_max−A_max×R≦2×V_dda−V_head1  (7)

[0091] where V_head1denotes the total headroom requirement for the lift amplifiers (i.e., the sum of the negative side and positive side headrooms). If it is assumed that V_dda=V_dd, Equation (7) is expressed as:

[0092] A_max−A_max×R≦2×V_dd−V_head1  (7)′

[0093] Equation (6) is the headroom requirement for the line driver on the positive and negative sides, and Equation (7) or (7)′ places a limit to the peak amplitude A_maxat the line driver output due to the finite lift amplifier supply voltage V_dda.

[0094] The main advantage of a dual supply system (such as shown in FIG. 9) over a single supply system (such as shown in FIG, 4) is the fact that both of the positive and negative supply voltages for the line driver can be lifted in the dual supply system. While in a single supply system the “negative” supply voltage is by definition fixed to 0V and thus cannot be lifted, in a dual supply system the positive supply voltage can be lifted up above a fixed positive board supply V_dd, while the negative supply voltage can also be lifted down below a fixed negative board supply V_ss. Thus, for the same amount of available fixed supply voltage range, the dynamically controlled supply voltage in a dual supply system can be increased more than that in a single supply system.

[0095] For example, consider a single supply system with the fixed supply voltage V_cc=6.0V and a dual supply system with the fixed supply and lift amplifier supply voltages V_dd=V_dda=3.0 and V_ss=V_ssa=−3.0V. That is, both supply systems have the same available fixed supply voltage range of 6.0V. It is also assumed that the diode drop V_diode=0.8V, headroom V_head=0.5V, and the signal dependent constant R=0.3 for the both systems. For the single supply system, the peak amplitude at the line driver output A_max=3.2V from Equation (5), while for the dual supply system, A_max=5.7V from Equation (6). Since the maximum available amplitude with the dynamic supply control is greater in a dual supply system than in a single supply system, in order to provide the same maximum voltage on the transmission line V_lmax, a smaller transformer turns ratio can be used in the dual supply system, leading to smaller current and consequently smaller power consumption.

[0096] In accordance with a preferred embodiment of the present invention, the lift amplifier(s) and the line driver are formed on the same Integrated Circuit (IC), so as to provide precise gain and delay matching between the line driver and the lift amplifier(s). FIG. 11 schematically illustrates a conventional IC chip 110 where an output device (n-type transistor 111) of the line driver is formed on a substrate 113. In standard IC chips, the substrate on which all the semiconductor devices or transistors are placed must be biased with the most negative voltage in the IC chip in order to prevent the transistors from latching up and from drawing currents which may be destructive to the IC devices. In a conventional fixed supply system, the substrate 113 is biased with the most negative voltage V_ss, as shown in FIG. 11.

[0097] However, when the dynamic supply control is employed, the output signal voltage (OUT) of the line driver can go below the most negative power supply voltage V_ssduring the lift operation. Since a parasitic diode 115 exists between an output node of the transistor 111 and the substrate 113, such a “super-negative” voltage would turn on the parasitic diode 115 to cause an undesirable current in the conventional substrate 113. Such a current may cause an output to the IC clip and also lead to latch-up of the transistor 111 depending on the circuit layout. In order to prevent such undesirable effects, a third negative supply can be used so as to bias the substrate 113 with the third negative supply, which may be chosen to be more negative than the most negative signal voltage expected at the line driver output. However, providing an additional voltage supply (with a greater negative voltage capability) is not desirable from a cost and complexity standpoint.

[0098] Therefore, in accordance with one embodiment of the present invention, a special process providing a deep N-well structure is employed. FIG. 12A schematically illustrates a semiconductor structure 120 in accordance with this process. As shown in FIG. 12A, an n-type well (deep N-well) 121 is provided in the semiconductor substrate 122, and a p-type well 125 is placed within the n-type well 121. The substrate 122 is biased with the fixed supply voltage (negative board supply voltage) V_ss. The n-type well 121 is biased at a certain positive voltage, for example, the first fixed supply voltage V_dd. The p-type well 125 is biased at the negative driving voltage, i.e., the dynamically controlled negative line driver supply voltage V_sslift. A transistor (NMOS-FET) 127 having a gate (G), a drain (D), and a source (S) is formed on the p-type well 125. Since the n-type well is biased at a certain positive voltage (in this case at V_dd), it isolates the p-type well 125 and the NMOS transistor 127 formed therein from the substrate 122. FIG. 12B shows an equivalent circuit diagram of the deep N-well structure 120.

[0099] FIG. 13 schematically illustrates an output stage of the line driver 61 shown in FIG. 9. As is well understood by one of ordinary skilled in the art, a line driver typically includes a preamplifier (not shown in FIG. 13), a pair of error amplifiers 131 and 133, and a pair of output transistors 135 and 137. As shown in FIG. 13, the output (NMOS) transistor 137 has the semiconductor structure 120 as described above. FIG. 14 illustrates the negative side of the output stage in more detail. As shown in FIG. 14, driver transistors 141 and 143, which are the associated with the output transistor 137, have the same deep N-well structure as the output transistor 137 to match their operation.

[0100] The output transistor 137, whose drain actually goes below the fixed negative voltage V_ss, and the transistors 141 and 143 which are supposed to match the output transistor 137, are placed in a p-well which in turn is placed inside the deep N-well, as described above. The p-well which constitutes the body of the respective NMOS transistor is biased with the dynamically controlled supply voltage (negative driving voltage) V_sslift. Since the negative driving voltage V_ssliftis always below the output signal voltage (OUT) of the line driver, the parasitic diode between the output node and the p-well is prevented from turning on. Thus, the resulting undesirable effect such as latch-up is also prevented. Furthermore, by providing the deep N-well structure, the output device can be locally biased at the “super” negative voltage without affecting or changing the substrate's fixed bias voltage.

[0101] The previous embodiments and associated figures are based on the single-ended structure. However, the dynamic supply control of the present invention is equally applicable to differential systems. In accordance with one embodiment of the present invention, the line driver circuit is configured as a differential driver circuit as opposed to a single-ended one. FIG. 15 schematically illustrates one example of a differential implementation of a line driver circuit 150 with dynamic supply control.

[0102] As shown in FIG. 15, the line driver circuit 150 includes a first line driver (positive portion: LDP) 151a, a second line driver (negative portion: LDM) 151b, a first supply controller 153 and a second supply controller 155 for the first line driver 151a, and a third supply controller 157 and a fourth supply controller 159 for the second line driver 151b. The first and second line drivers 151aand 151bform a differential line driver 151 which is coupled between an input port 161 (first and second input nodes 161aand 161b) and an output port 163 (first and second driver output nodes 163aand 163b). A differential input signal voltage (first input signal voltage INP and second input signal voltage INM) is received at the input port 161 and a differential-output signal voltage (first output signal voltage OUTP and second input signal voltage OUTM) from the line driver 151 is supplied to the output port 163. A first fixed supply voltage V_ddis provided by a positive voltage supply, and a second fixed supply voltage V_ss(=−V_dd) is provided by a negative voltage supply.

[0103] The first and second line drivers 151aand 151bhave an amplification factor (gain) G. The first line driver 151ais supplied with a first positive driving voltage V_ddlift1and a first negative driving voltage V_sslift1, and outputs the first output signal voltage (OUTP) to the first driver output node 163a. Similarly, the second line driver 151bis supplied with a second positive driving voltage V_ddlift2and a second negative driving voltage V_sslift2, and outputs the second output signal voltage (OUTM) to the second driver output node 163b.

[0104] The first supply controller 153 is coupled to the first input node 161aand to the first line driver 151a. The first supply controller 153 includes a first supply node 161 for supplying the first positive driving voltage V_ddlift1to the first line driver 151a, a first lift diode 163, a first lift capacitor 165, and a first lift amplifier 167. The first lift diode 163 is coupled between the first fixed supply voltage V_ddand the first supply node 161, and the first lift capacitor 165 is coupled between the first supply node 161 and a first output node 169. The first lift amplifier 167 has the amplification factor G, and is coupled between the first input node 161aand the first output node 169 (through an optional limiter circuit 211). The first lift amplifier 167 drives the first lift capacitor 165 when the first input signal voltage (INP) is greater than a first input threshold voltage LFTP_in.

[0105] The first dynamic supply controller 153 maintains the first positive driving voltage V_ddlift1a first predetermined amount (i.e., the diode drop of the first lift diode 163) below the first fixed supply voltage V_ddwhen the first output signal voltage (OUTP) is below a first output threshold voltage LFTP_out(=G×LFTP_in). When the first output signal voltage (OUTP) rises above the first output threshold voltage LFTP_out, the first dynamic supply controller 153 lifts the first positive driving voltage V_ddlift1a second predetermined amount (headroom) above the first output signal voltage (OUTP). By this lifting operation, the peak of the first positive driving voltage V_ddlift1becomes greater than the first fixed supply voltage V_dd, preventing clipping of the first line driver output. The first supply controller 153 operates substantially in the same manner as the first (positive) portion 60aof the dynamic supply control circuit 60 (FIG. 9) described above.

[0106] Similarly, the second supply controller 155 is coupled to the first input node 161aand to the first line driver 151a. The second supply controller 155 includes a second supply node 171 for supplying the first negative driving voltage V_sslift1to the first line driver 151a, a second lift diode 173, a second lift capacitor 175, and a second lift amplifier 177. The second lift diode 173 is coupled between the second fixed supply voltage V_ssand the second supply node 171, and the second lift capacitor 175 is coupled between the second supply node 171 and a second output node 179. The second lift amplifier 177 has the amplification factor G, and is coupled between the first input node 161aand the second output node 179 (through an optional limiter circuit 213). The second lift amplifier 177 drives the second lift capacitor 175 when the first input signal voltage (INP) is smaller than a second input threshold voltage LFTM_in.

[0107] The second dynamic supply controller 155 maintains the first negative driving voltage V_sslift1the first predetermined amount (i.e., the diode drop of the second lift diode 173) above the second fixed supply voltage V_sswhen the first output signal voltage is above a second output threshold voltage LFTM_out(=G×LFTM_in). When the first output signal voltage goes below the second output threshold voltage LFTM_out, the second supply controller 155 pulls down (i.e., lifts in the negative direction) the first negative driving voltage V_sslift1the second predetermined amount (headroom) below the first output signal voltage V_ss. By this pulling (negative lifting) operation, the peak of the first negative driving voltage V_sslift1becomes lower than the second fixed supply voltage V_ss. The second supply controller 155 operates substantially in the same manner as the second (negative) portion 60bof the dynamic supply control circuit 60 (FIG. 9) described above.

[0108] The third supply controller 157 is coupled to the second input node 161band to the second line driver 151b. The third supply controller 157 includes a third supply node 181 for supplying the second positive driving voltage V_ddlift2to the second line driver 151b, a third lift diode 183, a third lift capacitor 185, and a third lift amplifier 187. The third lift diode 183 is coupled between the first fixed supply voltage V_ddand the third supply node 181, and the third lift capacitor 185 is coupled between the third supply node 181 and a third output node 189. The third lift amplifier 187 has the amplification factor G, and is coupled between the second input node 161band the third output node 191 (through an optional limiter circuit 215). The third lift amplifier 187 drives the third lift capacitor 185 when the second input signal voltage (INM) is greater than the first input threshold voltage LFTP_in.

[0109] The third dynamic supply controller 157 maintains the second positive driving voltage V_ddlift2the first predetermined amount (diode drop of the third lift diode 183) below the first fixed supply voltage V_ddwhen the second output signal voltage (OUTM) is below the first output threshold voltage LFTP_out. When the second output signal voltage rises above the first output threshold voltage LFTP_outthe third dynamic supply controller 157 lifts the second positive driving voltage V_ddlift2the second predetermined amount (headroom) above the second output signal voltage (OUTM). By this lifting operation, the peak of the second positive driving voltage V_ddlift2becomes greater than the first fixed supply voltage V_dd. The third supply controller 157 operates substantially in the same manner as the first (positive) portion 60aof the dynamic supply control circuit 60 (FIG. 9) described above.

[0110] The fourth supply controller 159 is coupled to the second input node 161band to the second line driver 151b. The fourth supply controller 159 includes a fourth supply node 191 for supplying the second negative driving voltage V_sslift2to the second line driver 151b, a fourth lift diode 193, a fourth lift capacitor 195, and a fourth lift amplifier 197. The fourth lift diode 193 is coupled between the second fixed supply voltage V_ssand the fourth supply node 191, and the fourth lift capacitor 195 is coupled between the fourth supply node 191 and a fourth output node 199. The fourth lift amplifier 197 has the amplification factor G, and is coupled between the second input node 161band the fourth output node 199 (through an optional limiter circuit 217). The fourth lift amplifier 187 drives the fourth lift capacitor 195 when the second input signal voltage (INM) is smaller than the second input threshold voltage LFTM_in.

[0111] The fourth supply controller 159 maintains the second negative driving voltage V_sslift2the first predetermined amount (diode drop of the fourth lift diode 193) above the second fixed supply voltage V_sswhen the second output signal voltage is above the second output threshold voltage LFTM_out. When the second output signal voltage (OUTM) goes below the second output threshold voltage LFTM_outthe fourth supply controller 159 pulls down (lifts in the negative direction) the second negative driving voltage V_sslift2the second predetermined amount (headroom) below the second output signal voltage (OUT). By this negative lifting operation, the peak of the second negative driving voltage V_sslift2becomes lower than the second fixed supply voltage V_ss. The fourth supply controller 159 operates substantially in the same manner as the second (negative) portion 60bof the dynamic supply control circuit 60 (FIG. 9) described above.

[0112] For the same reasons described above, each of the first, second, third, and fourth supply controllers 153, 155, 157, and 159 may include a respective charger circuit 201, 203, 205, or 207 in accordance with one embodiment of the present invention. The first and third charger circuit 201 and 205 charge the first and third lift capacitors 165 and 185, respectively, so that the first and third output nodes 169 and 189 are maintained at a first predetermined voltage LAMINP, respectively, when the respective input signal voltage is lower than the first input threshold voltage LFTP_in. Similarly, the second and fourth charger circuits 203 and 207 charge the second and fourth lift capacitors 175 and 195, respectively, so that the second and fourth output nodes 179 and 199 are maintained at a second predetermined voltage LAMIM, respectively, when the respective input signal voltage is higher than the second input threshold voltage LFTM_in. The charger circuits 201, 203, 205, and 207 may be implemented using clamp circuits or switched current supplies as described above.

[0113] In accordance with one embodiment of the present invention, each of the first, second, third, and fourth supply controllers 153, 155, 157, and 159 may include a respective limiter circuit 211, 213, 215, or 217 when the lift amplifiers 167, 177, 187, and 197 are of Class AB type, as shown in FIG. 15. When each of the lift amplifiers 167, 177, 187, and 197 is of Class A type with a limited sink or source capability, these limiter circuits may be omitted as explained in the previous embodiments.

[0114] FIG. 16 schematically illustrates a line driver circuit 220 in accordance with one embodiment of the present invention. The line driver circuit 220 is a differential amplifier with a dual supply system, and its circuit structure is simplified compared with the line driver circuit 150 shown in FIG. 15. With this simplified structure, the number of lift amplifiers as well as the number of off-chip components is reduced.

[0115] As shown in FIG. 16, the line driver circuit 220 includes an input node 221 (first and second input nodes 221aand 221b), an output node 223 (first and second driver output nodes 223aand 223b), a first line driver (positive portion: LDP) 225acoupled between the first input node 221aand the first driver output node 223a, and a second line driver (negative portion: LDM) 225bcoupled between the second input node 221band the second driver output node 223b. The line driver circuit 220 further includes a first supply controller 227 providing a positive driving voltage V_ddliftto the first and second line drivers 225aand 225b, a second supply controller 229 providing a negative driving voltage V_ssliftto the first and second line drivers 225aand 225b, first and second limiter circuits 231 and 233 coupled to the first input node 221a, third and fourth limiter circuits 235 and 237 coupled to the second input node 221b, a first adder 239, and a second adder 241. A first fixed supply voltage V_ddis provided by a positive voltage supply, and a second fixed supply voltage V_ss(=−V_dd) is provided by a negative voltage supply.

[0116] The first and second line drivers 225aand 225bform a differential line driver 225 coupled between the input port 221 and the output port 223. A differential input signal voltage (first input signal voltage INP and second input signal voltage INM) is received at the input port 221. The first line driver 225a, which has an amplification factor G, receives the. first input signal voltage (INP) and outputs a first output signal voltage (OUTP) to the first driver output node 223a. Similarly, the second line driver 225b, which has an amplification factor G, receives the second input signal voltage (INM) and outputs a second output signal voltage (OUTM) to the second driver output node 223b. Thus, a differential output signal voltage (the first output voltage OUTP and second output signal voltage OUTM) of the line driver 225 is supplied through the output port 223.

[0117] The first limiter circuit 231 clips the first input signal voltage below a first input threshold voltage LFTP_in, and the second limiter circuit 233 clips the first input signal voltage above a second input threshold voltage LFTM_in. Similarly, the third limiter circuit 235 clips the second input signal voltage below the first input threshold voltage LFTP_in, and the fourth limiter circuit 237 clips the second input signal voltage above the second input threshold voltage LFTM_in.

[0118] The outputs of the first and third limiter circuits 231 and 235 are input to the first adder 239. The first adder 239 adds these outputs, i.e., combines the positive portions of the first and second input signal voltages above the first input threshold voltage LFTP_inso as to output a positive-sum input signal voltage 243. Similarly, the outputs of the second and fourth limiter circuits 233 and 237 are input to the second adder 241. The second adder 241 adds these outputs, i.e., combines the negative portions of the first and second input signal voltages below the second input threshold voltage LFTM_inso as to output a negative-sum input signal voltage 245.

[0119] The first supply controller 227 is coupled to the first adder 239 and receives the positive-sum input signal voltage 243. The first supply controller 227 includes a first supply node 251 for supplying the positive driving voltage V_ddliftto the first and second line drivers 225aand 225b, a first lift diode 253, a first lift capacitor 255, and a first lift amplifier 257. The first lift diode 253 is coupled between the first fixed supply voltage V_ddand the first supply node 251, and the first lift capacitor 255 is coupled between the first supply node 251 and a first output node 259. The first lift amplifier 257 has the amplification factor G, and is coupled to the first adder 239 and the first output node 259. The first lift amplifier 257 drives the bottom plate of the first lift capacitor 255 above a predetermined voltage level LAMINP, when the positive-sum input signal voltage 243 is greater than two times the first input threshold voltage LFTP_in(i.e., 2×LFTP_in). In other words, the first output node 259 is driven above a predetermined voltage level LAMINP, whenever, either the first input voltage INP, or the second input voltage INM goes above the first input threshold voltage LFTP_in.

[0120] The first supply controller 227 maintains the positive driving voltage V_ddlifta first predetermined amount (diode drop of the first lift diode 253) below the first fixed supply voltage V_ddwhen the positive-sum input signal voltage 243 is substantially equal to two times the first input threshold voltage LFTP_in, (i.e., 2×LFTP_in). When the positive-sum input signal voltage 243 rises above two times the first input threshold voltage LFTP_in, the first supply controller 227 lifts the positive driving voltage V_ddlifta second predetermined amount (headroom) above either one of the first output signal voltage (OUTP) and the second output signal voltage (OUTM). By this lifting operation, the peak of the positive driving voltage V_ddliftbecomes greater than the first fixed supply voltage V_dd.

[0121] The second supply controller 229 is coupled to the second adder 241 and receives the negative-sum input signal voltage 245. The second supply controller 229 includes a second supply node 261 for supplying the negative driving voltage V_ssliftto the first and second line drivers 225aand 225b, a second lift diode 263, a second lift capacitor 265, and a second lift amplifier 267. The second lift diode 263 is coupled between the second fixed supply voltage V_ssand the second supply node 261, and the second lift capacitor 265 is coupled between the second supply node 261 and a second output node 269. The second lift amplifier 267 is coupled between the second adder 241 and the second output node 269. The second lift amplifier 267 drives the top plate of the second lift capacitor 265 below a predetermined voltage level LAMINM, when the negative-sum input signal voltage 245 is smaller than two times the second input threshold voltage LFTM_in. In other words, the second output node 269 is driven below a predetermined voltage level LAMINM, whenever, either the first input voltage INP, or the second input voltage INM goes below the second input voltage LFTM_in.

[0122] The second supply controller 229 maintains the negative driving voltage V_ssliftthe first predetermined amount (diode drop of the second lift diode 261) above the second fixed supply voltage V_sswhen the negative-sum input signal voltage 245 is substantially equal to two times the second input threshold voltage LFTM_in. When the negative-sum input signal voltage 245 goes below two times the second input threshold voltage LFTM_in, the second supply controller pulls down (lift in the negative direction) the negative driving voltage V_ssliftthe second predetermined amount (headroom) below either one of the first output signal voltage (OUTP) and the second output signal voltage (OUTM). By this pulling operation, the peak of the negative driving voltage V_ssliftbecomes lower than the second fixed supply voltage V_ss.

[0123] FIG. 17A. schematically illustrates the waveforms of the positive and negative driving voltages V_ddliftand V_ssliftwith respect to the first and second output signal voltages (OUTP and OUTM), as described above. FIG. 17B schematically illustrates the waveforms of the output (OLP) of the first lift amplifier 257 and the output (OLM) of the second lift amplifier 267 with respect to the first and second input signal voltages (INP and INM), as described above. It should be noted that the output (OLP) of the first lift amplifier 257 follows the positive-sum input signal voltage 243 (with the proper level shifting and scaling), and the output (OLM) of the second lift amplifier 267 follows the negative-sum input signal voltage 245 (with the proper level shifting and scaling).

[0124] Similarly to the previous embodiments, each of the first and second supply controllers 227 and 229 may include a respective charger circuit 271 or 273 in accordance with one embodiment of the present invention. As shown in FIG. 16, the first charger circuit 271 is coupled to the first output node 259, and charges the first lift capacitor 255 so that the first output node 259 is maintained at a first predetermined voltage (LAMINP) when the positive-sum input signal voltage 243 is lower than two times the first input threshold voltage LFTP_in. Similarly, the second charger circuit 273 is coupled to the second output node 269, and charges the second lift capacitor 265 so that the second output node 269 is maintained at a second predetermined voltage (LAMINM) when the negative-sum input signal voltage 245 is higher than two times the second input threshold voltage LFTM_in. The charger circuits 271 and 273 may be implemented using clamp circuits or switched current supplies as described above.

[0125] Since the two output signal voltages (OUTP and OUTM) of the first and second line drivers 225aand 225bdo not go up (or go down) at the same time (as shown in FIG. 17A), the two line drivers 225aand 225bcan share one supply controller for their positive side operation (positive driving voltage V_ddlift) and another for their negative side operation (negative driving voltage V_sslift). In this embodiment, the limiter circuits 231-237 are used in order to capture positive peaks of the both first and second input signal voltages (INP and INM) for the first supply controller 227, and to capture negative peaks of the both first and second input signal voltages (INP and INM) for the second supply controller 229.

[0126] The lift amplifiers used in the embodiments described above can be configured as inverting type or non-inverting type. FIG. 18A schematically illustrates a lift amplifier 280 having an amplification factor (gain) G with a dual supply system. The above-described level-shift operation of the lift amplifier 280 can be implemented either by an extra current (I_lshift) supplied to the inverting node of the lift amplifier, by the lift amplifier input bias voltage (V_lshift), or both. FIGS. 18B and 18C illustrate an example of a non-inverting type implementation 281 and an inverting type implementation 283, respectively.

[0127] FIG. 19A schematically illustrates a lift amplifier 285 and an adder 287, which may be used as the first lift amplifier 257 and the adder 239 or the second lift amplifier 267 and the adder 241 in the line driver circuit 220. The two inputs A and B of the adder 287 may be the outputs from the first and third limiter circuits 231 and 235, for example. A combination of the lift amplifier 285 and the adder 287 can also be implemented in a non-inverting type structure 289 (FIG. 19B) or in an inverting type structure 281 (FIG. 19C).

[0128] In addition, the lift amplifiers may be powered down when they are not lifting. This will further reduce the power consumption of the line driver circuit.

[0129] FIG. 20 schematically illustrates a method for dynamically controlling a supply voltage for a line driver in accordance with one embodiment of the present invention. The line driver has an amplification factor (gain) G, and receives an input signal voltage IN and provides an output signal voltage OUT to a transmission line having a load R_Lvia a transformer having a turns ratio of 1:n. The method may be performed using the dynamic supply control circuit 20 shown in FIG. 4.

[0130] First, the input signal voltage IN is received (301), and the input signal voltage IN is compared with an input threshold voltage LFT_in(303). The input threshold voltage LFT_inis set higher than a common mode voltage (CM) of the input signal voltage IN. When the input signal voltage IN is smaller than the input threshold voltage LFT_in, a driving voltage V_cclifthaving a value a first predetermined amount (V_diode) below a fixed supply voltage V_cc(305) is supplied to the line driver (307). The corresponding driving current is supplied from the fixed supply voltage via a diode.

[0131] When the input signal voltage IN is greater than the input threshold voltage LFT_in) the driving voltage is lifted so that the lifted driving voltage V_cclifthas a value a second predetermined amount (V_head) above the output signal voltage OUT (309). By this lifting operation, the peak of the lifted driving voltage V_ccliftis greater than the fixed supply voltage V_cc. The lift operation may include generating a lift voltage by amplifying the input signal voltage IN by an amplification factor G, and supplying the lift voltage through a lift capacitor to the line driver. The corresponding driving current is thus provided through the electric discharge from the lift capacitor (311).

[0132] As shown in FIG. 20, a portion of the input signal voltage IN below the input threshold voltage LFT_inmay be clipped before or after the amplification in accordance with one embodiment of the present invention. By clipping the input signal voltage IN, the lift. voltage is prevented from going below a predetermined voltage when the lift voltage is generated using a lift amplifier of Class-AB type. The lift amplifier may be supplied with the lift amplifier supply voltage equal to or smaller than the fixed supply voltage.

[0133] In accordance with one embodiment of the present invention, the lift capacitor is charged outside the lifting operation (315) so as to make up the electric charge lost during the lift operation. By this charging operation, a voltage of a capacitor plate driven by the lift voltage is maintained at a predetermined voltage LAMIN when the input signal voltage IN is smaller than the input threshold voltage LFT_in. A clamp circuit or a switched current supply circuit may be used for the charging operation. For example, the charging may include enabling a clamp circuit coupled to the lift capacitor when the input signal voltage IN is smaller than the input threshold voltage LFT_in. The clamp circuit charges the lift capacitor until the capacitor plate has the predetermined voltage LAMIN. Alternatively, the charging may include providing a switched current supply, and electrically connecting the current supply to the lift capacitor when a voltage of the capacitor plate is greater than the predetermined voltage LAMIN and the input signal voltage IN is smaller than the input threshold voltage.

[0134] In accordance with one embodiment of the present invention, the input threshold voltage LFT_inis determined so as to minimize the total power consumption in the line driver circuit. For example, when the input signal voltage IN has Discrete Multitone (DMT) patterns as used in ADSL systems, the input threshold voltage LFT_inis about 30% of the maximum peak of the input signal voltage measured from the common mode voltage of the input signal voltage.

[0135] FIG. 21 schematically illustrates a method for dynamically controlling a dual-supply voltage for a line driver in accordance with one embodiment of the present invention. The line driver has an amplification factor (gain) G, and receives an input signal voltage IN and provides an output signal voltage (OUT) to a transmission line having a load R_Lvia a transformer having a turns ratio of 1:n.

[0136] First, the input signal voltage IN is received (321). The input signal voltage is compared with a first input threshold voltage LFTP_in(323), and when the input signal voltage IN is below the first input threshold voltage LFTP_in, a positive driving voltage V_ddlifthaving a value a first predetermined amount (V_diode) lower than a first fixed supply voltage V_dd(325) is supplied to the line driver (327). The first input threshold voltage LFTP_inis set higher than a common mode voltage (CM) of the input signal voltage IN.

[0137] When the input signal voltage IN is above the first input threshold voltage FLTP_in, the positive driving voltage V_ddliftis lifted so as to have a value a second predetermined amount (V_head) higher than the output signal voltage OUT (329). The peak of the lifted positive driving voltage V_ddliftbecomes greater than the first fixed supply voltage V_ddby this lifting operation. The lift operation may include generating a positive lift voltage by amplifying the input signal voltage IN by an amplification factor G, and supplying the positive lift voltage through a lift capacitor to the line driver. The corresponding driving current is thus provided through the electric discharge from the lift capacitor (331).

[0138] The input signal voltage IN is also compared with a second input threshold voltage LFTM_in(333). The second input threshold voltage LFTM_inis set lower than the common mode voltage of the input signal voltage IN. When the input signal voltage IN is above the second input threshold voltage LFTM_in, the line driver is supplied with a negative driving voltage V_sslifta value the first predetermined amount (V_diode) higher than a second fixed supply V_ss(335).

[0139] When the input signal voltage IN is below the second input threshold voltage LFTM_in, the negative driving voltage V_ssliftis pulled down so as to have a value the second predetermined amount (V_head) lower than the output signal voltage OUT (337). By this pulling-down (negative lifting) operation, the peak of the negative driving voltage V_ssliftbecomes lower than the second fixed supply voltage V_ss. The pulling-down operation may include generating a negative lift voltage by amplifying the input signal voltage IN by an amplification factor G, and supplying the negative lift voltage through a lift capacitor to the line driver. The corresponding negative driving current is provided by discharging the lift capacitor (339).

[0140] As shown in FIG. 21, in accordance with one embodiment of the present invention, a portion of the input signal voltage IN below the first input threshold voltage LFTP_inmay be clipped (341) when a lift amplifier of Class-AB type is used in generating the positive driving voltage V_ddlift. Similarly, a portion of the input signal voltage IN above the second input threshold voltage LFTM_inmay be clipped (343) when a lift amplifier of Class-AB type is used in generating the negative driving voltage V_sslift. By clipping or limiting the input signal voltage IN for generating the positive driving voltage V_ddlift, the output of the lift amplifier is prevented from going below a predetermined voltage during the time outside the lifting operation. Similarly, by clipping or limiting the input signal voltage IN for generating the negative driving voltage V_sslift, the output of the lift amplifier is prevented from going above a predetermined voltage during the time outside the pull-down operation.

[0141] In accordance with one embodiment of the present invention, the lift capacitor for the positive driving voltage V_ddliftis charged outside the lift operation so as to make up the electric charge lost during the lift operation. By this charging operation, a voltage of a capacitor plate driven by the positive lift amplifier is maintained at a predetermined voltage LAMINP when the input signal voltage IN is smaller than the first input threshold voltage LFTP_in(345). Similarly, the lift capacitor for the negative driving voltage V_ssliftis charged outside the pull-down operation so as to replace the electric charge lost during the pull-down operation. By this charging operation, a voltage of a capacitor plate driven by the negative lift amplifier is maintained at a predetermined voltage LAMINM when the input signal voltage IN is greater than the second input threshold voltage LFTM_in(347).

[0142] Clamp circuits or switched current supply circuits may be used for the charging/discharging operation. For example, the charging may include enabling a clamp circuit coupled to the lift capacitor when the system is not conducting the lift/pull-down operation. The clamp circuit charges the lift capacitor until the capacitor plate has the predetermined voltage LAMINP or LAMINM. Alternatively, the charging may include providing a switched current supply, and electrically connecting the current supply to the lift capacitor when the system is not conducting the lift/pull-down operation.

[0143] The output stage of a line driver typically includes an n-type semiconductor device or NMOS-FET. Such an n-type transistor is typically formed on a p-type semiconductor substrate biased at the second (i.e., negative) fixed supply voltage V_ss. In order to prevent a parasitic diode between an output node of the transistor and the substrate from turning on during the lift operation, a special process is employed to form a specific well structure in accordance with one embodiment of the present invention. As shown in FIG. 22, by this process, an n-type well is provided in the semiconductor substrate (351), a p-type well is placed within the n-type well (353), and a transistor is formed on the p-type well (355). With this specific well structure, the substrate is biased at the negative supply voltage V_ss(357), and the n-type well is biased at the first fixed supply voltage V_ss(359) in order to isolate the p-type well from the substrate. The p-type well is biased at the negative driving voltage V_sslift(361), and the output signal voltage of the line driver is output via the transistor (363). Since the negative driving voltage V_ssliftis always below the output signal voltage of the line driver, the parasitic diode between the output node and the p-well is prevented from turning on.

[0144] While embodiments and applications of this invention have been shown and described, it would be apparent to those skilled in the art having the benefit of this disclosure that many more modifications than mentioned above are possible without departing from the inventive concepts herein. The invention, therefore, is not to be restricted except in the spirit of the appended claims.