Switching circuitry providing improved signal performance at high frequencies and method of operation thereof

[0002] The present invention relates to switching circuitry, and in particular, to a transistor switching circuitry providing improved signal performance at high frequencies.

[0003] Analog or digital switches are employed for connecting two points in signal path to one another. Usually, such switches employ a semiconductor device which is connected between the two points. When the semiconductor device is rendered conductive, it completes a circuit between the two points, and when the device is rendered non-conductive, it opens the circuit between the two points. It is common to use transistors as switches, the most widely used examples including bipolar junction transistor (BJT), field effect transistor (FET), junction FET (JFET), metal-oxide semiconductor FET (MOSFET) and other known types of transistors used in electronic industry.

[0004] A typical example of the prior art switch 10 is shown in FIG. 1. It comprises a switching means represented by an NMOS transistor 12, whose source 14 is connected to an input node 18 of an input means 20, the input node being a first circuitry point, and whose drain 16 is connected to an output node 22 of an output means 24, the output node being a second circuitry point. The input means 20 has an output resistance R₂, and the output means 24 has an input resistance R₃, usually R₃being much larger than R₂. The gate 26 of the transistor 12 is connected to the control means 28 which sends a control signal to the gate 26 to open or close the transistor 12. The control means has an internal resistance R₀which is small compared to the output resistance R₂and the input resistance R₃. When the control signal is below a predetermined threshold value, the transistor 12 is cut off and therefore is non-conducting, which means that the input and output nodes 18 and 22 along a signal path are disconnected. Alternatively, when the control signal is above the threshold value, the transistor 12 is conducting, and the two nodes 18 and 22 along the signal path are connected to each other.

[0005] There is a drawback associated with the above circuitry. It is known that semiconductor devices have internal distributed resistance and capacitance inherently coupled into the devices due to their internal structure. For example, MOSFET transistors have dominant capacitive effect due to the gate-to-channel capacitance which can be modeled by a single capacitor between the gate and the conducting channel. The corresponding distributed capacitor C_chassociated with the internal structure of the transistor 12 is designated by numeral 29 in FIG. 1 (the corresponding distributed resistance R_chof the transistor 12 is not shown). The presence of the distributed capacitor C_chinfluences the frequency characteristics of the switching circuitry 10. The capacitor C_chand the resistor R₂form a resistor-capacitance circuit R₂C_chwhich results in the following transfer function of the switching circuitry 10:

[0006] T=V_out/V_in=1/(1+jωC_chR₂)  (1)

[0007] where V_inand V_outare input and output voltages at the input and output nodes 18 and 22 respectively, ω is frequency of the transmitted signal, and j={square root over (−1)}. As follows from equation (1), the RC circuit operates as a low pass filter, and the switching circuitry 10 cuts off high frequency signals starting approximately at ω˜1/(R₂C_ch). As a result, the circuitry 10 exhibits substantial signal degradation at high frequencies which is not acceptable in many practical situations.

[0008] Accordingly, there is a need in electronic industry for designing alternative switching circuitry which would reduce or eliminate the influence of the parasitic capacitance and provide no or substantially reduced signal degradation at high frequencies.

[0009] It is therefore an object of the invention to provide a switching circuitry which would avoid the afore-mentioned problem.

[0010] According to one aspect of the invention there is provided a switching circuitry, comprising:

[0011] a semiconductor device having an input node and an output node, the input node being connected to an input means having an output resistance R₂, and the output node being connected to an output means having an input resistance R₃;

[0012] the semiconductor device having a first state where the first node and the second node are substantially electrically connected, and a second state where the nodes are substantially electrically disconnected;

[0013] the semiconductor device being responsive to a control signal generated by a control means, having an output resistance R₁, to provide switching of the device between the first and the second states;

[0014] the switching circuitry being characterized in that the output resistance of the control means R₁is greater than the output resistance of the input means R₂i.e. R₁>R₂, thereby ensuring that switching of an electrical signal is provided so that signal degradation at high frequencies is substantially reduced or eliminated.

[0015] Preferably, the control means has the output resistance which is much greater than the output resistance of the input means, i.e. R₁>>R₂. It provides voltage gain and phase shift between the output and input voltages approaching unity and zero respectively at high frequencies, thus ensuring no signal degradation. Usually the output resistance of the control means is of the order of magnitude or greater than the input resistance of the output means, i.e. R₁≧R₃. Other arrangements when R₁>>R₃and/or R₃>>R₂are also possible. Conveniently, the resistance R₁can have a variable magnitude, e.g. being formed as a digitally controlled resistive network to provide digital control of resistance of the resistor R₁. It is implied that magnitude of R₁may vary depending on the signal generated by the control means and/or on the state of semiconductor device (conducting or non-conducting). It is beneficial to have resistances R₁and R₂that are matched to provide stability of the circuitry characteristics. It would also be beneficial for the resistance R₃to have a layout matched to the resistances R₁and R₂.

[0016] Advantageously, the semiconductor device of the switching circuitry comprises a transistor which is selected from the group consisting of BJT transistor, FET transistor, JFET transistor, MOSFET transistor, depletion type MOSFET transistor, enhanced type MOSFET transistor and MESFET transistor. Conveniently, the switching circuitry comprises one of the PMOS and NMOS transistors, where the control means are connected to the gate of the transistor, the source of the transistor being the input node and the drain being the output node. It is also possible to interchange source and gain of the transistor, using them as output and input nodes respectively. Alternatively, the circuitry may comprise the semiconductor device including a first transistor and a second transistor, the transistors having a complimentary structure and arranged so that the source of the first transistor is connected to the drain of the second transistor to form one of the input and output nodes, and the source of the second transistor is connected to the drain of the first transistor to form the other node;

[0017] the control means comprising a first output connected to the gate of the first transistor through a first output resistance R,a, and a second output connected to the gate of the second transistor through a second output resistance R_1b, the first and second outputs generating complimentary control signals for switching the semiconductor device between the first and second states, and the circuitry being characterized in that R_1,a, R_1,b>R₂, and R_1,a, R_1,b˜R₃.

[0018] Beneficially, R_1,a,R_1,b>>R₂and R_1,a,R_1,b≧R₃. Alternatively, instead of one control means having two complimentary outputs, the control means may comprise a first control means and a second control means generating complimentary signals.

[0019] Though control means output resistance R₁may be formed by means of a resistor, various modifications are also possible. For example, the control means output resistance may be formed by means of an output resistance of a long channel inverter, by means of a resistor and a long channel inverter coupled in series and/or by means of a known semiconductor device having a resistance.

[0020] The switching circuitry described above has numerous applications, e.g. in a power down circuit, input/output amplifier, input/output pad, telephone switch and video switch, being capable of switching signals whose transmittance frequency is higher than the frequency of switching. Depending on the application being used, the switching circuitry is capable of switching digital or analog signals.

[0021] According to another aspect of the invention there is provided a method of connecting/disconnecting a first and second circuitry points along a signal path, comprising the steps of:

[0022] generating an electrical signal by an input means having an input node and characterized by an output resistance R₂, the input node being the first circuitry point;

[0023] the signal to be received by an output means having the output node and characterized by an input resistance R₃, the output node being the second circuitry point;

[0024] transmitting the signal through a switching means disposed between the first and second circuitry points, the switching means having a first state where the circuitry points are substantially electrically connected, and a second state where the circuitry points are substantially electrically disconnected, the switching means being responsive to a control signal generated by a control means having an output resistance R₁;

[0025] arranging that R₁>R₂, thereby ensuring that connection and disconnection of the first and second circuitry points takes place without substantial signal degradation at high frequencies.

[0026] By arranging R₁>>R₂(and conveniently using R₁≧R₃) it is provided that the voltage gain of the circuitry approaches unity while phase shift between output and input voltages tends to zero, thus ensuring high quality circuitry characteristics. Conveniently, it may be arranged that R₁>>R₃and/or R₃>>R₂. The method suitable for operation with digital signals. It also can be used for operation with analog signals.

[0027] According to yet another aspect of the invention there is provided a method of controlling voltage gain and phase shift for a high frequency electrical signal passing through a switching circuitry, comprising the steps of:

[0028] generating an electrical signal by an input means having the input node and characterized by an output resistance R₂. the signal to be received by an output means having the output node and characterized by an input resistance R₃;

[0029] transmitting the signal through a switching means disposed between the input node and the output node along a signal path, the switching means having a first state where the nodes are substantially electrically connected, and a second state where the nodes are substantially electrically disconnected, the switching means being responsive to a control signal generated by a control means having an output resistance R₁;

[0030] varying the ratio of R₂/R₁to provide predetermined signal characteristics at high frequencies, whereby the lower the ratio, the higher the voltage gain and the smaller phase shift between output and input voltages resulting in the lower deterioration of the signal.

[0031] Conveniently, it is arranged that R₁≧R₃or R₁>>R₃and/or R₃>>R₂. Similar to the above, the method is suitable for operation with digital and analog signals.

[0032] Thus, the switching circuitry of the invention provides connecting/disconnecting of two nodes of the circuitry along the signal path in such a manner that the influence of the parasitic capacitance of the semiconductor device is reduced, which results in substantially reduced or completely eliminated signal degradation at high frequencies. The operation of the circuitry also provides a method of connecting/disconnecting two circuitry points along a signal path ensuring reduced or eliminated signal degradation at high frequencies. Additionally, it provides a convenient method of controlling high frequency voltage gain and phase shift of an electrical signal passing through a switching circuitry by varying the ratio of the output resistance of the input means over the output resistance of the control means, i.e. by varying R₂/R₁.

[0033] Embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings in which:

[0034] FIG. 1 illustrates a switching circuitry according to the prior art;

[0035] FIG. 2 illustrates a switching circuitry according to a first embodiment of the invention;

[0036] FIGS. 3 to 12 illustrate switching circuitry according to second to eleventh embodiments of the invention respectively; and

[0037] FIGS. 13aand 13billustrate frequency dependence of voltage gain and phase shift respectively for the switching circuitry of FIGS. 1 and 2.

[0038] A switching circuitry 100 according to the first embodiment of the invention is shown in FIG. 2. It is similar to that shown in FIG. 1, and like elements are referred to by same reference numerals incremented by 100, for example, transistor 112 whose source and drain are connected between input and output means 120 and 124 respectively, and whose gate is connected to the control means 128. The circuitry 100 of the first embodiment differs from the circuitry 10 of FIG. 1 in that an additional control means output resistor 130, having resistance R₁, is introduced between the control means 128 and the gate 126. It is arranged that the magnitude of R₁is much greater than R₂and of the order of R₃(or greater), i.e. the output impedance of the control means is now not negligibly small compared to the impedance R₂and R₃of the input and output means respectively as it is arranged in the prior art.

[0039] Introduction of an additional control means output resistor R₁results in a dramatic improvement of the circuitry characteristics. To demonstrate it, below are the calculations of the transfer function of the switching circuitry 100, derived in an assumption that R₃>>R₂, R₁(where the assumption is made for the simplicity of derivations only).

[0040] The transfer function of the circuitry 100 defined as T=V_out/V_inis as follows: $\begin{array}{cc}\begin{array}{c}T=\frac{V_{\mathrm{out}}}{V_{\mathrm{in}}}=\frac{\frac{1}{j\omega C_{\mathrm{ch}}}+R_1}{\frac{1}{jw{C}_{\mathrm{ch}}}+R_1+R_2}=\frac{1+j\omega R_1{C}_{\mathrm{ch}}}{1+j\omega \left(R_1+R_2\right)C_{\mathrm{ch}}}\\ =\frac{\left(1+j\omega R_1{C}_{\mathrm{ch}}\right)\left(1-\mathrm{j\omega }\left(R_1+R_2\right)C_{\mathrm{ch}}\right)}{1+{{\omega }^2\left(R_1+R_2\right)}^2{C}_{\mathrm{ch}}^2}\\ =\frac{1+{\omega }^2{C}_{\mathrm{ch}}^2{R}_1\left(R_1+R_2\right)-\mathrm{j\omega }{C}_{\mathrm{ch}}{R}_2}{1+{{\omega }^2\left(R_1+R_2\right)}^2{C}_{\mathrm{ch}}^2}\end{array}& \left(2\right)\end{array}$

[0041] Accordingly, gain G of the circuitry, defied as an absolute value of the transfer function T, and a phase shift φ between the output and input voltages are expressed as follows: $\begin{array}{cc}G=\left|\frac{V_{\mathrm{out}}}{V_{\mathrm{in}}}\right|=\frac{\sqrt{{\left(1+{\omega }^2{C}_{\mathrm{ch}}^2{R}_1\left(R_1+R_2\right)\right)}^2+{\omega }^2{C}_{\mathrm{ch}}^2{R}_2^2}}{1+{{\omega }^2\left(R_1+R_2\right)}^2{C}_{\mathrm{ch}}^2}& \left(3\right)\\ \mathrm{tg}\varphi =\frac{\omega C_{\mathrm{ch}}{R}_2}{1+{\omega }^2{C}_{\mathrm{ch}}^2{R}_1\left(R_1+R_2\right)}& \left(4\right)\end{array}$

[0042] Accordingly, at high frequencies, when ω→∞, $\begin{array}{cc}G\approx \frac{R_1}{R_2+R_1}=\frac{1}{1+\frac{R_2}{R_1}},& \left(5\right)\end{array}$

[0043] and

[0044] φ→0.  (6)

[0045] From equations (5) and (6) it follows that when R₁>>R₂, then G→1 and φ→0, which corresponds to a perfect transfer function of the circuitry 100. Generally, when R₃is comparable in magnitude to R₁, equations (3) to (6) are valid provided that R₁is substituted with R₁R₃/(R₁+R₃), which corresponds to the resistances R₁and R₃connected in parallel.

[0046] By way of example, the switching circuitry 100 described above has been implemented with the following components: R₂=1 kOhms, R₃=100 kOhms, R₁ranging from 1 Ohms to 10 kOhms, and transistor 112 having a width of w=80 micron and length of 1=0.35 micron. The input and output means may have included an input and output stage amplifiers respectively (not shown in detail in Figures).

[0047] FIGS. 3 to 12 illustrate alternative embodiments of the invention. For convenience, like elements are referred to by same reference numerals incremented by 100 for each consecutive embodiment.

[0048] A switching circuitry 200 of a second embodiment, shown in FIG. 3, is similar to that of the first embodiment of FIG. 2 except for the NMOS transistor 112 being replaced with a PMOS transistor 212.

[0049] Switching circuitry 300 and 400 of third and fourth embodiments, shown in FIGS. 4 and 5 respectively, are similar to that of the first and second embodiments described above except for the control means output resistors 130 and 230 being replaced with long channel inverters 330 and 430 respectively. Accordingly, the output resistances of the inverters are used as the output resistances R₁of the corresponding control means.

[0050] A circuitry 500 of a fifth embodiment, shown in FIG. 6, is similar to that of the first and second embodiments. The difference is how the control means output resistance 530 (R₁) is formed. In this embodiment the resistor 530 is formed by a long channel inverter 531 and a resistor 533 connected in series, i.e. R₁=R_inverter+R_resistor. It is also assumed that R₁>>R₂and R₁≧R₃as discussed above.

[0051] A circuitry 600 of a sixth embodiment repeats the circuitry 500 of the fifth embodiment, being different in that the NMOS transistor 512 is substituted by a PMOS transistor 612, as shown in FIG. 7.

[0052] A circuitry 700 of a seventh embodiment, shown in FIG. 8, is commonly known as a transmission gate. It includes a NMOS transistor 712aand a PMOS transistor 712barranged so that the drain of one transistor is connected to the source of the other, and vice versa. The transistor gates 726aand 726bare controlled by two complimentary signals generated by corresponding first and second control means 728aand 728b, the signals being denoted V_Cand {overscore (V)}_ccomplimentary respectively. When V_cis at the low level, the gate 726aof the NMOS transistor 712awill be at the low level, thus preventing the NMOS transistor 712afrom conducting. Simultaneously, the gate 726bof the PMOS transistor 712bwill be at the high level voltage {overscore (V)}_c, which also prevents that transistor from conducting. Thus, with V_clow, the input and output nodes 718 and 722 of the switching circuitry 700 are disconnected. In order to connect the nodes, we have to raise the control signal V_cto the high level. Correspondingly, the NMOS transistor 712awill have its gate at V_chigh level and will thus be conducting. Simultaneously, the PMOS transistor 712bwill have its gate at {overscore (V)}_clow level and will also be conducting. Thus, with V_chigh, the input and output nodes 718 and 722 of the switching circuitry 700 are connected. Similar to the embodiments described above, the control means 728ais connected to the gate 726aof the transistor 712athrough the additional resistor 730a, as shown in FIG. 8. Accordingly, the control means 728bis connected to the gate 726bof the transistor 712bthrough the additional resistor 730b. The resistors 730aand 730bhave resistances R_1aand R_1brespectively, where R_1a,R_1b>>R₂and R_1a,R_1b≧R₃. Instead of two control means 728aand 728bshown in FIG. 8, it is also possible to have one control means 728 (outlined in FIG. 8 by a dashed line) having two outputs which generate complimentary signals V_cand {overscore (V)}_c.

[0053] A switching circuitry 800 of an eight embodiment repeats the circuitry 700 of the seventh embodiment except for the control means output resistors 730aand 730bbeing substituted with the corresponding long channel inverters 830aand 830bas illustrated in FIG. 9. The inverters 830aand 830bhave output resistances R_1aand R_1brespectively, which satisfy the conditions of R_1a,R_1b>>R₂and R_1a,R_1b≧R₃similar to the above. As discussed with regard to the seventh embodiment, it is also possible replace two control means 828aand 828bby a single control means 828, outlined in FIG. 9 by a dashed line, which has two complimentary outputs.

[0054] A switching circuitry 900 of a ninth embodiment, shown in FIG. 10, differs from the circuitry 700 and 800 in that each of the control means output resistors is now formed of two elements connected in series. For example, the control means output resistor 930a,having resistance R_1a, is formed by an inverter 931aand a resistor 933a, having resistances R^(a)_inverterand R^(a)_resistorrespectively, i.e. R_1a=R^(a)_inverter+R^(a)_resistor. Similarly, the control means output resistor 930b, having resistance R_1b, is formed by an inverter 931band a resistor 933b, having resistances R^(b)_inverterand R^(b)_resistorrespectively, i.e. R_1b=R^(b)_inverter+R^(b)_resistor. It is assumed that R_1a,R_1b>>R₂and R_1a,R_1b≧R₃similar to other embodiments. It is also assumed that either two control means 928aand 928bor a single control means 928 generating two complimentary signals can be used.

[0055] A switching circuitry 1000 of a tenth embodiment of the invention is shown in FIG. 11. It includes a control means output resistance 1030 (R₁) which has a variable magnitude as a function of the control means output signal. By way of example, the resistance R₁shown in FIG. 11 comprises a resistive network having N resistors R₁₁, R₁₂, R₁₃, . . . R_1N. Assuming that the above resistors are controlled by digital signals, generated by the control means 1028, it is possible to vary the magnitude of the composed resistor R₁in accordance with the digital signal by using a predetermined combination of the resistors. For example, depending on the control means signal, it is possible to arrange that only one of N resistors is being used, or alternatively, that any predetermined number of resistors is being used or combination thereof.

[0056] The eleventh embodiment of the invention, having designations similar to other embodiments and shown in FIG. 12, illustrates an implementation of the switching circuitry 1100 having a variable resistor R₁designated by reference numeral 1130. The resistor 1130 is formed by a parallel connection of a resistor 1130ahaving resistance R_1aand a MOS transistor 1130bwhose drain to source path resistance is R_1b. When the MOS transistor 1130bis conducting, the resistance R_1bis low, and the switch 1100 is closed. When the MOS transistor 1130bis not conducting, R_1bis high, and R₁˜R_1a. Thus, the magnitude of R₁is changes depending on the state of the transistor 1130b.

[0057] Alternatively, for the embodiments described above, R₁may comprise a combination of resistors or any other combination of known semiconductor devices having resistive characteristics.

[0058] It is worth mentioning that for all the above embodiments the resistance R₁also changes its magnitude depending on the state of the transistor, i.e. when the transistor is in its conducting state R₁is greater, and when the transistor is non-conducting, R₁is smaller.

[0059] Although the above preferred embodiments have been described for R₁>>R₂, it is also contemplated that alternative embodiments may also be implemented with R₁>R₂.

[0060] While the embodiments of the invention have been implemented by use of PMOS and NMOS transistors, it is also contemplated that other types of transistors known in electronic industry, e.g. BJT, FET, JFET, depletion type MOSFET, MESFET, may also be used in alternative embodiments. Additional modifications include substitution of the transistor with another known semiconductor device capable of operating as a switch. While in the embodiments described above the output resistors of the control means have been implemented by use of resistors and long channel inverters, it is also contemplated that in the alternative embodiments they may be formed by use of other known semiconductor elements, or a combination thereof.

[0061] It is also important to ensure that resistances R₁and R₂(or alternatively R₁, R₂and R₃) are matched, e.g. manufactured by the same technology by using same materials and having same shape, size and orientation, located in close vicinity, to provide similar characteristics.

[0062] It is implied that all the improvements to the circuitry characteristics described above relate to the switching circuitry being in its conducting state, while circuitry characteristics in non-conducting state remain unchanged. The embodiments on the invention allow to diminish parasitic effects not only due to the gate to channel capacitance as mentioned above, but also due to the channel to substrate capacitance which produces parasitic effects in a similar manner.

[0063] Principles of operation of the switching circuitry will be illustrated on the circuitry 100 of the first embodiment. FIGS. 13aand 13billustrate frequency dependence for voltage gain and phase shift of the circuitry 100 calculated in accordance with equations (3) and (4) respectively for different values of R₁. Curves 2000, 2010, 2020 and 2030 for gain response in FIG. 13acorrespond to curves 2040, 2050, 2060 and 2070 for phase shift response in FIG. 13band are respectively calculated for the following values of R₁:R₁=1 Ohm, R₁=100 Ohms, R₁=1 kOhm, R₁=10 kOhms. Values of R₂and R₃are set constant and equal to R₂=1 kOhm and R₃=100 kohms respectively. Virtually, curves 2000 and 2040 represent the prior art results.

[0064] As follows from FIG. 13, the lower the ratio of R₂/R₁, the closer the voltage gain to its maximal value of unity (FIG. 13a, curve 2030) and the smaller the phase shift between the output and input voltages approaching zero (FIG. 13b, curve 2070). It provides a convenient method of obtaining a predetermined circuitry characteristics (voltage gain and phase shift) and avoiding signal degradation at high frequencies. When G→1 and φ→0 the circuitry 100 provides high quality switching without signal distortion.

[0065] Operation of circuitry 200 to 1100 of other embodiments is similar to that of the first embodiment and provides same advantages.

[0066] Thus, the switching circuitry of the invention provides connecting/disconnecting of two nodes of the circuitry along the signal path in such a manner that the influence of the parasitic gate-to-channel capacitance of the semiconductor device is reduced, which results in substantially reduced or completely eliminated signal degradation at high frequencies. The operation of the circuitry also provides a method of connecting/disconnecting two circuitry points along the signal path ensuring high quality of the signal at high frequencies. Additionally, it provides a convenient method of controlling high frequency voltage gain and phase shift of the electrical signal passing through the switching circuitry by varying the ratio of the output resistance of the input means over the output resistance of the control means over, i.e. by varying R₂/R₁.

[0067] The switching circuitry of the invention is suitable for operation with digital and analog signals, which provides numerous applications. For example, the circuitry may be used in power down circuits, video or telephone switches, input/output amplifiers and pads, or any other application where the required frequency of switching is lower than the transmitted frequency.

[0068] The switching circuitry of the invention and methods of operation thereof provide many advantages. It ensures high quality signal switching without gain deterioration and phase shift between the output and input voltages. Having zero phase shift is extremely important for phase lock loops, clock recovery circuits or any other application including phase shift sensitive circuitry.

[0069] When the resistances R₁, R₂(and R₃) are matched, the circuitry has more stable gain characteristics, being less dependent on external conditions. Moreover, it makes the stability requirements for the internal control means resistance much less stringent in view of the circuitry characteristics being mainly determined by the resistances R₁, R₂,R₃.

[0070] Introduction of the resistance R₁also allows to reduce the complexity of the input means. In the prior art, without introduction of R₁, there were stringent requirements for R₂to keep it as low as possible in order to have small values for R₂C_chchain time constant. For the circuitry of the invention it is not critical any more as follows from equations (3) and (5). As a result, the requirement for low values of R₂is eliminated, and the complexity of the input means is reduced by having smaller chip area and lower power consumption.

[0071] Thus, it will be appreciated that, while specific embodiments of the invention are described in detail above, numerous variations, combinations and modifications of these embodiments fall within the scope of the invention as defined in the following claims.