Abstract
A tapped transmission line for distributing an electrical signal, such as an RF signal, to multiple modules, and/or aggregating signals from multiple modules. Embodiments of the invention provide a tapped transmission line based on a transmission-line medium along which tap elements are dispersed, so that the tap elements have a predominantly capacitive loading of the transmission-line medium. Methods for compensating the loss of the transmission-line medium are presented as well. Applications for distribution of transmitted signals, of local oscillator signals, and to aggregation of signals from multiple oscillators are disclosed. The invention is particularly applicable to integrated circuits (IC, ASIC, RFIC), and to multichannel RF systems such as phased array and MIMO systems.
Claims
1. A tapped transmission line for distributing a transmitted signal to a plurality of modules, the tapped transmission line comprising: a first transmission-line medium for carrying a first electrical signal; a second transmission-line medium for carrying a second electrical signal; a first plurality of operational tap elements connected to the first transmission-line medium at intervals along a physical length of the first transmission-line medium; and a second plurality of operational tap elements connected to the second transmission-line medium at intervals along a physical length of the second transmission-line medium; wherein: the operational tap elements exhibit predominantly capacitive impedance to the respective transmission-line media; and each module of the plurality of modules is connected to a respective operational tap element of the first plurality of operational tap elements and is connected to a respective operational tap element of the second plurality of operational tap elements.
2. The tapped transmission line of claim 1, wherein: the first electrical signal is an in-phase replica of the transmitted signal; and the second electrical signal is a quadrature replica of the transmitted signal.
3. The tapped transmission line of claim 1, wherein an operational tap element is selected from a group consisting of: an amplifier, a mixer, a frequency multiplier, and a vector modulator.
4. The tapped transmission line of claim 1, wherein a transmission-line medium is a differential transmission-line medium.
5. The tapped transmission line of claim 1, wherein the transmitted signal is a local oscillator signal.
6. The tapped transmission line of claim 1, wherein the transmission-line media distribute respective multiple electrical signals to the modules.
7. The tapped transmission line of claim 1 wherein a transmission-line medium is tapered.
8. The tapped transmission line of claim 1, wherein the transmitted signal is a radio frequency (RF) signal.
9. The tapped transmission line of claim 1, wherein outputs of the plurality of modules are not connected to a common transmission-line medium.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
[0012] The subject matter disclosed may best be understood by reference to the following detailed description when read with the accompanying drawings in which:
[0013] FIG. 1 illustrates a prior art scheme for signal distribution to multiple consumers, based on a splitting tree.
[0014] FIG. 2 illustrates a prior art scheme for signal aggregation from multiple sources, based on a combining tree.
[0015] FIG. 3 illustrates signal distribution to multiple consumers by a voltage-sensing tapped transmission-line medium according to an embodiment of the invention.
[0016] FIG. 4 illustrates signal aggregation from multiple sources by a current-injection tapped transmission-line medium according to an embodiment of the invention.
[0017] FIG. 5 is an illustration of signal distribution to multiple consumers, accounting for the input capacitance of the consumers according to an embodiment of the invention.
[0018] FIG. 6 illustrates signal aggregation from multiple sources, accounting for the output capacitance of current-injecting tap elements according to an embodiment of the invention.
[0019] FIG. 7a illustrates sections of transmission-line medium, periodically loaded with tap element capacitances according to an embodiment of the invention.
[0020] FIG. 7b illustrates transmission-line medium sections as approximated by L-C sections according to various embodiments of the invention.
[0021] FIG. 7c illustrates L-C approximation of a transmission-line medium, as further loaded by tap element capacitances to increase the capacitive component according to an embodiment of the invention.
[0022] FIG. 8a illustrates a multi-port receiver having a local oscillator drive signal transmission-line medium distributed to multiple downconverting mixers, according to an embodiment of the invention.
[0023] FIG. 8b illustrates an embodiment of passive switching down-converter mixer with gates of the switch transistors being driven by an LO signal distributed through the transmission-line medium according to an embodiment of the invention.
[0024] FIG. 8c illustrates a Gilbert-cell-based down-converter mixer according to an embodiment of the invention.
[0025] FIG. 9a illustrates a differential transmission-line medium periodically loaded with differential transistor pairs serving as tap element amplifiers according to an embodiment of the invention.
[0026] FIG. 9b illustrates the case where a differential pair forms part of a polarity-controlled cascade circuit according to an embodiment of the invention.
[0027] FIG. 9c shows a related embodiment of the invention, where the polarity of the tapped signals can be chosen.
[0028] FIG. 10a illustrates a multi-port transmitter with embedded vector modulator tap elements within each tap module according to an embodiment of the invention. The 90-degree split of the LO is shared among all the units.
[0029] FIG. 10b illustrates a vector-modulated tap element amplifier according to an embodiment of the invention.
[0030] FIG. 11a illustrates compensation for the transmission-line medium loss by use of a tapered transmission-line medium according to an embodiment of the invention, wherein the increasing characteristic impedance of the line increases the voltage swing and compensates for transmission-line medium loss.
[0031] FIG. 11b illustrates compensation for transmission line medium loss by use of a reflective termination to enhance the voltage by using standing waves according to an embodiment of the invention.
[0032] FIG. 12 conceptually illustrates negative resistance compensation for transmission-line medium loss according to an embodiment of the invention.
[0033] FIG. 13 illustrates negative resistance compensation for transmission-line medium loss by cross-coupled transistor pairs according to an embodiment of the invention.
[0034] FIG. 14 illustrates multiport transmitter with neutralized tap element amplifiers to reduce effective input capacitance and introduce negative admittance compensation for transmission-line medium loss at the input of each tap element amplifier a positive feedback produced by neutralization capacitors, according to an embodiment of the invention.
[0035] FIG. 15 illustrates a transmission-line medium loaded with negative resistances for adjustable loss compensation, having end-switches to test for stability, according to an embodiment of the invention.
[0036] FIG. 16 illustrates a prior art distributed amplifier having tapped transmission-line media for combining amplified signals to achieve broadband amplification of a single signal.
[0037] FIG. 17 illustrates a synthesizer with multiple voltage-controlled oscillators (VCOs) aggregated into a transmission-line medium using common synthesizer components (divider, phase detector, loop filter) and modulator, according to an embodiment of the invention.
[0038] For simplicity and clarity of illustration, elements shown in the figures are not necessarily drawn to scale, and the dimensions of some elements may be exaggerated relative to other elements. In addition, reference numerals may be repeated among the figures to indicate corresponding or analogous elements.
DETAILED DESCRIPTION
[0039] Modern RF communications and radar systems use smart antennas (such as beamforming or MIMO antennas) having numerous antenna elements attached to numerous transceiver modules. In such systems there is a need to distribute an RF signal (or an auxiliary signal such as a local oscillator signal) to multiple modules. A common prior-art solution, illustrated in FIG. 1, places an input 101 into a cascade of splitters 102-104a/104b-105a/105b/105c/105d, typically with amplifiers (e.g., an amplifier 103), to produce multiple replicas of the desired signal at outputs 106a/106b/106c/106d/106e/106f/106g/106h. However, the use of such passive power splitters is space-consuming, as power splitters are commonly of a size commensurate with the wavelength of the signal. Moreover, each split into 2 introduces a 3 dB power loss (due to power conservation), thus requiring amplifiers (e.g., amplifier 103) to compensate for this loss—amplifiers which consume both power and space. The transmission-line media of tree distribution network 100 may become space-consuming and introduce routing difficulties.
[0040] Similar issues arise when signals need to be aggregated. FIG. 2 illustrates a commonly-encountered prior art tree of combiners 200, which requires numerous combiners (e.g., a combiner 201), amplifiers (e.g., an amplifier 202), connected by an intricate network of transmission-line media.
[0041] FIG. 3 illustrates an embodiment of the present invention, in which a transmission-line medium 301 periodically loaded with operational tap elements is used to distribute and/or aggregate signals along a physical length 310 which is divisible into length intervals (such as an interval 311, an interval 312, and an interval 313). Transmission-line medium 301 has a signal input 302 at one end and a terminating load 303 at the other end, and is tapped with multiple operational tap elements, including initial operational tap elements 304a/304b/304c, intermediate operational tap elements indicated by an ellipsis 304d, and a final operational tap element 305e. In a related embodiment, operational tap elements 304a, 304b, 304c, 304d, and 304e are amplifiers or buffers. In another related embodiment, operational tap elements 304a, 304b, 304c, 304d, and 304e are substantially identical. In an additional related embodiment, operational tap elements 304a, 304b, 304c, 304d, and 304e are substantially identical, whatever their structure or function. In yet a further related embodiment, operational tap elements 304a, 304b, 304c, 304d, and 304e are distributed at substantially identical physical inter-tap distances along transmission-line medium 301. In the non-limiting example of this embodiment shown in FIG. 3, the physical inter-tap distance between tap element 304a and tap element 304b is a distance d 305. Likewise, the inter-tap distance between tap element 304b and tap element 304c is substantially the same distance d 305. The general case is illustrated for n tap elements 304c/304d/304e, where n≥2, and the physical inter-tap distance between tap element 304c and tap element 304e is substantially d(n−1) for any value of n, so that the physical inter-tap distance between any two adjacent taps is substantially the same distance d.
[0042] In the case of signal distribution (FIG. 3), a signal propagates along transmission-line medium 301, from source input 302, to first amplifier 304a, then to second amplifier 304b, and so on. Each amplifier causes minimal disturbance to the signal on transmission-line medium 301, allowing the signal to propagate substantially unchanged. In still another related embodiment, load 303 has a resistance equal to the characteristic impedance of transmission-line medium 301 to prevent signal reflection back toward input 302. Signal reflection may have undesired effects unless controlled or utilized as discussed below.
[0043] FIG. 4 illustrates signal aggregation via a transmission-line medium 401 according to an embodiment of the present invention. Tap elements 404a/404b/404d and additional tap elements indicated by an ellipsis 404c are each current sources that inject a signal into transmission-line medium 401. The injected signals combine and propagate in both directions along transmission-line medium 401. Tap elements 404a/404b/404c/404d present high shunt impedances to transmission-line medium 401, and therefore the contribution of any one tap element is not significantly attenuated by the other tap elements. The combined signal eventually appears at both ends of the line. At one end, a first output 402 feeds another circuit (not shown), and at the other end a second output 403 either feeds another circuit (not shown) or is terminated by a matched load (not shown).
[0044] FIG. 3 and FIG. 4 show simplified views of the system; in practice the impedance presented by a tap element amplifier or source is not negligible compared to the characteristic impedance of the line. FIG. 5 illustrates the case where amplifiers 504a/504b/504c have inputs exhibiting respective input capacitance 505a/505b/505c, even if the capacitance is just that of a field-effect transistor (FET) gate. Similarly, FIG. 6 illustrates the case where current sources 604a/604b/604c have outputs exhibiting respective output capacitance 605a/605b/605c, even if the capacitance is just that of a transistor drain.
[0045] FIG. 7a illustrates multiple distributed tap element capacitances 704a/704b/704c on a transmission-line medium 701, causing a cumulative effect. As the signal propagates along transmission-line medium 701 past each tap, the related capacitance causes cumulative disturbances.
[0046] The effect of capacitive inclusions along the line is best understood by recalling that a continuous transmission-line medium can be approximated by a ladder network of series inductances and shunt inductances. FIG. 7b illustrates just such a ladder network, with series inductances 711a/711b/711c/711d and shunt capacitances 714a/714b/714c/714d. A network composed of identical two-ports exhibits at each frequency a characteristic impedance and a per-stage response, which can then be mathematically raised to a power to obtain the composite response. Lossless networks can exhibit passbands (at which the signal is passed along, possibly with a phase shift, but without appreciable reflection or attenuation) and stopbands (at which the signal is reflected and attenuated). The L-C network approximates a transmission-line medium with characteristic impedance Z
[00001]
and having a delay per stage of
[00002]
up to a cutoff frequency
[00003]
implying that a transmission-line medium, which is composed of infinitesimal L-C segments, has an infinite bandwidth. It also implies that the L-C ladder approximation holds so long as each section is shorter than ≈λ/6. If the line is loaded periodically with capacitances C.sub.tap, and each line section is shorter than λ/6, then the line can be approximated by a ladder network in which the inductances are L.sub.line=τZ.sub.0, and the capacitances are C.sub.line=τ/Z.sub.0. The line loaded with the C.sub.tap capacitors can now be approximated by an equivalent L-C line, in which the composite capacitances have the value C.sub.line+C.sub.tap. The implications of this is that:
[0047] (1) the characteristic impedance is lowered according to
[00004]
[0048] (2) the delay of each section is increased according to
[00005]
[0049] (3) the composite line now has a cutoff frequency of
[00006]
which gets lower as the tap element capacitance is increased. Therefore, the maximal value of tap element capacitance that can be absorbed into an equivalent transmission-line medium is closely related to the operating frequency.
[0050] A particular design example is a distribution network for millimeter wave RFIC including multiple T/R modules, spaced 200 microns from each other. With a effective dielectric constant of ˜4, the delay of 200 micron line segment is 1.3 picoseconds, and it can be represented (for Z.sub.0=50Ω) by an L-C approximation with L=0.066 nH and C=26 if. The target operating frequency is ˜80 GHz. For this operating frequency, approximating each 200 micron section with a single LC section is too coarse, but simulation confirms that tap element capacitances up to ≈50 fF can be used, while having a cutoff frequency high enough (above 120 GHz).
[0051] Several embodiments of the invention are illustrated below. The examples below use differential transmission-line media, as the use of differential mode components is common in RFICs to reduce signal leakage and to reduce RF signal in ground and in supply lines. It should be further noted that the schematics of the amplifier circuits are shown in a basic form, omitting numerous details, such as tail-current sources, bias voltages, and drain loads.
[0052] A first embodiment, illustrated in FIG. 8a, attaches local oscillator (LO) 803 inputs of multiple mixers 804 and 805 to transmission-line media 801 and 802, to efficiently implement a multichannel receiver. Several examples of mixers can be considered. FIG. 8b illustrates a “switching mixer” composed of two pairs 811 and 812 of FET switches, the gates of which are driven by the local oscillator. A second embodiment, illustrated in FIG. 8c, uses Gilbert cell multipliers 821 and 822 as mixers—in this case the local oscillator signal can be applied either to top transistors 823 or to bottom transistors 824, while the signal is applied to the other input to the mixers. This embodiment obviates the need to use a LO buffer amplifier in each receiver; only one buffer amplifier is needed at the input to the transmission-line medium to develop an adequate voltage swing at the LO port of the mixer.
[0053] A second embodiment, illustrated in FIG. 9a for a differential transmission-line medium 901/902 periodically loaded with differential transistor pairs (typically like a pair 903, with an ellipsis 904 indicating additional similar differential transistor pairs) serving as tap element amplifiers, such as for distributing a transmitted signal to multiple Tx/Rx modules. These tap element amplifiers can be all active at the same time, as in phased array radars, or one at a time, as in MIMO radars. The amplifier can be a single differential pair 903, as illustrated in FIG. 9a, or it can utilize a cascode amplifier 913, as shown in FIG. 9b, with the advantage of lower Miller capacitance being reflected to the inputs. The term “cascode” refers to a combination of a common-drain and common-gate stages (in the case of FETs), well known in the art. The cascode configuration can be further augmented as shown in FIG. 9c, to include two top transistors 923 (as in a Gilbert cell 922). Selecting one of the two top-transistors 923 allows controlling signal polarity (as in BPSK modulation). Applying bias to both pairs 923 and 924 allows both gain and polarity control.
[0054] FIG. 10a illustrates an extension of the embodiments of FIGS. 8a-8c and 9a-9c to implement a phased array transmitter with phase control in each transmit module, such as a transmit module 1000. In this related embodiment two transmission-line media are used, one including a differential transmission-line medium 1003/1004 for carrying an in-phase replica of the transmitted signal, and a second, differential transmission-line medium 1001/1002 for carrying a quadrature replica of the transmitted signal. At module 1000, both lines 1001/1002 and 1003/1004 are tapped with a mixer 1005, e.g. Gilbert cell 922 in FIG. 9c. By combining the currents of multiplied signals at the drains of top transistors 1015, as illustrated in FIG. 10b, and applying cos φ and sin φ gains to the two mixers, a vector-modulated signal is obtained, with a phase shifted by φ. In this case, a substantial saving in complexity and cost is obtained by implementing the 0-90 degree split of the signal only once, at the input to the distribution lines, rather than on a per-module basis.
[0055] A third exemplary embodiment is related to aggregation of multiple signal sources. An exemplary application in mind is creating a broadband frequency synthesizer, which requires (for phase noise optimization) multiple voltage controlled oscillators (VCOs), of which one is active at a time, as exemplified in FIG. 17. By attaching a voltage-to-current converter to each VCO (e.g. a differential transistor pair with output at the drain), and connecting the current outputs to a transmission-line medium, an aggregation network is obtained. As an example, both ends of the transmission can be utilized, one for the divider of the synthesizer, and one for a vector modulator to generate the desired signal.
[0056] It is worth noting that at the circuit level implementation commonly the tap elements require same DC voltage—for example V.sub.G when gates of transistors are connected to the transmission-line medium, or V.sub.DD the drains are connected. In these cases, the tap element circuits can be DC-coupled to the line; the supply of this voltage can be performed jointly through the transmission-line medium, and the voltage generation and stabilization circuitry (such as current mirrors, LDOs (low drop-off regulators) etc.) can be implemented only once rather than per-module.
Loss Compensation
[0057] In practical implementations of the above-described principle, the transmission-line media exhibit loss. This has the detrimental effect of the signal not being of same amplitude at all the taps. For example, transmitted signals can differ in amplitude, or the receive mixers can exhibit different conversion loss due to a differing LO drive.
[0058] FIG. 11 illustrates alternative techniques for compensating for transmission-line medium loss. Tapered transmission-line media are known to operate as transformers which convert lower voltage at lower impedance to a higher voltage at higher impedance. FIG. 11a illustrates compensation for the transmission-line medium loss by use of a tapered transmission-line medium 1101, wherein the increasing characteristic impedance of the line increases the voltage swing and compensates for transmission-line medium loss. FIG. 11b illustrates compensation for the transmission-line medium loss by use of a reflective termination 1111 to enhance the voltage through utilization of standing waves towards the end of the line.
[0059] FIG. 12 conceptually illustrates compensating for transmission-line medium loss by attaching negative resistances along the transmission-line medium. Negative resistances according to various embodiments of the present invention are exhibited by negative resistance operational tap elements. The loss within a section of transmission-line medium is roughly modeled by a resistance to ground. In turn, a resistance to ground can be countered by a negative resistance of the same magnitude connected in parallel to the loss resistance. Therefore, according to an embodiment of the present invention, operational negative resistance tap elements, which exhibit negative resistance (or conductance) are connected to successive small sections of the transmission-line medium, so that the resulting structure behaves as a transmission-line medium of reduced loss per unit length. In related embodiments, the negative resistance is implemented by active circuits of the operational tap elements, which typically consume power in order to replenish the energy of the signal which is lost in the transmission-line medium. According to these embodiments, when the magnitude of negative resistance is high enough, the transmission line exhibits amplification, with the signal increasing as it propagates. Amplification is generally undesirable, as it detracts from the uniformity of signal level experienced by the tap elements. Another undesirable side-effect of amplification is oscillation.
[0060] FIG. 13 illustrates an embodiment of the invention that provides compensation for transmission-line medium loss by attaching cross-coupled transistor pairs along the transmission-line medium. The cross-coupled transistor pair is known to exhibit a negative differential transconductance—an effect commonly used in oscillators. Here it is used to compensate for the transmission-line medium loss. Preferably, the differential pair is augmented with frequency-selective circuitry so that the negative transconductance is predominant at higher frequencies, at which the transmission-line medium loss is high, and is lower at the lower frequencies. According to a related embodiment, in the case of constant-envelope signal along the transmission line, the cross-coupled pairs operate in a saturated, rather than linear, regime.
[0061] FIG. 14 illustrates an alternative embodiment of compensating for transmission-line medium loss by producing negative resistance component at the input of each operational tap element amplifier using a positive feedback produced by neutralization capacitors. The neutralization capacitors counter the effect of Miller capacitance due to C.sub.DG capacitance of the transistors, thus decreasing the loading of the transmission-line medium. If the value of the neutralization capacitors exceeds C.sub.DG, a positive feedback is created so that the input exhibits some negative resistance, which can be utilized to reduce the effective loss of the transmission-line medium.
[0062] In certain embodiments of the present invention, the negative resistances are provided by dedicated operational negative resistance tap elements. In other embodiments, however, negative resistances are provided by operational tap elements which may also provide other functionalities in addition to negative resistances.
[0063] Structures containing negative transconductance/resistance are prone to instability and oscillation. This tendency can be countered by controlling the value of the transconductance, to the extent that the losses are compensated but the stability is not compromised. As a non-limiting example, a common way to control the negative transconductance is by controlling the tail current of a cross-coupled differential pair. The value of the transconductance needs to be set to a value that compensates for the losses but avoids instability and oscillation. FIG. 15 illustrates an arrangement according to an embodiment of the present invention for testing a loss-compensated transmission-line medium 1501/1502 for stability so that the amount of compensation can be adjusted. The method is based on temporarily terminating the transmission-line media with reflective impedances (such as short-circuiting switches 1503 and 1504) and checking whether oscillation has occurred. Embodiments of the invention provide sensing of oscillation, for example, by a power detector that detects the presence of a signal with no input signal applied. Once the value of the transconductances is adjusted below the oscillation threshold, short-circuiting switches 1503 and 1504 are opened and the circuit is restored to the operational state, in which the effective loss of the transmission-line medium is significantly lower, and the signal along the line is more nearly uniform.
APPLICATIONS
[0064] Embodiments of the present invention have numerous potential applications in multichannel RF systems, such as MIMO and phased array transmitters, receivers, transceivers, radars etc. The embodiments disclosed herein for signal distribution and aggregation are applicable to a variety of technologies, and in particular to integrated circuits (IC, ASIC, RFIC), where space savings and power consumption savings are paramount.
