[0001] This relates generally to voltage controlled oscillator (VCO) technology, and more particularly to a VCO with a trifilar inductive coil.

BACKGROUND

[0002] A VCO is a device (i.e., oscillator) that outputs an oscillating signal with a frequency that is controlled by the level of an input voltage applied to the VCO. Therefore, a fixed DC input voltage to the VCO should ideally produce a fixed output frequency signal, but that input voltage also may be varied to vary the VCO output frequency. As to the latter, therefore, a modulating input signal may be applied to cause the VCO to output a signal with a modulating frequency (or phase).

[0003] As further background, FIG. 1 illustrates a schematic of a conventional VCO shown generally at 10. VCO 10 includes bias control circuitry 12 that may be constructed according to known principles for biasing VCO 10, as further explored below. One connection from bias control circuitry 12 is to a gate of a first nMOS transistor 14, which has its source connected to ground. The drain of first nMOS transistor 14 is connected to a source of a second nMOS transistor 16 and to a source of a third nMOS transistor 18.

[0004] VCO 10 also includes a transformer 20, shown in a dashed box and including a first inductor II and a second inductor 12, where polarities as between inductors II and 12 are shown according to the well-known dot convention. A first terminal Till of inductor II is connected to the drain of nMOS transistor 16, a second terminal T2I1 of inductor II is connected to the drain of nMOS transistor 18, and a center tap of inductor II is connected to a fixed voltage potential, shown as VDD. A first terminal T1I2 of inductor 12 is connected to the gate third of nMOS transistor 18, a second terminal T2I2 of inductor 12 is connected to the gate of second nMOS transistor 16, and a center tap of inductor 12 is connected to bias control circuitry 18. The oscillator output signal, vout, is provided as a differential signal between the respective drains of second nMOS transistor 16 and third nMOS transistor 18.

[0005] Generally, in operation, VCO 10 provides a frequency response in vout based on the inductance and parasitic capacitance of transformer 20, the parasitic capacitance of nMOS transistors 16 and 18, and the bias voltages from bias control circuitry 12, which further control a contribution to vout based on the biasing of nMOS transistor 14. Thus, energy oscillates between the inductance and capacitance, giving rise to the oscillating output vout. In the circuit, resistance also exists (which itself would tend to diminish the circuit's response), but the VCO 10 has a negative conductance (sometimes also referred to as a -R) to compensate for this resistance. In VCO 10, the negative conductance is achieved via the positive feedback provided by the cross-coupled configuration of nMOS transistors 16 and 18, relative to inductor 12. More specifically, the inductance of inductor II combines with capacitance to provide a resonating output while also inducing a signal into inductor 12, which is cross-coupled and thereby provides in-phase positive feedback to the gates of nMOS transistors 16 and 18, thereby sustaining vout.

[0006] The above and related approaches have served various needs, but they also provide various drawbacks. For example, when VCO 10 is implemented in an (e.g., silicon) integrated circuit, the transformer inductors are typically constructed using different layers of the back end metal process. Therefore, for the two inductor transformer, each inductor is typically built in a separate metal layer, thereby consuming a considerable amount of two-dimensional area, where area in itself can be a critical design consideration for numerous devices and applications. Moreover, various performance measures are desirable, having dedicated such transformer area for the VCO. A first and key such measure is power consumed. A second measure is phase noise, which is a figure of merit on accuracy of vout frequency for a given bias voltage, where such accuracy also includes susceptibility to jitter around the intended frequency tone at a given bias voltage.

SUMMARY

[0007] In described examples of a voltage controlled oscillator (VCO) for providing an oscillating output signal, the VCO includes a first inductor, and the oscillating output signal is responsive to a changing current through the first inductor. The VCO also includes: a second inductor, proximate the first inductor, coupled to a first cross-coupling stage; and a third inductor, proximate the first inductor, coupled to a second cross-coupling stage.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] FIG. 1 illustrates an electrical schematic of a conventional voltage controlled oscillator.

[0009] FIG. 2 illustrates an electrical schematic of a voltage controlled oscillator according to an example embodiment. [0010] FIG. 3A illustrates an exploded perspective view of three inductors in an example embodiment voltage controlled oscillator.

[0011] FIG. 3B illustrates a cross-sectional view of three inductors in an example embodiment voltage controlled oscillator.

[0012] FIG. 4 again illustrates the VCO of FIG. 2, with a few additional illustrated aspects.

DETAILED DESCRIPTION OF EXAMPLE EMBODFMENTS

[0013] FIG. 2 illustrates a schematic of an example embodiment voltage controlled oscillator (VCO) shown generally at 30. VCO 30 includes bias control circuitry 32 that may be constructed according to known principles for biasing a VCO and more particularly for biasing VCO 30, as further explored below. One connection from bias control circuitry 32 is to a gate of a first nMOS transistor 34, which has its source connected to a reference potential, such as ground. The drain of first nMOS transistor 34 is connected to a source of a second nMOS transistor 36 and to a source of a third nMOS transistor 38. Another connection from bias control circuitry 32 is to a gate of a first pMOS transistor 40, which has its source connected to a fixed supply voltage, designated VDD. The drain of first pMOS transistor 40 is connected to a source of a second pMOS transistor 42 and to a source of a third pMOS transistor 44.

[0014] VCO 30 also includes a trifilar transformer 30, which is a transformer with three different inductor coils as shown in a dashed box and including a first inductor II, a second inductor 12, and a third inductor 13, where polarities as between inductors II, 12, and 13 are shown according to the well-known dot convention. A first terminal Till of inductor II is connected to the drain of second nMOS transistor 36, a second terminal T2I1 of inductor II is connected to the drain of third nMOS transistor 38, and a center tap (or alternatively some other intermediate point between its terminals) of inductor II is connected to bias control circuitry 32. A first terminal T1I2 of inductor 12 is connected to the gate of third nMOS transistor 38, a second terminal T2I2 of inductor 12 is connected to the gate of second nMOS transistor 36, and a center tap (or alternatively some other intermediate point between its terminals) of inductor 12 is connected to bias control circuitry 32. A first terminal T1I3 of inductor 13 is connected to the gate of third pMOS transistor 44, a second terminal T2I3 of inductor 13 is connected to the gate of second pMOS transistor 42, and a center tap (or alternatively some other intermediate point between its terminals) of inductor 13 is connected to bias control circuitry 32. The drain of second pMOS transistor 42 is connected to the drain of second nMOS transistor 36 and also to terminal Till. The drain of third pMOS transistor 44 is connected to the drain of third nMOS transistor 38 and also to terminal T2I1. The oscillator output signal, vout, is provided as a differential signal between terminal Till (i.e., the same nodes as the drains of second nMOS transistor 36 and second pMOS transistor 42) and terminal T2I1 (i.e., the same nodes as the drains of third nMOS transistor 38 and third pMOS transistor 44).

[0015] Generally, in operation, VCO 30 provides a frequency response in vout based on the inductance and parasitic of transformer 30, the parasitic capacitance of nMOS transistors 36 and 38 and pMOS transistors 42 and 44, and the bias voltages from bias control circuitry 32, which further control a contribution to vout based on the biasing of first nMOS transistor 34 and the biasing of first pMOS transistor 40. In an alternative example, explicit capacitance through discrete devices also could be added, thereby further influencing the characteristics of vout. In all events, therefore, energy oscillates between the trifilar inductance and capacitance, giving rise to the oscillating output, and a negative cross-conductance is achieved through two different cross-coupled configurations, one with respect to inductor 12 and another with respect to inductor 13.

[0016] In view these points, the example embodiment VCO 30 provides numerous advantages.

[0017] One benefit of VCO 30 is that the power required to achieve an oscillating output of vout is reduced relative to conventional techniques, possibly by a factor greater than two. The example embodiment includes two cross-coupling stages, shown by example as an nMOS cross-coupling with nMOS transistors 36 and 38 and a pMOS cross-coupling with pMOS transistors 42 and 44. These stages thereby double the net get, as compared to VCO 10 of FIG. 1. Also, depending on implementation, a gain may be achieved between multiple coil pairs; for example, if inductor II is a primary coil, then it can induce a voltage (or current) boost into one or both of inductors 12 and 13, such as via the relative amount of turns as between the pair of inductors II and 12 or II and 13. Such additional magnetic boosting can further reduce DC power requirements to VCO 30.

[0018] Another benefit of VCO 30 is that separate biasing is available for both the nMOS transistor 34 to ground and the pMOS transistor 40 to VDD. In this respect, the sensitivity of each to noise can be separately or independently suppressed, via the respective gate potentials of nMOS transistor 34 and pMOS transistor 40. Indeed, this benefit has an additional potential benefit to relax standards of the voltage supply to VCO 30. Often, a low drop out (LDO) supply is used for VDD, and strict and cost-influencing requirements are placed on the LDO supply to allow it to suppress noise. The example embodiment's ability to separately suppress noise, therefore, permits the requirements on such an LDO to be reduced, thereby improving cost and efficiency considerations.

[0019] Another benefit of VCO 30 is that separate biasing is available for the center tap of all inductors II, 12, and 13. Again, therefore, noise influence associated with one device can be separated from noise influence associated with the other. Moreover, the example embodiment provides an improvement in gate swing, one for the PMOS side and one for the NMOS side.

[0020] FIG. 3 A illustrates a perspective exploded view, and FIG. 3B a side cross-sectional view, of a configuration in which each of inductors II, 12 and 13 may be formed in connection with known semiconductor and integrated circuit fabrication processes. In this example embodiment, each inductor is generally a same shape and may be formed so that a majority of the metal for the inductor is positioned in a different respective metal layer in a semiconductor process. As shown in the exploded view of FIG. 3 A, therefore, in the metal layers of a semiconductor process, inductor 12 would be formed from metal, below the formation of a metal inductor II, and inductor 13 would be formed from metal, above inductor II; this is also shown by cross-section in FIG. 3B, where intermediate (e.g., insulating) layers IL are formed between the inductors, for simplification. Therefore, given the two-dimensional (e.g., from a top-down view) shape and orientation in FIGS. 3 A and 3B, the shape and borders of inductors II, 12 and 13 are vertically aligned, so that the area consumed by the trifilar device in those two dimensions is no greater than for a conventional two-inductor device. This gives rise to another benefit, because typical inductors consume a considerable amount of two-dimensional area, particularly relative to the rest of the circuitry required to implement a VCO (and related circuitry). The example embodiment of FIG. 2 may be achieved via FIGS. 3A and 3B, with its additional inductor formed in a same two-dimensional space, by aligning it in the third dimension (e.g., vertically) in line with the other inductors. Accordingly, the various benefits described above are achieved without a two-dimensional increase in surface area.

[0021] FIG. 4 again illustrates VCO 30 of FIG. 2, with a few additional illustrated aspects. Specifically, as a trifilar coil VCO, an output of VCO 30 can be tapped from the respective differential signal across any of the three inductors II, 12, and 13. Thus, in FIG. 4, respective outputs voutl, vout2 and vout3 are shown. Also, across each such output is a respective tuning (i.e., variable) capacitor, CI, C2 and C3. In an example embodiment, each such capacitor is a combination of switched capacitors (for band tuning) and varactors (for continuous tuning). As an alternative, any of capacitors CI, C2 or C3 may be replaced with an explicit switched capacitor plus a MOS varactor in parallel. Moreover, continuous (or analog) tuning is not required for all of capacitors CI, C2 or C3, and each can be scaled differential and can have different bit sizes and other parameters, depending on the frequency of oscillation and whether multiple oscillation modes exist. Thus, VCO 30 can have multiple oscillation modes and selection of one (and suppression of the rest) also will factor in the choice and tuning on capacitors CI, C2, and C3.

[0022] From the above, the example embodiments provide a VCO with a trifilar inductive transformer with cross-coupling stages to improve numerous metrics, as compared to conventional techniques. In one example embodiment, a first cross-coupling stage is formed by nMOS transistors with respect to one inductor of the trifilar transformer, while a second cross-coupling stage is formed by pMOS transistors with respect to another inductor of the trifilar transformer. Separate biasing devices (e.g., transistors) exist in an example embodiment for respective ones of the cross-coupled stages and respective inductor center taps. The example embodiment construction may use area comparable in two dimensions to that used by a conventional configuration, while considerably outperforming that conventional configuration. Moreover, while various embodiments have been provided, various measures and architectures are adjustable according to application and other considerations. For example, FIGS. 3A and 3B show the each inductor in the trifilar transformer in a separate metal layer, but (in an alternative example) two or more inductors may be formed in the same layer, with connections thereto potentially extending to other metal layers.

[0023] Modifications are possible in the described embodiments, and other embodiments are possible, within the scope of the claims.