Varactor, in particular for radio-frequency transceivers

[0002] The present invention relates to a varactor, and in particular, a varactor for radio frequency transceivers.

[0003] As known, varactors are electronic devices having a variable capacity, selectable in particular through a biasing voltage, applied between a pair of terminals.

[0004] Varactors are used, for example, for tuning LC circuit radio-frequency transceivers. In fact, these circuits are adversely affected by production inaccuracies, and by the low quality factor of the passive components, and thus, in use, require adaptation of the actual reactance value, to adjust it to the planned values.

[0005] For this purpose a POS (Polysilicon-Oxide-Semiconductor) varactor has been recently proposed, described for example in the article “A ±30% Tuning Range Varactor Compatible with future Scaled Technologies”, by R. Castello, P. Erratico, S. Manzini, F. Svelto, VLSI Symp. on Circuits, Dig. Techn. Papers, June 1998, pp. 34-35, and illustrated in FIG. 1. In detail, a varactor 1 is formed in a wafer 2, only partially shown for easy of representation, comprising a substrate 3, of P-type, having a surface 8 and accommodating a well 4 of N⁻ type. In turn, the well 4 accommodates two biasing regions 5aand 5b, of N⁺ type, spaced from one another (but electrically connected, as shown in FIG. 1, by an electrical connection 20) and bias the well 4 to voltage V_N.

[0006] A gate region 6, made of polycrystalline silicon, extends above surface 8 of wafer 2, at an intermediate wafer portion 7 between biasing regions 5aand 5b, and is electrically isolated from the wafer portion 7 by a gate oxide region 9. The gate region 6 is biased to a voltage V_G, via an electrical connection line 21.

[0007] As apparent to those skilled in the art, the conductivity of the wafer portion 7 depends on the voltage difference V_VAR=V_G−V_Nexisting between gate region 6 and well 4. In particular, when the gate region 6 is biased to a positive voltage V_G, charges (electrons) are accumulated in the wafer portion 7 and the capacity of the varactor 1 increases to a value CMAX, beyond which it remains constant, and can no longer be modulated. This capacity value CMAX, equivalent to the capacity of gate oxide layer 9, is reached for values of V_VARslightly higher than 0 V (for example 0.3 V-0.4 V). If, on the other hand, voltage V_VARis negative, a depletion region 10 is formed in the wafer portion 7, the depth whereof increases along with the absolute value of inverse voltage V_VAR, thus causing the capacity of the varactor 1 to decrease down to a minimum value CMIN, when the inverse voltage V_VARreaches a negative threshold value V_T(for example of approximately −1.6 V). In fact, below the threshold value V_T, in the wafer portion 7, next to the surface 8 of the wafer 2, an inversion layer 11 is formed, comprising minority carriers (here gaps), thermally generated in the depletion region 10. The wafer portion 7 thus defines a capacity modulation region. For example, in a typical MOS process, CMAX is approximately equivalent to 500 nF/cm², and CMIN is approximately equivalent to 250 nF/cm².

[0008] However, the present production technology for integrated transceivers requires more extensive variability.

[0009] An embodiment of the present invention provides a varactor, the capacity of which can be modified within a wider range than those obtained at present.

[0010] According to the present invention, an embodiment of a varactor is provided, as defined in the claims.

[0011] For the understanding of the invention, an ebodiment is now described, purely by way of non-limiting and non-exhaustive examples, with reference to the attached drawings, in which:

[0012] FIG. 1 shows a cross-section through a wafer, incorporating a varactor of a known type;

[0013] FIG. 2 shows a perspective cross-section through a wafer, incorporating a varactor according to a first embodiment of the present invention;

[0014] FIG. 3 shows a plan view of a wafer comprising a varactor, in a second embodiment of the present invention; and

[0015] FIG. 4 shows a perspective cross-section through a wafer, incorporating a varactor according to a third embodiment of the present invention.

[0016] With reference to FIG. 2, in which parts equivalent to FIG. 1 are indicated with the same reference numbers, a varactor 50 is formed in a wafer 2 of semiconductor material, accommodating a well 4 of N⁻ type. Similar to FIG. 1, the well 4 comprises two biasing regions 5aand 5b, of N⁺ type, extending parallel, spaced from one another, for a length L1. The two biasing regions 5aand 5bare also connected electrically via a first electrical connection line 20, and bias the well 4 to voltage V_N.

[0017] In FIG. 2, the gate region 6, again of polycrystalline silicon, extends parallel to the biasing regions 5a, 5b, above the wafer portion 7, and is insulated by the gate oxide region 9, for a length L2, greater than L1. The gate region 6 is biased to voltage V_Gby a second electrical connection line 21.

[0018] In addition, as shown in FIG. 2, two charge extraction regions 15aand 15b, of P⁺ type, extend in continuation of one of the ends of the biasing regions 5aand 5b, for a length L3, equal to the difference between L2 and L1, for example to about one hundredth of L1. For example, L2 may have a length of 10 μm, L3 may be the minimum lithographic dimension (for example 0.35 μm), and L1 is 9.65 μm.

[0019] Thus, the charge extraction regions 15aand 15band the well 4 form a first pair of PN junctions 16aand 16a, and, the charge extraction regions 15aand 15band the biasing regions 5aand 5bform a second pair of PN junctions 17aand 17brespectively. The charge extraction regions 15aand 15bare electrically connected to one another by a third electrical connection line 22 and are biased to voltage V_p.

[0020] For example, the well 4 may have a doping level of 2×10¹⁷−3×10¹⁷atoms/cm³; the biasing regions 5aand 5band the charge exraction regions 15aand 15bmay have a doping level of approximately 10²⁰atoms/cm³, with ions imparting opposite types of conductivity.

[0021] The varactor 50 functions as follows.

[0022] Similarly to the varactor 3 of FIG. 1, the capacity of the varactor 50 is determined by voltage V_VAR=V_G−V_Napplied between the gate region 6 and the well 4. In addition, voltage V_pof the charge extraction regions 15aand 15bis lower than voltage V_N, such that the pairs of PN junctions 16aand 16b, 17aand 17bare always inversely biased when in use. For example, if the biasing regions 5aand 5bare connected to ground (V_N=0 V), and V_Gvaries between +0.4 V and −3 V, the voltage V_pis advantageously −3 V.

[0023] When voltage V_VARis positive or negative, but is greater than threshold voltage V_T, the charge extraction regions 15a, 15bdo not intervene, and the varactor 50 acts in a manner similar to that described with reference to the varactor 1 of FIG. 1.

[0024] When the voltage V_VARreaches the threshold value V_T, the minority carriers (holes) thermally generated in the depletion region 10, accumulate near the gate oxide region 9, by virtue of the electrical field. However, since the extraction regions 15aand 15bare set to a lower voltage than the well 4, they attract the minority carriers (holes), thus creating a diffusion current I_Dthrough the first pair of PN junctions 16aand 16b, inversely biased. Consequently, the minority carriers are removed from the depletion region 10, and the formation of an inversion layer below the gate oxide region 9 is prevented.

[0025] The phenomenon of thermal generation of carriers can be considered virtually stationary, compared with the frequency of the voltage signals typical of the applications commonly using varactors (of approximately hundreds or thousands of MHz). The variations of the diffusion current I_Dthus do not influence variations in the capacity of the varactor 50, which therefore are determined only by the depth of the depletion region 10, set by voltage V_VAR.

[0026] The described varactor 50 has the following advantages. First, the values selectable for the capacity of the varactor 50 may vary more extensively than for conventional varactors. In fact, as previously stated, removal of minority carriers from the depletion region 10 prevents the formation of the inversion layer 11 (see, e.g., FIG. 1). Consequently, the capacity of the varactor 50 can assume lower values than the value reached when the inverse voltage V_VARis V_T. For example, the capacity of the varactor 50 can vary between approximately 4.4 pF and 1.7 pF, when V_VARvaries between +0.4 V and −3 V. In addition, in the voltage range comprised between the maximum voltage and the threshold voltage V_T, junctions 16a, 16bare inversely biased (thus forming diodes in the off state), and thus do not modify the normal behavior of the varactor 50.

[0027] In addition, the reduction of the minimum capacity CMIN, obtained by removing the minority carriers from the depletion region 10, improves the linearity of the varactor 50, thus eliminating the distortion effects occurring in known varactors, when voltage V_VARapproaches threshold voltage V_T.

[0028] The fact that the extraction regions 15aand 15bare integrated with the device constitutes a further advantage. In fact, these regions 15aand 15bcan be biased to inverse voltages compatible with the supply voltages provided to the circuitry interacting with the varactor 50 and do not require specific components for biasing.

[0029] Finally, it is apparent that many modifications and variants can be made to the embodiment of the varactor 50 described and illustrated here, within the scope of the invention, as defined in the attached claims.

[0030] In particular, as illustrated in FIG. 3 (not to scale for the sake of simplicity), the gate region (here indicated at 6′) can comprise a plurality of fingers 30, electrically connected at a respective end by a connection region 31, such that, in practice, the gate region 6′ is comb-shaped. The fingers 30 are arranged parallel to one another, each overlapping a respective capacity modulation portion. When seen from above, the fingers 30 alternate with zones 32, formed inside the well 4, and each zone comprises a biasing region 32a, of N⁺ type, and an extraction region 32b, of P³⁰type. All biasing regions 32aare biased to the same voltage V_N, as shown schematically by a connection line 20′, and all the extraction regions 32bare biased to the same voltage V_p, as shown schematically by a connection line 22′. In this case, the overall capacity of the varactor is equivalent to the sum of the capacities associated with each finger 30, since the fingers are in parallel to each other.

[0031] In addition, with reference to FIG. 4, intermediate zones 40 of N-type, with a lower doping level than the biasing regions 5a, 5b, but higher than the well 4 (for example of approximately 10¹⁸atoms/cm³-10¹⁹atoms/cm³) can be formed respectively between the biasing region 5aand the respective extraction region 15a, and between the biasing region 5b, and the respective extraction region 15b. In this case, the length of the gate region 6 is given by the sum of the dimensions, in the same direction, of each of the biasing regions 5a, 5bof the respective intermediate zone 40, and of the respective extraction region 15a, 15b.

[0032] The intermediate zones 40 advantageously do not require corresponding masking steps, but can be formed simultaneously with regions belonging to the components of the circuit connected to the varactor 50 (typically the drain extension regions of NMOS transistors). Thereby, the capability of withstanding voltages is improved between each biasing region 5a, 5b, and the respective extraction region 15a, 15b.

[0033] Finally, the varactor 50 can be manufactured in dual form, in particular by forming inside a well of P-type, biasing regions of P⁺ type, and extraction regions of N⁺ type. In this case the voltages applied to the different regions have an opposite polarity to the described one.

[0034] Therefore, the above description of illustrated embodiments is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Accordingly, the invention is not limited by the disclosure, but instead the scope of the invention is to be determined entirely by the following claims, which are to be construed in accordance with established doctrines of claim interpretation.