CN200983296Y - Non equal plane spiral inductor - Google Patents

Abstract

A non-equal planar spiral inductor in an integrated circuit, at least one metal spiral coil in the spiral inductor is in a different plane from any other metal spiral coil, and the coils in different planes are realized on dielectric layers with different heights. Compared with the traditional planar inductor, the non-planar spiral inductor of the utility model has lower influence of eddy current effect and proximity effect at high frequency, resulting in a decrease of series resistance Rs, thereby improving the quality factor Q value of the inductor; the preparation process is compatible with the conventional CMOS process ; Can improve the performance of important functional units of the CMOS radio frequency front end.

Description

Non-equal Plane Spiral Inductor

technical field

The utility model belongs to the technical field of integrated circuit device manufacturing, specifically relates to the structure and manufacture of a non-equal planar spiral inductor, and is suitable for deep submicron/micron integrated circuit technology.

Background technique

In the CMOS radio frequency integrated circuit (RFIC), the spiral inductor is a key component and one of the most difficult components to design and master in the circuit. Its performance parameters directly affect the performance of the radio frequency integrated circuit. The on-chip inductor can realize the integration of the inductor in the radio frequency integrated circuit, thus contributing to the realization of the system on chip of the radio frequency integrated circuit.

On-chip spiral inductors are mostly wound on silicon substrates by metal thin films. Compared with traditional wire-wound inductors, on-chip spiral inductors have the advantages of low cost, easy integration, low noise, and low power consumption. More importantly, they can be used with Compatible with today's CMOS processes. In recent years, with the development of mobile communication towards miniaturization and low power consumption, there are more and more researches on making high-quality on-chip passive devices compatible with CMOS technology. However, the parasitic effects in the spiral inductor, such as the parasitic capacitance and resistance of the substrate, the parasitic capacitance and resistance of the metal conductor, and the parasitic resistance due to effects such as eddy current loss, will all have an impact on the performance of the inductor. For substrate loss, two major categories of solutions have been proposed: on the one hand, people use methods such as ion implantation or making porous silicon layers to achieve selective formation of semi-insulating silicon substrates; A graphic ground shielding layer is inserted between the substrates, or a PN junction isolation is formed in the silicon substrate to reduce substrate loss.

If the substrate loss is ignored, the spiral inductor mainly considers the loss of the metal conductor. Under normal circumstances, people generally use a planar spiral inductor, that is, all the metal coils of the inductor are on the same plane, and the structure is relatively simple, which causes loss at this time. The main reason is the skin effect and proximity effect of metal conductors. The skin effect is inevitable at high frequencies, and its strength is determined by the material and frequency; the proximity effect is due to the fact that the metal coils are close to each other and the electromagnetic interference causes the current to be uneven in the conductor cross-section. The flow makes the resistance increase, and the proximity effect can be reduced by increasing the distance between the inductor coils to reduce the loss and improve the performance of the inductor. Increase the distance of the inductance coil, if the outer coil is moved outward, the size of the inductance will be increased, which is not conducive to the miniaturization of the device; if the inner coil is moved inward, the electromagnetic field in the center of the inductor is the strongest, and the coil Moving to the inner center will increase the loss, which is not conducive to the improvement of inductance performance. Therefore, a new type of inductance structure is needed to achieve the purpose of improving inductance performance without increasing the size of the inductance and causing other losses.

Utility model content

The purpose of the utility model is to provide a non-equal planar spiral inductor in an integrated circuit that can improve the Q value of the inductance quality factor.

The utility model is realized through the following technical solutions: a non-equal planar spiral inductor in an integrated circuit, at least one turn of the metal spiral coil in the spiral inductor is in a different plane from any other metal spiral coil, and the The coils are realized on dielectric layers of different heights.

Wherein, the innermost coil of the spiral inductor is in a different plane from the rest of the metal spiral coils.

The utility model moves the inner coil of the spiral inductor upwards, that is, places it on different wiring layers, so that the distance between the coils can be increased and the proximity effect can be reduced without increasing the size of the inductor and causing other losses. Thereby reducing loss. Compared with the traditional planar inductor, the spiral inductor of the present invention reduces the influence of the eddy current effect and the proximity effect at high frequencies, resulting in a decrease in the series resistance R _s , thereby improving the quality factor Q value of the inductor; the preparation process is compatible with the conventional CMOS process; It can improve the performance of important functional units of the CMOS radio frequency front end.

Description of drawings

FIG. 1A is a cross-sectional view of current distribution in a metal wire.

FIG. 1B is a top view of current distribution in a metal wire.

Figure 2 is an illustration of the eddy current effect of a spiral inductor.

Fig. 3 is a vector distribution diagram of the spiral inductor magnetic field.

Fig. 4 is a top view of the structure of the square spiral inductor involved in the present invention.

5A is a cross-sectional view of a planar spiral inductor.

5B is a cross-sectional view of a non-planar spiral inductor.

Among them, N refers to the number of turns of the metal coil of the inductor, D _out refers to the outer diameter of the inductor, D _in refers to the inner diameter of the inductor, w _n refers to the wire width of the metal conductor of the inductor, and s _n refers to the distance between the two metal conductors of the inductor. The spacing, M ₁ is the metal layer of the inductor used as a lead, M ₂ is the metal lower layer of the inductor coil, M ₃ is the metal upper layer of the inductor coil, Δh is the vertical distance between the two metal layers.

Detailed ways

In integrated circuits, the loss of radio frequency passive devices is mainly composed of two kinds of losses: substrate loss and metal conductor loss. The utility model mainly aims at reducing the loss of metal conductors and proposes an optimized inductance structure with increased number of metal layers. The design method of this structure is mainly based on reducing the proximity effect of the metal conductor in the spiral inductor.

Any metal is resistive under normal conditions. When the high-frequency signal current flows through the metal conductor, ohmic loss and eddy current loss are generated at the same time. The former is caused by the non-zero resistivity of the metal material, and the latter is caused by the eddy current.

Ohmic loss is related to the resistance of the metal conductor, and the resistance value is proportional to the resistivity and total length of the metal conductor material, and inversely proportional to the width and thickness of the metal conductor. In order to reduce the ohmic loss, the metal conductor material with low resistivity is generally selected and the thickness of the metal layer is appropriately increased without changing the geometric structure of the inductor. Because the total length of the metal conductor is closely related to the number of turns of the inductor coil, and the width of the metal conductor is closely related to the inner and outer diameters of the inductor coil, these will affect other parameters of the inductor.

Eddy current losses are related to the frequency of the signal and the inductive structure. When the high-frequency alternating current passes through the straight wire, the distribution of the current on any cross section of the straight wire is not uniform, most of the current is transmitted along the surface layer of the wire, and there is basically no current transmission near the central axis of the wire, as shown in Figure 1A As shown in Figure 1B, this is the so-called skin effect. The skin depth formula is as follows:

$\mathrm{\δ}=\sqrt{\frac{\mathrm{\ρ}}{\mathrm{\π\μf}}}$ Formula 1)

It can be seen from formula (1) that the higher the frequency of alternating current, the smaller the skin depth will be. Since the inductance series equivalent resistance Rs is inversely proportional to the skin depth, the greater Rs will be.

As a result of the current flowing adjacent to the metal, an alternating magnetic field is generated that passes through the metal. According to Lenz's law, the metal generates eddy currents to suppress changes in the magnetic field, a phenomenon known as the proximity effect. The proximity effect is essentially caused by eddy currents. Unlike the skin effect, the skin effect is caused by the electromagnetic field of the metal itself passing through the current, while the proximity effect is caused by the adjacent metal flowing through the current. Adjacent metals through which no current flows create eddy currents induced by electromagnetic fields. The generation of eddy current changes the current distribution and current density of the coil adjacent to the inductor, and increases the equivalent series resistance of the inductor.

When the operating frequency reaches several GHz, the eddy current loss caused by the inductive coupling between the coils will be very serious, as shown in Figure 2.

In Figure 2, three metal wires from outside to inside are shown from top to bottom. I _coil is the alternating current flowing through the inductance coil. According to Faraday's law, this current will generate an alternating magnetic field B _coil , and at the center of the coil The alternating magnetic field B _coil reaches the maximum value, as shown in the vector distribution diagram of the inductive magnetic field in FIG. 3 . The closer this magnetic field is to the center of the coil, the denser it becomes. According to Faraday's law of electromagnetic induction, the alternating magnetic field B _coil will generate an eddy current I _eddy inside the coil, and the alternating current will generate an alternating magnetic field _Beddy . According to Lenz's law, in order to hinder the change of the alternating magnetic field B _coil , the direction of the alternating magnetic field _Beddy will be opposite to that of the alternating magnetic field B _coin . Moreover, on the inner side of the coil, I _eddy is in the same direction as I _coil , and the total current density becomes larger; on the outer side of the coil, I _eddy is in the opposite direction to I _coil , and the total current density becomes smaller, which will lead to The current density is more uneven. And the closer the metal wire to the center of the coil, the more uneven the current density, and the higher the frequency, the more serious this phenomenon will be.

The thicker the coil, the greater the induced current I _eddy and the more energy it consumes. For a wire in a time-varying magnetic field, the power dissipated by eddy currents is

$P_e=\frac{{\left(\mathrm{\πtfB}\right)}^2}{6\mathrm{\ρ}}$ Formula (2)

In the formula: t is the width of the conductor, f is the working frequency, B is the magnetic field where the conductor is located, ρ is the resistivity of the conductor, Pe is the power consumed by the conductor per unit volume, multiply Pe by the total volume of the conductor at the same magnetic field, Then summing the whole coil, the total energy Pa consumed by the eddy current in the coil can be obtained. And the eddy current loss is proportional to the square of the operating frequency.

The magnetic field where the nth turn of the wire of the inductive metal coil is located has 3 components:

B _n = B _{n, n} + B _{n, in} + B _{n, out} formula (3)

In the formula: B _n is the total magnetic field where the nth turn of wire is located, B _{n, n} is the magnetic field generated by the nth turn of wire itself here, B _{n, in} is the position of each turn coil within the nth turn of wire The generated magnetic field, B _{n, out} is the magnetic field generated here by the coils other than the nth turn of wire. They are calculated as follows

$B_{\mathrm{no},\mathrm{no}}=\frac{{\mathrm{\μ}}_{0}I}{4{\mathrm{\πd}}_{\mathrm{no}}}(\frac{\sqrt{5}}{2}+\frac{\sqrt{2}}{4})$ Formula (4a)

${\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; in</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}\mathrm{<span\; class="notranslate">\; I</span>}}{\mathrm{<span\; class="notranslate">\; 4</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span><span\; class="notranslate">\; -</span>\frac{\sqrt{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\sqrt{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\sqrt{\mathrm{<span\; class="notranslate">\; 2</span>}}\frac{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

$-\frac{\sqrt{{(d_{\mathrm{no}}-d_k)}^2+{(d_{\mathrm{no}}+d_k)}^2}}{d_{\mathrm{no}}^2-d_k^2})$ Formula (4b)

${\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; out</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}\mathrm{<span\; class="notranslate">\; I</span>}}{\mathrm{<span\; class="notranslate">\; 4</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; (</span>\frac{\sqrt{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\frac{\sqrt{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; +</span>\sqrt{\mathrm{<span\; class="notranslate">\; 2</span>}}\frac{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}}{{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

$+\frac{\sqrt{{(d_k-d_{\mathrm{no}})}^2+{(d_k+d_{\mathrm{no}})}^2}}{d_k^2-d_{\mathrm{no}}^2})$ Formula (4c)

Among them, μ ₀ is the vacuum magnetic inductance coefficient, I is the current in the coil, i.e. I _coil , dn=ln/8, ln is the length of the nth coil, as shown in the following formula:

$l_{\mathrm{no}}=l_{\mathrm{no}-1}+{\mathrm{the\; s}}_{\mathrm{no}-1}+{6\mathrm{the\; s}}_{\mathrm{no}}+{\mathrm{the\; s}}_{\mathrm{no}+1}+\frac{{7w}_{\mathrm{no}-1}+{9w}_{\mathrm{no}}}{2}$ Formula (5)

On the other hand, from the analysis of the eddy current effect, it can be seen that the magnetic field between two adjacent metal conductors will change the current distribution of the metal conductors. When the distance between two adjacent metal conductors is smaller, the magnetic field interaction between them is stronger. It will cause the current density in the metal conductor to be more uneven, which will further increase the series resistance Rs of the metal conductor, thereby affecting the decrease of the quality factor Q value of the inductance; in addition, when other parameters of the inductance remain unchanged, if the two metal conductors The larger the spacing, the larger the total length of the metal conductor, and the smaller the inner diameter, which will also increase the series resistance Rs of the metal conductor, and the quality factor Q value of the inductance will decrease.

From the above analysis, it can be seen that the closer the inductance helical coil is to the center of the coil, the greater the magnetic induction loss it bears, that is, the eddy current loss of the innermost inductance coil is the largest. And the higher the operating frequency of the inductor, the more serious the proximity effect. In this regard, the distance between the metal conductors can be increased by adding a metal layer to reduce the proximity effect so as to reduce the loss. At the same time, in order to avoid the structure being too complicated and not conducive to tape-out, the utility model is that the innermost circle of conductors and other conductors Not on a plane, this is because the innermost magnetic field of the coil is the strongest, and this can achieve good results.

Please refer to Fig. 4. In this utility model, the metal wire coils of the spiral inductor are not on the same plane, and its structural parameters are respectively the number of metal coil coils N, the outer diameter D _out , the inner diameter D _in , the metal wire width w _n , and the two metal coils. Inter-spacing s _n . In order not to make the structure too complicated, it is considered that the inductor is composed of metal layers M1, M2, and M3. The metal layers M1 and M2 are respectively connected to the signal input and output ends, and the metal layer M3 is the innermost coil of the inductor, which is not on the same plane as the other coils. Two-tier structure.

Please refer to Figure 5A and Figure 5B, comparing the cross-sectional view of the traditional planar spiral inductor and the non-planar spiral inductor of the utility model, the coil of the inner coil of the non-planar spiral inductor coil of the utility model is not on the same plane as the other coils, two layers There is a vertical distance Δh between them, and coils in different planes are implemented on dielectric layers of different heights.

The implementation process of the utility model is as follows: according to the required inductance and outer diameter D _out of the design, use software simulation to determine the number of metal coil coils N of the inductance, the metal conductor line width w _n , and the spacing between metal conductors s _n , and The vertical distance Δh between the two layers completes the manufacture of the inductor when manufacturing the corresponding metal layer.

Claims (2)

1, planar spiral inductor such as non-, it is characterized in that: at least one circle metal spiral coil in the described spiral inductance and all the other arbitrary metal spiral coils are in Different Plane, and the coil that is in Different Plane is realized on the dielectric layer of differing heights.

2, planar spiral inductor such as non-as claimed in claim 1, it is characterized in that: coil and all the other metal spiral coils of the inner ring of spiral inductance are in Different Plane.

Images