Electrical circuits used for radio communication often comprise semiconductor components such as transistors or PIN diodes. However at microwave frequencies the performance of such transistors and pin diodes may be less than satisfactory.

For example PIN diodes can suffer from relatively high insertion losses, and transistors can suffer from relatively low isolation values. Additionally the insertion losses and isolation values of these prior art devices varies depending on the frequency of the signal passing through the switches.

In an attempt to solve these and other problems micro-electromechanical switches have been developed, and these often have relatively low insertion loss and a high isolation value.

High isolation values for MEMS switches are achieved by having a large switch gap between the contact electrodes of the switch. Typically the gap between the contact electrodes is 1 to 10 microns. In electrostatically actuated devices this large switch gap results in the need for relatively high voltages to actuate the switch.

Typical actuation voltages for prior art MEMS switches are of the order of 50 volts.

For applications in which relatively low power batteries are typically used, such as in mobile telephony, the need to reduce actuation voltages, whilst maintaining low power consumption, is marked.

European Patent Application EP1146532A2 describes a micro-mechanical element, such as a switch, in which first and second control signals are applied to the element to reduce the voltage levels required. The first signal holds the micro- mechanical element in an active position and the second signal sets the element to the active position using either mechanical or electrical resonance to reduce the voltage level required.

It is an objective of the present invention to provide an improved micro- electromechanical switch.

According to a first aspect, the invention provides a switch device comprising a first and second contact electrode, and a first and second actuation electrode, the first actuation electrode and first contact electrode being mounted on at least part of, or forming at least part of, a resiliently flexible arm, the switch further comprising an electrical actuation means for, when in use, applying a first inter-actuation electrode potential difference between the first actuation electrode and the second actuation electrode ; the electrical actuation means further comprising a resonance means for, when in use, varying the first inter-actuation electrode potential difference in such a manner that mechanical resonance is caused to occur in the resiliently flexible arm characterised in that the resonance means comprises an electrical feedback circuit.

The switch device conveniently comprises a substrate and the resiliently flexible arm, first actuation electrode, second actuation electrode, first contact electrode and second contact electrode are arranged in such a manner that, when the first inter-actuation electrode potential difference is zero, and when the arm, is in a state of equilibrium, at least part of the arm is spatially separate from a substrate, i. e. with no signal applied the arm is spaced away from the substrate and the switch is open-the first and second contact electrodes are not touching. The second actuation electrode and the second contact electrode may be mounted on, or form at least part of, the substrate. The construction of the switch device may be such that the position of at least part of the arm is fixed relative to that of the substrate.

The resiliently flexible arm and substrate may be arranged in such a manner that at least part of the arm may be made to contact the substrate by deformation of the arm. Contact between the arm and the substrate, resulting from deformation of the

arm, results in contact between the first and second contact electrodes, thus closing the switch.

The resonance means causes the resiliently flexible arm to resonate, i. e. to oscillate at a certain frequency. The oscillation causes a variation in the distance between the first and second actuation electrodes. As the arm oscillates the minimum distance between the first and second actuation electrodes decreases as the amplitude of oscillation increases.

The electrical actuation means comprises a latch means for, when in use, applying the latch potential difference between the first and second actuation electrodes.

As mentioned resonance of the arm causes the distance between the first and second actuation electrodes to fluctuate. Enough fluctuation may result in the proximity of the first and second actuation electrodes being such that the latch potential is sufficient to hold at least part of the arm, that having the first contact electrode disposed thereon, against the substrate.

The latch potential may therefore be greater than or equal to the minimum potential difference required to hold at least part of the arm in contact with the substrate.

The potentials applied to the first and second actuation electrodes to cause this resonance and contact are much smaller in magnitude than would have to be applied to cause contact in the absence of resonance.

The resonance means comprise an electrical feedback circuit. A positive electrical feedback circuit is a simple and effective way of achieving resonance with a quickly increasing amplitude of oscillation until the latched position is achieved. Use of a positive feedback circuit removes the need for a complex driving circuit allowing simpler and less expensive switches to be produced.

Preferably the resonance means comprises a gain means for, when in use, superimposing a resonance potential difference, which is proportional to the rate of change of capacitance between the first and second actuation electrodes, to the latch potential difference in such a manner that resonance in the resiliently flexible arm occurs.

In other words the inter-electrode potential difference for the first and second actuation electrodes may be equal to the sum of the latch potential difference and the resonance potential difference.

Advantageously the electrical actuation means has a construction such that the magnitude of the potential difference between the first and second actuation electrodes is always less than 100 volts. More advantageously the electrical actuation means has a construction such that the magnitude of the potential difference between the first and second actuation electrodes is always less than 50 volts. Yet more advantageously the electrical actuation means has a construction such that the magnitude of the potential difference between the first and second actuation electrodes is always less than 10 volts.

Preferably the first and second actuation electrodes are arranged in such a manner that gap between the first and second actuation electrodes, when the first inter- actuation electrode potential difference is zero, and when the resiliently flexible arm is in a state of equilibrium, is between 0.1 and 100 microns.

Preferably the first and second contact electrodes are arranged in such a manner that gap between the first and second contact electrodes, when the first inter- actuation electrode potential difference is zero, and when the resiliently flexible arm is in a state of equilibrium, is between 0.1 and 100 microns.

Preferably the length of the resiliently flexible arm may be between 10 and 1000 microns. More preferably the length of the resiliently flexible arm may be between 20 and 500 microns.

Advantageously the smallest cross-sectional dimension of the resiliently flexible arm is between 0. 1 and 50 microns. More advantageously the cross-sectional area of the of the resiliently flexible arm is between 0.1 and 20 microns.

The resiliently flexible arm may comprise an insulator or a semiconductor. The resiliently flexible arm may comprise polycrystalline silicon or silicon nitride.

The switch device may form part of a radio frequency circuit. The switch device may comprise part of an optical communications system The value of Q for the fundamental mode of the resiliently flexible arm, when mounted on the substrate and prior to contact between the first and second contacts, is preferably greater than 10. The value of Q for the fundamental mode of the resiliently flexible arm, when mounted on the substrate and prior to contact between the first and second contacts, is more preferably greater than 100.

The switch device may further comprise a gain stop means, the gain stop means comprising an electrical circuit and having a construction such that, when the at least part of the arm contacts the substrate, as a result of resonance in the arm, the gain variable potential difference is turned off.

The invention will now be described, by way of example only, with reference to the following diagrams : Figure 1 shows various views of a mechanical switch which forms part of a switch device according to the invention; Figure 1 a shows a side view, Figure 1 b a perspective view and Figure 1 c a plan view, Figure 2 shows a schematic block diagram of switch device with feedback control means according to the invention, Figure 3 shows an example circuit which may be used to control the switch, Figure 4 shows the applied voltage and tip deflection as simulated for a switch according to the present invention.

Figure 1 shows several views of a mechanical switch, generally indicated by 10, according to an embodiment of the invention. Figure 1 a shows a side view, Figure 1 b a perspective view and Figure 1 c a plan view. In all views like numerals are used to denote like components. The mechanical switch 10 comprises a first actuation electrode 11, and a second actuation electrode 12. The first actuation electrode 11 is mounted on the resiliently flexible arm 13. Note, as shown, the arm 13 need not be a continuous structure but can have gaps or channels to aid the mechanical properties. The second actuation electrode 12 is mounted on a substrate 14. The mechanical switch further comprises a first contact electrode 16 and a second contact electrode 17. The first contact electrode 16 is mounted on the tip of the resiliently flexible arm and the second contact electrode 17 mounted on the substrate at an appropriate point so that deformation of the arm can bring the contact electrodes 16,17 into contact.

Supply of an inter-electrode potential difference to the first and second actuation electrodes will causes deformation of the resilient arm 13. The applied voltage is

varied in such a way to cause mechanical resonance of the resilient arm 13. Thus the deformation of the arm 13 will increase until the deflection is such that the first and second actuation electrodes 11,12 are close enough that they can be held in proximity by a constant latch voltage. The switch is designed such that when the first and second actuation electrodes 11, 12 are held in proximity by the latch voltage the first and second contact electrodes 16,17 are in contact. In this state the switch is closed and electrical signals can pass through the switch via the first and second contact electrodes 16,17. When the first and second actuation electrodes are held in proximity they may or may not actually be in contact.

The constant voltage applied is sufficient to hold the first and second actuation electrodes in proximity or in other words hold the first and second contact electrodes 16,17 in contact. However it would be insufficient to bring the actuation electrodes into proximity from the open position. Therefore application of a varying voltage to cause mechanical resonance reduces the voltage levels required to close the switch. Mechanical resonance is achieved using a positive feedback circuit.

Figure 2 shows a schematic block diagram switch device according to the invention. The switch device comprises four components: a latch means 21, a capacitance monitoring means 22, a gain means 23, and a mechanical switch 10.

The mechanical switch 10 is identical to that shown in figure 1. The latch means 21 may apply a constant latch potential difference between the first and second actuation electrodes 11 and 12 shown in figure 1. This latch potential difference is greater or equal to the minimum potential difference required to keep the first and second contact electrodes 16,17 together once they have been bought into contact and are in a state of equilibrium. The capacitance monitoring means 22 outputs a voltage proportional to the rate of change of capacitance between the first and second actuation electrodes 11, 12. The gain means 23 increases the size of the signal from the differential of the capacitance monitoring means. The signal from the gain means 23 is added to the constant latch potential. The resonance variable potential difference between the first and second actuation electrodes 11,12

causes the resiliently flexible arm 13 to oscillate, and further causes the capacitance of the mechanical switch 10 to vary. The system thus describes uses the variation in capacitance due to movement of the resilient arm to vary the applied voltage. The system therefore applies positive feedback to achieve mechanical resonance.

Figure 3 shows an example circuit which may be used to control the switch according to the invention. This circuit comprises an OP-Amp 31, a mechanical switch 10, a fixed capacitor 32 and a fixed resistor 33.

Figure 4 shows results of a dynamic simulation of a switch according to the invention. It can be seen that the applied voltage quickly builds from a constant level to a varying voltage at the resonant frequency of the device. The tip deflection also starts to build until after less than 0.25ms the first and second contact electrodes are in contact. At this time the arm can be held in position, the switch held closed merely by application of the constant voltage.