Abstract

One of the largest components on miniaturized, multichip modules is the chip inductor associated with the onboard DC to DC power converter. It is an essential component for efficient conversion of the system voltage to the operating voltage of the module. In a previous paper, we described how vibrating capacitor structures could be used to perform inductor functions. In this paper, we focus on the specific application of an on-board power converter. Our vision is for a MEMS scale device that could be attached to the back of a chip as an appliqué or embedded in an interposer on which the chip was mounted. It would function as a DC to DC converter to extract power from the system buss and supply it to the chip at its required operating voltage. This device consists of a long, flat, rectangular, vibrating beam that is sandwiched between two fixed electrodes. It is clamped at its two ends and electrically connected to ground. On one of the fixed electrodes, which we refer to as the source electrode, a bias potential is applied through a resistor connected to a battery. The motion of the beam relative to the source electrode generates a sinusoidal like fluctuation in the potential between them. The amplitude of this alternating voltage is set by the amplitude of the beam vibration and the nominal gap between the beam and source electrode. It can be made insensitive to variations in battery voltage, by controlling the amplitude of the beam vibrations. The sinusoidal portion of the bias voltage can be extracted through a DC blocking capacitor connected to the source electrode. By passing it through a rectifier and filter capacitor, it can be used as a stable DC supply voltage. The beam vibration is initiated and sustained by voltage pulses applied to the other fixed electrode, which we refer to as the drive electrode. A logic circuit connected to the source electrode monitors its sinusoidal voltage and applies a pulse to the drive electrode when two criteria are met. First, the sinusoidal voltage must be rising, which occurs when the beam is moving away from the source electrode and towards the drive electrode. Second, the time average of the sinusoidal voltage must be below its desired set point, which means that more energy must be injected into the beam. When voltage is applied to the drive electrode, it exerts a force on the vibrating beam that pulls it towards the drive electrode. We have constructed a Mat Lab model of a vibrating beam power converter and have been using it to examine a number of design factors including, physical size, vibration frequency and amplitude, values for the DC blocking and output filter capacitors, as well as the logic used to apply pulses to the drive electrode. Our intent is to develop a sufficient understanding of the device operational and physical design requirements to enable us to build a prototype device.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call