Author Archives: Andi

Underwater, Battery-less, Wireless Camera

At the Massachusetts Institute of Technology, engineers have built a wireless camera that does not require a battery to operate underwater. The necessity arose when scientists wanted to observe life under the oceans. They realized they knew less about earth’s oceans than the surface of Mars or the far side of the moon.

An underwater camera must remain tethered to a research vessel for receiving power or sent to a ship periodically to recharge its batteries. This limitation is a big challenge, preventing easy and widespread explorations underwater.

MIT engineers took up the challenge of overcoming the problem. They came up with a camera that does not require batteries, works underwater, and transmits wirelessly. Compared to other underwater cameras, the new camera is more than 1000,000 times more energy-efficient. The device even takes color photos, transmitting them wirelessly through the water.

Sound powers the new autonomous camera—converting the mechanical energy of sound waves into electrical energy for powering its imaging and communications circuitry. After capturing the image, the camera encodes the data and uses underwater waves to transmit it to a receiver for reconstructing the image.

As the camera does not need a rechargeable power source, it can run for a long time before retrieval. This enables scientists to explore remote areas under the ocean. The camera is helpful in capturing images of ocean pollution and monitoring the health and growth of fish.

For a camera that can operate autonomously underwater for long periods, engineers required a mechanism for harvesting underwater energy by itself while using up very little power internally.

The camera uses transducers made of piezoelectric materials that the engineers placed around its exterior. The transducers produce an electrical signal when sound waves hit them. The sound waves may come from any source, such as from marine life or a passing ship. The camera then stores the energy it has harvested, until it has enough for powering its electronics.

The camera has ultra-low-power imaging sensors to keep its power consumption at the lowest possible levels—but these sensors capture only gray-scale images. Moreover, underwater environments are mostly dark, so the camera also needs a low-power flash.

MIT engineers solved both problems simultaneously by using three LEDs of red, blue, and green colors. For capturing an image, the camera first uses the red LED, then repeats the process with a blue LED, and finally with the green LED.

Although each image is black and white, the white part of each photo has the reflection of its respective colored light. Combining the image data during post-processing reconstructs the color image.

The engineers use an underwater backscatter process to transmit the captured image data after encoding it as bits. A nearby receiver transmits sound waves through the water to the camera, reflecting it back just as a mirror would. The camera can choose to either reflect the sound back to the receiver or act as an absorber and not reflect it.

The transmitter has a hydrophone next to it. If it senses a reflected signal from the camera, it treats it as a bit 1. If there is no signal, then it is a bit 0. The receiver uses this binary information for communicating with the camera.

What are Radar Sensors?

Autonomous driving requires the car to have radar sensors as its ears. Originally, the military and avionics developed radar for their applications. Automobiles typically use millimeter wave radar, with a working frequency range of 30-300 GHz, and wavelengths nearer to centimeter waves. These millimeter wave radar offer advantages of photoelectric and microwave guidance to automobiles, because of their significant penetration power.

Automobile collision avoidance mainly uses 24 GHz and 77 GHz radar sensors. In comparison with centimeter wave radar, millimeter-wave radar offers a smaller size, higher spatial resolution, and easier integration. Compared to optical sensors, infrared, and lasers, millimeter wave radar has a significantly stronger ability to penetrate smoke, fog, and dust, along with a good anti-interference ability. Although the millimeter band radar is essential for autonomous driving, heavy rain can significantly reduce the performance of radar sensors, as it produces a large interference. 

Automobiles first used radar sensors in a research project about 40 years ago. Commercial vehicle projects started using radar sensors only in 1998. Initially, they were useful only for adaptive cruise control. Later, radar sensors have developed to provide collision warnings also.

Radar sensors are available in diverse types, and they have a wide range of applications. Automobile applications typically use them as FMCW or frequency-modulated continuous wave radars. FMCW radars measure the air travel time and frequency difference between the transmitted and received signals to provide indirect ranging.

The FMCW radar transmits a frequency-modulated continuous wave. The frequency of this wave changes with time, depending on another triangular wave. After reflection from the object, the echo received by the radar has the same nature of frequency as the emitted wave. However, there is a time difference, and this tiny time difference represents the target distance.

Another radar in common use is the CW Doppler radar sensor. These sensors use the principle of the Doppler effect for measuring the speed of targets at various distances. The radar transmits a microwave signal towards the target, analyzing the frequency change of the reflected signal. The difference between the two frequencies accurately represents the target’s speed relative to the vehicle.

Autonomous vehicles use radar sensors as their basic but critical technical accessories. The radar sensor helps the vehicle to sense objects surrounding it, such as other vehicles, trees, or pedestrians, and determine their relative positions. Then the car can use other sensors to take corresponding measures. Radar sensors provide warnings like front vehicle collisions and the initial adaptive cruise. Vehicles with autopilot radars require more advanced radar sensors such as LIDARs that offer significantly faster response speeds.

Autonomous vehicles must develop technologically. Autonomous driving basically requires an autonomous vehicle to quickly understand and perceive its surrounding environment. This requires the coordination of various sensors, allowing the car to see six directions and hear all. Reliable and decisive driving by an autonomous vehicle requires timely and accurate sensing of roads, other vehicles, pedestrians, and other objects around the vehicle.

Automotive electronics mainly uses radar sensors to avoid forward collisions, sideways collisions, backward collisions, automatic cruises, automatic start and stop, blind spot monitoring, pedestrian detection, and automatic driving of vehicles.

What are Multilayer Chip Capacitors?

The electronics industry uses various types of capacitors in its circuits. These capacitors provide different capabilities and functionality depending on the type and construction. One of the most prevalent types of capacitors is the MLCC or multilayer ceramic capacitor.

Most MLCCs are applicable to circuits that require small-value capacitance. They are preferably useful as filters, in op-amp circuits, and bypass capacitors. This is because MLCC offers small parasitic inductance as compared to aluminum electrolytic capacitors. Therefore, MLCC offers better stability over temperature, subject to their temperature coefficient.

MLCC is available in three categories or classes. The Class I type of ceramic capacitors offer low losses and high stability in resonant circuits. Although they do not require aging corrections, their volumetric efficiency is low. Class II and Class III offer high volumetric efficiency, but their stability is not as good as that of Class I capacitors. Once outside the referee time of the manufacturer, Class II and Class III capacitors may require aging corrections. Manufacturers specify the referee time during which the capacitor will remain within the tolerance range.

Alternating layers of dielectric ceramic and metallic electrodes make up an MLCC. This structure makes them physically small but does not provide them with volumetric efficiency. Design engineers selecting MLCC for electronic applications look for two important parameters—voltage rating and temperature coefficient.

The voltage rating of the MLCC indicates the maximum safe voltage the circuit can apply across the capacitor terminals. For enhanced reliability, designers use a capacitor with a voltage rating higher than it will experience in the circuit. One advantage over electrolytic capacitors is that MLCCs are non-polarized. Therefore, it is possible to connect MLCC in any position without damage.

The temperature coefficient of an MLCC depends on its Class category. If the capacitor contains Class I ceramic material, it will have a very low-temperature coefficient, which means, a change in temperature will minimally affect the capacitance. Class I MLCC also tends to have low dielectric constants, which means the material offers very small capacitance per volume. For instance, C0G and NP0 type Class I MLCC feature a 0 temperature coefficient with a tolerance of ±30 ppm.

Class II MLCC, although less stable over temperature, contains ceramic material with a higher dielectric constant. That means Class II MLCC can have more capacitance in the same volume compared to that of Class I. Class II MLCC are available in X, Y, and Z temperature coefficients. For instance, X7R is a common Class II MLCC, and can operate within a temperature range of -55 °C and +125 °C with a tolerance of ±15%. X5R MLCC can operate within a temperature range of -55 °C and +85 °C with a tolerance of ±15%. Y5V MLCC can operate within a temperature range of -30 °C and +85 °C with a tolerance of +22/-82%. MLCCs with wider temperature ranges are also available with the higher stability of temperature characteristics. However, these capacitors tend to cost more.

Engineers use several capacitors with various values in parallel or series for providing the requisite impedance over a wide range of frequencies.

Three-Phase Monitor Relays Protect Expensive Machinery

Three-phase motors power many industrial and commercial machines. One can find these machines in material handling, water treatment, air conditioning systems, ventilation, heating, marine, machine tools, and aviation applications. However, a range of fault conditions can damage these reliable devices when not addressed quickly. This can lead to a shortened operating lifetime or even a failure, resulting in significant repair costs and downtime.

Phase monitoring relays can detect these faults, notify the operators, and stop the machinery before it develops permanent damage. These relays detect the presence of all three phases, their correct sequence, and that all phase voltages are within the specified range. Should an error develop, the relay opens a set of contacts, initiating an alarm condition, and powers down the machine. There are many types of phase-sensing relays. They can handle a wide range of phase configurations, voltages, and errors.

Among the common failure modes of three-phase motors, are those related to their three-phase power source and their effects on the motor. An imbalance in the phase voltages, or a loss in one of the three phases, can result in the remaining phases driving higher-than-normal currents into the motor. This can lead to a loss of rotational power and excessive vibrations. Likewise, over-voltages and under-voltages can force the motor to draw excess current for driving the same load, and this can shorten the life of the motor. An incorrect phase sequence may cause the motor to reverse the direction of rotation. This can have significantly disastrous results on the load connected to the motor.

Phase monitoring relays monitor the state of the three-phase power source. The three-phase lines that they monitor also power them. Apart from the phase sequence, they also monitor the loss of any phase voltage. Only when all the phases are present, and are in the correct sequence, do the relays activate. Whenever there is a loss of any phase, or the phase sequence is incorrect, the relays de-energize.

Some phase monitoring relays also have the capability to monitor the voltage levels of all three phases. This typically uses a true root-mean-square measurement. The relay deactivates whenever the voltage drops below a preset threshold. Some relays also offer adjustable limit settings along with voltage detection. Other relays monitor phase asymmetry along with tolerance. Typically, phase monitoring relays offer a delay before actuation. This prevents spurious activation from temporary voltage levels or asymmetry issues. In some models, the delay is adjustable.

The DPA01CM44 is an example of a three-phase monitoring relay meant for three-wire configurations. The three-phase source powers the relay. Relay models available operate at voltages of 208, 230, 400, 600, and 690 VAC. Although relays for mounting on DIN rails are typical, plugin models are also available. The relay output configuration can be single or dual SPDT contacts.

Normal voltage and phase conditions allow the relay to remain activated. That means, the normally open contacts of the relay output remain closed. Abnormal conditions make the relay operate within 100 milliseconds. The front panel on the relay has status LEDs to indicate relay activation and power on.

What is Capacitance to Digital Converter Technology?

The healthcare industry has witnessed many advancements, innovations, and improvements in electronic technology in recent years. Healthcare equipment faced challenges like developing new treatment methods and diagnoses, home healthcare, remote monitoring, enhancing flexibility, improving quality and reliability, and improving ease of use.

A comprehensive portfolio of these technologies includes digital signal processing, MEMS, mixed-signal, and linear technologies that have helped to make a difference in healthcare instrumentation in areas such as patient monitoring and imaging. Another is the capacitance to digital converter technology that offers the use of highly sensitive capacitance sensing in healthcare applications. For instance, a capacitive touch sensor is a novel user input method that can be in the form of a slider bar, a push button, a scroll wheel, or other similar forms.

In a typical touch sensor layout, a printed circuit board may have a geometric area representing a sensor electrode. This area forms one plate of a virtual capacitor, while the user’s finger forms the other plate. For this system to work, the user must essentially be grounded with respect to the sensor electrode.

Analog Devices has designed their CapTouch controller family of ICs, the AD7147/ AD7148, to activate and interface with capacitance touch sensors. The controller ICs measure capacitance changes from single-electrode sensors by generating excitation signals to charge the plate of the capacitor. When another object, like the user’s finger, approaches the sensor, it creates a virtual capacitance, with the user acting as the second plate of the capacitor. A CDC or capacitance to digital converter in the ICs measures the change in capacitance.

The CDC can measure changes in the capacitance of the external sensors and uses this information to activate a sensor. The AD7147 has 13 capacitance sensor inputs, while the AD7148 has eight. Both have on-chip calibration logic for compensating for measurement changes due to temperature and humidity variations in the ambient environment, thereby ensuring no false alarms from such changes.

Both CDCs offer many operational modes, very flexible control features, and user-programmable conversion sequences. With these features, the CDCs are highly suitable for touch sensors of high resolution, acting as scroll wheels or slider bars, requiring minimum software support. Likewise, no software support is necessary for implementing button-sensor applications with on-chip digital logic.

The CDCs function by applying an excitation signal to one plate of the virtual capacitor, while measuring the charge stored in it. They also make the digital result available to the external host. The CDCs can differentiate four types of capacitance sensors by changing the way they apply the excitation.

By varying the values of these parameters, and/or observing the variations in their values, the CDC technology directly measures the capacitance values. The distance between the two electrodes affects the output of the CDCs in inverse proportions.

The family of Analog Device CDCs, the AD714x, AD715x, and AD774x, are suitable for applications involving a wide range of functions. These involve various input sensor types, input ranges, resolutions, and sample rates. Applications involve liquid level monitoring, sweat detection, respiratory rate measurement, blood pressure measurement, and more.

Space Saving Molex Connectors

With manufacturing processes and semiconductor materials going through new developments at break-neck speeds, we now have a proliferation of increasingly smaller sensors, devices, and processors. However, some areas are still facing hindrances in technological advancements because of space limitations, thereby slowing down user adoption.

One such area is the AR/VR or augmented- and virtual reality applications. These technologies, typically AR, superimpose an image over a view of the user’s actual environment. A handheld device can accomplish this, such as a smartphone. Others can be user-worn glasses, headsets, or a projection such as heads-up displays in vehicles.

AR technology commonly includes offering information about the environment around the user, for gaming or for safety reasons. On the other hand, VR technology immerses the user in a virtual environment. That means, VR implementation typically requires the use of a headset, completely covering the user’s eyes, thereby blocking out the world around them.

However, the adoption of AR and VR has so far been limited, and these have remained relatively niche markets. The primary reason for this is their footprint. For instance, AR use requires wearing bulky glasses, lenses, or headwear, or, holding the smartphone up to view the AR environment. Wearing such heavy, unbecoming devices for any duration can be very uncomfortable.

For engineers, the size of connectors has been one of the biggest challenges when they try to limit the size of devices for embedded and wearable systems. Although semiconductor sizes have progressively reduced, communication devices have stayed the same. Therefore, even with custom cabling, the cable size and its corresponding connector are the factors limiting the system size.

For the success of AR and VR solutions, it is necessary for their form factor to be small, comfortable, and lightweight for the user. These technologies also demand significant processing power as well as high-quality displays. Meeting this demand requires design engineers to use connectors that offer not only robust communication capabilities, but also minimize the weight and footprint.

Molex is now offering a quad-row connector that meets the above needs. The package is significantly smaller than those available in the market while offering many connectivity options.

The quad-row connector from Molex offers its performance gains because of its staggered-circuit layout that offers a 30% space-saving over the design of its competitors. The quad-row connector achieves this as it positions its pins across four rows with a pitch of 0.175 mm. Such a staggered-circuit layout is a substantial space-saver in many applications involving wearable, smartphones, smartwatches, and AR and VR devices.

According to Molex, users can also have a soldering pitch of 0.35 mm in the quad-row connectors. This matches with the standard surface-mount technology processes. That means that as electronic devices gain popularity and size reduction, manufacturers can scale their products by shifting to the 0.175 mm soldering pitch. These connectors from Molex can also integrate into moving objects, and withstand drops, vibrations, and other harsh conditions of use. Molex builds its quad-row connectors with interior armor and insert-molded power nails, making them substantially reliable and robust. The connectors are available in 32- and 36-pin varieties, with 64-pin configurations for the future.

Magnetic Position Sensing in Robots

Robots often operate both autonomously and alongside humans. They greatly benefit the industrial and manufacturing sectors with their accuracy, efficiency, and convenience. By monitoring motor positions at all times, it is possible to maintain not only system control but also prevent unintentional motion, as this can cause system damage or bodily harm.

Such monitoring of motor positioning is possible to implement by contactless angle encoding. It requires a magnet mounted on the motor shaft and provides an input for a magnetic encoder. As dirt and grime do not influence the magnetic field, integrating such an arrangement onto the motor provides a compact solution. As the encoder tracking the rotating magnet provides sinusoidal and 90-degree out-of-phase components, their relationships offer quick calculations of the angular position.

As the magnet rotates on the motor shaft, many magnetic encoding technologies can offer the same end effect. For instance, Hall-effect and magnetoresistance sensors can detect the changing magnetic field. 3D linear Hall effect sensors can help with calculating angular positions, while at the same time, also offering compensations for temperature drift, device sensitivity, offset, and unbalanced input magnitudes.

Apart from signal-chain errors, the rotation of the magnet also depends on mechanical tolerances. This also determines the quality of detection of the magnetic field. A final calibration process is necessary to achieve optimal performance, which means either harmonic approximation or multipoint linearization. With calibration against mechanical error sources, it is possible for magnetic encoding to achieve high accuracy.

The driving motor may connect directly to the load, through a gearbox for increasing the applied torque, through a rack and pinion, or use a belt and screw drive for transferring energy elsewhere. As the motor shaft spins, it transfers the kinetic energy to change the mechanical position somewhere in the system. In each case, the angle of the motor shaft correlates directly to the position of the moving parts of the system. When the turns ratio is different from one, it is also necessary to track the motor rotations.

Sensorless motor controls and stepper motors do not offer feedback for the absolute position. Rather, they offer an estimate of the position on the basis of the relative change from the starting position. When there is a loss of power, it is necessary to determine the actual motor position through alternate means.

Although it is possible to obtain the highest positional accuracy through the use of optical encoders, these often require bulky enclosures for protecting the aperture and sensor from contaminants like dirt and dust. Also, it is necessary to couple the mechanical elements to the motor shaft. If the rotational speed exceeds the mechanical rating of the encoder, it can lead to irreparable damage.

No mechanical coupling is necessary in the case of magnetically sensed technologies like magnetoresistive and Hall-effect sensors, as they use a magnet mounted on the motor shaft. The permanent magnet has a magnetic field that permeates the surrounding area, allowing a wide range of freedom for placing the sensor.

RF MEMS Switches for 5G Networks

For high-power RF designs like 5G networks, Menlo Micro has added an RF MEMS switch that contains an integrated driver circuit for a charge pump. The RF MEMS switch operates from DC to 6 GHz.

The new RF switch from Menlo Micro is one of a family of SP4T or single-pole/four-throw, DC-t0-6 GHz switch, and is meant for 5G infrastructure, measurement, and testing equipment involving high-power RF switching applications. Menlo Micro is using its own Ideal Switch technology for the high integration MM5140 SP4T switch. The technology gives the new switch power handling capability up to 25 W, an ultra-low insertion loss, and the highest linearity in the industry. The MM5140 SP4T switch easily outperforms all types of traditional solid-state switches and electromechanical relays.

The MM5140 SP4T switch performs RF operations at high power levels over a wide temperature range of -40 °C to +85 °C, delivering superb linearity from DC to 6 GHz. 5G RF applications demand significant reductions in distortion, which the switch’s IP3 of 95 dBm provides conveniently.

Menlo Micro has custom designed a built-in high-voltage charge pump or driver circuit and integrated it into the LGA package of the MM5140 SP4T switch. The charge pump circuit has both GPIO and SPI digital interfaces so that any test system or host processor can keep control over it.

Although the new module has the existing MM5130 at heart, it also has the CMOS charge pump driver ASIC, driver circuitry, and other peripheral passive components in its 5.2 X 4.2 mm package.

As the MM5140 SP4T switch is a single-pole four-throw device, the voltage must route over to each of the four gate lines. This requires the presence of either a MOSFET drive circuit or a dedicated multiplexer IC. Along with the integrated charge pump and the driver circuitry, the MM5140 SP4T switch saves board area and bill of materials.

The integrated passive components include a large capacitor that the charge pump requires for handling the high voltage that drives the MEMS. This helps reduce the BOM for passive components.

The difference between the MM5130 and the MM5140 is their operational speed. The MM5130 is a design meant to operate at higher frequencies, such as the microwave band. The design of the MM5140 is meant for a sub-6 GHz application. The MM5140 comes in an LGA package rather than the WLCSP of MM5130. That makes it easier for customers to design their boards, as the LGA package has a bigger pitch.

Moreover, Menlo Micro has eliminated some external components for the MM5140 design reduces its complexity. This helps in simplifying RF front-end development including receivers and transmitters, beamforming antennas, and RF filters. These are necessary for radar systems and advanced radio architectures.

5G base stations typically use RF/microwave solid-state switches and RF electromagnetic relays that the MM5140 SP4T switch can replace. The replacement offers significant improvements over the competing technologies, especially in the integrated capability, BOM count reduction, and real-estate savings on the board. Moreover, the MM5140 SP4T switch exhibits far better reliability over the other competing technologies.

Wi-Fi 5.0 to Wi-Fi 7.0

Both on smartphones and in living rooms, the audio & video streaming revolution is producing an insatiable demand for speed and bandwidth. To satisfy this demand, in the early 2010s, we had the Wi-Fi 5. However, this lasted only for a decade or so, because by then, consumers had bidirectional video applications such as Webex, WhatsApp, and other social media uploads like TikTok. These had begun to alter not only the consumer landscape but also that of the enterprise.

That led to the catapulting of Wi-Fi 6 to the arena for better management of the huge traffic of streamlining wireless transmissions. This was followed by Wi-Fi 6E which literally extended the benefits of its predecessor with the availability of the 6 GHz band. The pandemic of Covid-19 in 2020 was the moment for Wi-Fi 6 and Wi-Fi 6E, as is evident from the 1+ billion chips of Wi-Fi 6 and Wi-Fi 6E that Broadcom shipped in the past three years.

And still, the demand for higher bandwidth and speed continues only to increase. A recent study has shown that consumer spending on games has increased by 40%. This involves not only devices operating at higher speed, but also the use of newer technology like AR or augmented reality and VR or virtual reaility headsets as new gaming devices. While these devices demand unprecedented levels of immersion while playing, they also call for deterministic and reliable wireless data.

So, we are now moving towards Wi-Fi 7. It has the ability to incorporate 320-MHz channels into the 6 GHz band and employ the 4096-QAM modulation technique, thereby effectively doubling the channel bandwidth. Additionally, it employs better technologies for lowering latency and bolstering determinism. These include AFC or automatic frequency coordination and MLO or multi-link operation.

Wi-Fi 7 comes with spectrum flexibility spanning three bands. However, the critical role is played by the incorporation of 320 MHz channels into the 6GHz band for doubling the speed. For boosting the coverage and the overall network performance, there is the 4096-QAM technique that plays a crucial role.

Wi-Fi 7 can rapidly aggregate channels in congested, high-density networks. This is due to its MLO or multi-link operation that significantly improves its deterministic performance. By rapidly switching traffic among several channels, Wi-Fi 7 can drive greater capacity, thereby facilitating commercial-grade QoS or quality of service in its networks.

Another technology that Wi-Fi 7 utilizes is AFC or automatic frequency coordination. This technique allocates optimum spectrum, thereby enabling high-power access points and extending the 6 GHz range outdoors and indoors. According to Broadcom, its Wi-Fi 7 designs with AFC are capable of 63 times greater transmitting power. This helps not only to extend the range but also the coverage of the 6 GHz band in use.

Therefore, with its immense focus on speed, latency, and determinism, Wi-Fi 7 has entered our lives and is here to stay. According to the forecast of industry technology analysts, revenue from Wi-Fi 7 will supersede that from any other Wi-Fi technology so far in the next five years.

What are Piezoelectric Audio Devices?

The piezoelectric effect is a versatile and extremely useful phenomenon. Engineers have adopted this phenomenon in various transducer applications. Some of these applications involve transforming the applied voltage to mechanical strain output, for use as a basic source of sound. In a complementary mode, the application of mechanical stress to the Piezo material causes the rugged sensor to produce a voltage. Piezoelectric devices are low-cost, reliable, and rugged, and this allows engineers to exploit their unique properties.

Piezo-based speakers offer many attributes as sound sources. Unlike electrodynamic speakers, piezo-based speakers can be relatively thin, yet create very high sound pressure levels. However, mechanical and physical material issues can limit their audio quality. Now, a team at MIT is changing all this. They have developed a dense array of tiny dome speakers that they have based on Piezo technology. They have significantly transformed the classic analog function of loudspeakers. Their new loudspeakers are paper-thin, very flexible, and fully capable of turning any surface into an active audio source.

Although there are conventional thin-film loudspeakers, the basic requirement is the film must be free to bend to produce sounds. Firmly mounting such thin-film loudspeakers to a surface would attenuate their output and dampen the vibrations, while limiting their frequency response tremendously.

However, the ingenious approach of the MIT team has solved the problem in a rather unique way. Their new loudspeaker does not have to vibrate the entire material surface. Rather, they have fabricated tiny domes on a thin layer of piezoelectric material, such that each dome can vibrate independently. Each dome is about 15 µm in height, and they move up and down by only half a micron when vibrating. As each dome forms a single sound-generating unit, it requires thousands of these tiny domes to vibrate together to produce audible sounds. While the basic loudspeaker is only 120-µm thick, it weighs only 2 grams. Only standard processes are necessary to manufacture this loudspeaker at low costs.

Spacer layers surround the domes on the bottom and top of the film. This helps to protect the domes from the mounting surface, and at the same time allows them to freely vibrate. These spacer layers also protect the domes from impact and abrasion during daily handling, thereby enhancing the durability of the loudspeaker.

To make the film loudspeakers, the researchers used a thin sheet of PET or polyethylene terephthalate. This is a standard plastic used for a variety of applications. They used a laser to cut tiny holes in the sheet while laminating its underside with an 8-µm thick film of PVDF or polyvinylidene fluoride. This is a common industrial and commercial coating. Then they applied vacuum and heat to bond the two sheets.

As the PVDF layer is very thin, the pressure difference that the vacuum creates together with the heat causes it to bulge, but it cannot force its way through the PET layer. This makes the tiny domes protrude through the holes. The researchers laminated the free side of the PVDF layer with another layer of PET and this acts like a spacer between the bonding surface and the domes. Regardless of the rigid bonding surface, the film loudspeaker could generate a sound pressure level of 66 dB at 30 cm.