Skip to main content

The Rise of the (Micro) Machines

If you’re the kind of person who is drawn to technology, there’s a strong probability that you grew up on an entertainment diet that included large dollops of science fiction. If you were born in the last century (before network executives realised the size and majesty of the market desperate for quality SciFi), there’s a good chance that Star Trek was one of those dollops.

There’s no doubt of the inordinate influence the show has had over western culture since its inception in 1966. There aren’t many people who won’t get a reference to ‘beaming up’ or ‘warp speed’, even if they despise SciFi with all their being. But it’s not just the lingo that has been influenced by Trek: many scientists and engineers have been inspired to bring what they saw on TV as a child to life in their careers.

For example, warp theory (i.e. the idea of contracting space in front of a spacecraft and re-expanding space behind it, to travel faster than light) is a real thing. The Alcubierre warp drive theory was proposed as a PhD thesis in 1994 – research that was inspired by the show. A lot of serious work has been done since that has taken the idea from ‘impossible’ to within the realms of probability for engineers to create one day.

Perhaps the other best-known technology in the series was the ubiquitous ‘tricorder’, the hand-held device that performed sensor scans of the environment, recorded data and analysed it: hence 'tricorder' for the three functions of sensing, recording, and computing. The more specialised version, the medical tricorder - as used by doctors McCoy (TOS) and Crusher (TNG) - has also become something of a reality thanks to the Qualcomm Tricorder XPRIZE which has produced some promising results in the field of hand-held diagnostics that were widely reported in 2017.

The unsung heroes behind both present and future Star Trek gizmos are the sensors. Whether they are collecting real-time data for the control loops that create a stable warp field or monitoring a patient suffering from Pa'nar Syndrome (or just regular old Earth flu), without all those sensors, there’s simply nothing to compute.

Sensor and actuator technology that could lead to the kind of world envisioned by the Star Trek creators has advanced in leaps and bounds since filming stopped on Star Trek TNG (1994) and DS9 (1999), especially in the area of Micro-Electro-Mechanical Systems or MEMs.

MEMS are miniature systems manufactured from a variety of materials (including semiconductors, plastics, ceramics, ferroelectric, magnetic, and biomaterials) and usually feature a combination of both electrical and mechanical functions.

Mechanical features are created by depositing layers of polycrystalline silicon along with ‘sacrificial layers’ of silicon dioxide. The layers are then optically patterned and etched before the sacrificial layers are dissolved, leaving three-dimensional structures such as gears, wheels, microscopic cantilevers, chambers, nozzles, and mirrors.

MEMS micro mirror

Gearing for a moving micro-mirror in a mirror array. Courtesy: Sandia National Laboratory

If the process sounds eerily similar to how integrated circuits are made, that’s because it is: MEMs processes and materials technology have generally been developed downstream of integrated-circuit (IC) manufacturing developments, and so have closely shadowed each new IC milestone. But to understand the journey, we need to go back a little further.

It’s Been a Long Road…Getting From There to Here

Way back in 1856 William Thomson (Lord Kelvin) had been investigating the changes in resistance in iron and copper with elongation as strain-induced conductivity changes had become a major headache for telegraph companies, causing pronounced signal propagation variations. In his Bakerian lecture to the Royal Society of London that year, Kelvin reported on his experiments where parallel lengths of copper and iron wires had been stretched with a weight and the difference in their resistance change was measured with a modified Wheatstone bridge. He concluded that, as the elongation was the same for both wires, “the effect observed depends truly on variations in their conductivities.”

Lord Kelvin

Lord Kelvin (1824 to 1907)

Piezo Resistance

In 1935, J W Cookson coined the term ‘piezo-resistance’ for this effect (from the Greek ‘piezen’, to press) not long before the bonded metallic strain gauge was developed by Arthur Ruge in 1938. For us, however, the story starts to get interesting in 1954 when C. S. Smith, a researcher who was visiting Bell Laboratories from Case Western Reserve University and interested in the anisotropic electrical properties of materials noted, in his seminal paper, a piezoresistive effect in silicon and germanium several magnitudes larger than that in metals.

A few smart cookies realised this was a game-changing revelation for the measurement of strain, and by extension, pressure. One of those who saw the potential of silicon transducers was Dr Anthony D. Kurtz who founded Kulite Inc. (which is still a leading manufacturer) and was producing commercial silicon strain gauges before the end of the 1950’s.

Resonant Gate Transistor

The next great MEMS milestone was a direct development of integrated circuit technology. Although Jack Kilby (working at Texas Instruments) and Robert Noyce (working at Fairchild Semiconductor) share credit for independently inventing the integrated circuit, it was the planar process (evaporating lines of conductive metal “wires” directly onto a silicon wafer’s surface) identified by Noyce and patented by Fairchild (after a lot of litigation) that led to the creation and patenting in 1968 of the resonant gate transistor or RGT.

Resonant Gate Transistor

The RGT was a field-effect transistor that incorporated a cantilevered beam which resonated at a specific frequency to provide high-Q-frequency discrimination. Although designed as a narrow-band filter, it was also the first example of a micro electrostatic actuator and the first demonstration of surface micro machining techniques.

Thermal Inkjet

The innovation that made MEMs ubiquitous, even though few people realised the significance at the time, was the introduction of the micro machined thermal inkjet nozzle by Hewlett Packard in 1979. Hands up if you remember the introduction of the inkjet (for thermal inkjet) in 1984 that truly commercialised this technology.

Thermal Inkjet

The thermal inkjet works by passing current through a thin-film resistor at the back of the ink chamber to heat the ink very rapidly: in the order of 100°C/µsec. A small amount of ink close to the heating resistor rises to about 340°C in around 3-µsec. This small amount of superheated ink vaporizes forming a bubble that pushes ink through a micro machined (to around 70-um) nozzle. The majority of ink does not boil - only heating to around 40-45°C – allowing it to safely eject onto your paper as part of that strongly-worded letter to the editor.

Improved Processes

The 1980’s and 1990’s were marked by some major improvements in micro-manufacturing processes that allowed innovation to flourish. Some highlights include:

LIGA – A fabrication technology used to create high-aspect-ratio (in the order of 100:1) microstructures. LIGA is a German acronym of Lithographie, Galvanoformung, Abformung (i.e. lithography, electroplating, and moulding) for the process developed in the early 1980s by a team under the leadership of Erwin Willy Becker and Wolfgang Ehrfeld at the Karlsruhe Nuclear Research Centre.

LIGA Process

High aspect ratios achievable using LIGA

1988 saw the first rotary micro electrostatic side drive motors fabricated at a UC Berkeley.

1989 saw the lateral comb drive emerge, used as MEMs linear actuators utilising electrostatic forces acting between two electrically conductive combs.

SCREAM - Single Crystal Reactive Etching and Metallization (SCREAM) was developed in 1992 at Cornell University as a bulk micromachining process that uses a single lithography step and reactive ion etching (RIE) of a substrate to fabricate suspended, movable single-crystal beam structures from silicon and gallium arsenide. These can be used in sensors and actuators.

MUMPs – before being broken up in the mid 1990’s, the Microelectronics Center of North Carolina (MCNC) was a state-funded microelectronics research entity (Note: the entity holding this name now is a privately funded nonprofit that specializes in rural broadband) that created a foundry in 1993 with the intension of making microsystems processing accessible and cost-effective for a large variety of users. They developed a process called MUMPs (MultiUser MEMS Processes) which is a three-layer, polysilicon surface micromachining process.

DRIE – In 1994 German industrial powerhouse Robert Bosch GmbH patented its own Deep Reactive-Ion Etching (DRIE) process. This is a pulsed or time-multiplexed etching process, alternating repeatedly between two modes: 1. an isotropic plasma etch, and 2. deposition of a chemically inert passivation layer. The process allows the creation of nearly vertical structures, like etching trenches (10–20 µm deep) for high-density capacitors on Dynamic Random Access Memory (DRAM) wafers.

Vertical channels for capacitors

Vertical channels for capacitors created with the deep reactive ion etching process

SUMMiT - In 1998, another surface micromachining foundry was constructed at Sandia National Laboratories to make use of a process called Sandia Ultra-planar, Multi-level MEMS Technology or SUMMiT IV. This has since evolved into SUMMiT V which is a 1.0 micron, 5-level, surface micromachining technology featuring four layers of poly-Si (for mechanical structures) fabricated above a highly doped poly-silicon layer for electrical interconnect and ground plane.

film deposition

An iterative process of film deposition, patterning, etching and Chemical-mechanical Polishing (CMP) is used to fabricate complex mechanical structures. Structural polysilicon layers are fabricated on top of sacrificial oxide layers which are “release” etched to remove sacrificial layers.

Chemical-mechanical Polishing (CMP)

Game Changer

The real game-changing application that led to the widespread adoption of MEMs was probably Analog Devices’ first MEMS product, the ADXL-50 accelerometer which was small, highly reliable, inexpensive and was able to sense sudden accelerations of at least 50 Gs1, making it an ideal candidate for initiating airbag deployment in cars. It sold by the millions to first-tier automotive suppliers like Delco in North America and Siemens in Europe. Other players, like Motorola in the US, Bosch in Europe and Denso in Japan also joined the MEMs party, increasing the competitive pressure for innovation.

With so many major players involved, it was inevitable that these accelerometers would leak over to a host of non-automotive applications like video games, security devices, appliances, navigational devices and, of course, mobile phones where motion of the entire device was data of interest.

Personal Tricorders

MEMS technology has continued to develop at a stunning pace, making smaller, cheaper and more accurate measurements available everywhere: perhaps nowhere more so than in our personal devices. Modern smartphones contain an array of MEMs sensors that would have been beyond the wildest imaginations of early Star Trek writers, even if tachyons are not yet on the detection menu.

Some of the MEMS highlights include:

Accelerometer - Of course! one of a phone’s most important sensors. Acceleration sensing across all 3 axes, including acceleration caused by gravity, to determine phone movement and orientation.

Gyroscope - Provides more spatial awareness by measuring direction and magnitude of rotation. Sky Mapping apps use gyroscopes to determine the direction in space the phone is pointed.

Magnetometer - Another spatial awareness sensor detecting magnetic field strength across all three axes. Usually, this is the earth's magnetic field so that your mapping app knows which way you are facing on the map but it will detect hard distortions from permanent or electromagnets and soft distortions from lumps of metal, allowing them to be used for metal detection in some instances.

Microphone - MEMS microphones have several advantages over the electret condenser microphones (ECM) traditionally used in phones, including lower power consumption and higher signal-to-noise ratio (SNR) which makes it easier for voice recognition apps to determine wanted sound from background noise.

Environmental – Temperature, humidity and barometric pressure are becoming more commonly available measurements on many phones: your own localised weather station!

Then there are a number of sensors that aren't necessarily MEMs, like the proximity sensor [an infrared LED and IR light detector to find out how close the phone is to your ear], the ambient light sensor [detecting local light levels to adjust the display brightness accordingly], capacitive touchscreen, fingerprint scanner, camera, Soli sensor [essentially a miniature radar system for motion sensing], plus a host of radio gear for your Global Positioning System (GPS), wifi, Bluetooth, 4G/5G, Near Field Communication (NFC) [for your Apple/Google Pay] and ultra-wideband (UWB). It’s pretty astounding what manufacturers have crammed in there to ‘one-up’ their competition. This is just one area where MEMs technology is making new applications possible.

QVAR

MEMs technology continues to evolve at a heady pace and to open up new vistas, with the hope of creating whole new markets, many manufacturers are going off in new and hitherto untried sensor directions. One that caught my eye is STMicro’s QVAR sensor.

The term QVAR is an amalgamation of 'Q', the standard letter used for electric charge, and 'var', short for 'variation'. So, as you might have guessed, we’re talking about an electrostatic variation sensor. To appreciate how useful this can be, we need to consider some physics.

Triboelectric Effect

Pretty much everyone will have had the experience of rubbing a balloon on their hair and, after moving the balloon away from their head, seeing their hair stick up towards the balloon. As far back as 600 BC the Greek philosopher Thales of Miletus recorded that amber, when rubbed with wool, would attract bits of paper. This is the triboelectric effect in action. The prefix tribo- (Greek for ‘rub’) refers to ‘friction' and our word electricity derives from the Greek word for amber, ēlektron: from when English physicist William Gilbert (1544 - 1603) was studying static electricity using amber and decided to call its effect the 'electric' force.

Triboelectric Effect

When two materials composed of different molecules are in direct contact with each other, there is a transfer of electrons that creates an electrostatic attraction between the molecules holding them together. As the electron transfer is not immediately reversible, when the materials are separated, one will have a surplus of electrons (negatively charged) while the other will have a deficit (positively charged), dependant on the properties of each material. The level of charge is somewhat unpredictable, but it can be detected by means of an electrode and an electronic signal conditioning circuit.

So how can this be useful?

When a human being is in any environment, static electricity is produced as a result of even their smallest movements and an electric potential is charged within the human body itself. This static potential differential dissipates within several milliseconds because the human body is capacitively coupled to the ground through the air (Cx) or shoe soles (Cs) and floor (CF).

Capacitive Effect

An electrostatic induction sensor can be thought of as an equivalent capacitive sensor; the charged object being sensed (i.e. the human) can be modelled as the plate of a capacitor while the electrode itself is the other plate. The movement of the charged object with reference to the electrode changes the distance between the two plates and hence the value of the capacitance. For a full treatment on the subject, there is an application note.

Electrostatic charges in the environment

Qvar sensor readings in an office

When the electrodes are on a human, either directly contacting skin or not, this offers enhanced activity detection for things like fitness apps. When used in the general proximity of people, as ‘QVAR radar’ then it is a great charge-based human presence detector which has already found its way into enhancing some ST products like:

LSM6DSV16X - (Coming soon to RS) iNEMO 3D accelerometer and 3D gyroscope

ILPS22QS - (240-0600) Dual full-scale digital pressure sensor

And the upcoming LIS2DUX12 high-performance three-axis linear accelerometer.

Final Thoughts

There’s no two ways about it: MEMS is now increasingly where the analogue and digital worlds meet. This is a trend that shows no signs of abating, as automation increases and device sizes decrease. As new sensing methodologies continue to be developed, the world envisioned by the Star Trek writers seems less like science fiction and more like reality with each innovation. Now, if only I could get a replicator to make my Earl Grey….

Mark completed his Electronic Engineering degree in 1991 and worked in real-time digital signal processing applications engineering for a number of years, before moving into technical marketing.
DesignSpark Electrical Logolinkedin