The shrinkage of electronics allows for more powerful devices to be constructed in smaller packages, and this is best evidenced by modern computing systems. The first computer models took up entire floors of buildings, similar to today’s supercomputing systems, but while the size of supercomputers doesn’t change, the amount of data they can process has scaled exponentially. But the reduction of electronic components doesn’t just allow for electronic devices to become more powerful; it also allows them to shrink in size.
Devices that can be made smaller make them more practical from a consumer point of view, and a good case study is headphones. The first headphones were large and bulky, which made them impractical to carry around. As technology improved, headphones were shrunk in size until they could be comfortably fitted into the ear. Now, technology has improved so much that these in-ear devices can be wireless.
Reducing the size of electronics is more than just making components smaller; the energy source for devices also needs to shrink. This brings us to a conundrum with power sources: Reducing the size of a power source reduces the amount of energy it can store. Therefore, designers of electronic devices are often playing a game of cat and mouse with modern electronics, as they need to shrink the power source while also reducing the energy consumption of the device to maintain the battery operational time.
So what can designers look toward in the future for creating even smaller devices, and what can designers do now while these technologies continue to be developed?
Battery technologies
Designers have a multitude of battery technologies available to them, but only one remains viable for small, portable devices: lithium-ion (Li-ion). Current Li-ion technology allows for batteries with a high energy density that can be charged quickly. This allows for Li-ion batteries to be designed significantly smaller than other technologies (such as alkali and lead-acid); however, this comes at a price.
Lithium-ion
Li-ion batteries require multiple safety systems, as they are easily damaged. When this happens, the result can be extremely violent. When Li-ion batteries are pierced or damaged, they almost always form an internal short-circuit between their electrodes. This causes a short-circuit that is very large in size which, in turn, generates hydrogen gas and heat. The result of this is that the battery swells up, vents the hydrogen, and catches fire. This fire can damage anything nearby as well as cause additional fires in other batteries.
Solid-state
Solid-state batteries are a new technology in development that could very easily replace current Li-ion batteries (which use a liquid electrolyte). As the name suggests, solid-state batteries are made entirely from solid materials. This makes them much more resistant to damage and internal short-circuits. Furthermore, it is believed by researchers that solid-state batteries are safer, capable of storing more energy, and, as such, will allow for smaller batteries.
However, solid-state batteries are not without their flaws. One major flaw is the formation of internal dendrites. Lithium-based, solid-state batteries are prone to forming small crystals on their anodes that grow toward the cathode (called a dendrite). If this crystal makes contact with the cathode, the cell short-circuits and stops functioning.
Supercapacitors
Supercapacitors are another option that could power devices of the future, thanks to their ability to store large amounts of energy at low voltages. Furthermore, supercapacitors can charge and discharge extremely fast, making them ideal for systems needing quick charge features or those that need to suddenly store large amounts of energy (such as a regenerative braking system).
The problem with supercapacitors, though, is that this technology is unable to store energy at the same scale as most other battery technologies. As such, it may not be able to provide long-term energy storage for high-power devices. Furthermore, the fact that it discharges quickly also raises safety concerns with the possibility of sparks from a short-circuit.
Alternative power sources
Batteries store energy, but generating energy on the fly can be far more advantageous, as device operators do not need to remember to charge the device and the technology can be mounted remotely and away from any power source, as well as provide greater safety.
Solar
Small solar cells are commonly found on calculators, which can easily operate in daylight as well as office light. While these are unable to generate enough power for smart devices or laptops, they can easily be used to power smaller devices like IoT sensors and wrist-worn sports watches. But long term, don’t expect solar technologies to be powering anything that needs more than 1 W of power.
Compressed hydrogen
Compressed hydrogen could be another stored energy solution for the future. If hydrogen and oxygen are passed through a special proton exchange membrane, they can be made to directly generate electricity. This creates a highly efficient process that can take advantage of the energy density offered by hydrogen. However, the storage of compressed hydrogen, even in a small portable device, raises safety concerns similar to those associated with Li-ion batteries. Furthermore, it is difficult to minimize the size of these fuel cells, and it therefore may not be a viable power source for several decades.
Thermal
Thermal energy can be directly converted into electrical energy via the Peltier effect. Researchers have been able to create energy sources that can be worn like a ring on a finger, and the temperature difference between the air and body is enough to create electricity. However, Peltier generators are notoriously inefficient, and such devices only operate efficiently with a large temperature difference (about 80˚C).
Radiation
Another future energy source is radiation, and researchers have designed potential diamond batteries that directly convert radioactive decay into electricity. A battery designed with this technology could be made small and provide power for generations without ever needing a recharge. However, such a power source would provide minuscule amounts of power (nanowatts). It also presents as a potential environmental hazard.
Power consumption reduction techniques
For designers who are looking to create smaller devices today, only battery technologies that have been proven can be used, and that most likely means the use of lithium-ion. However, it is not all doom and gloom, as such batteries can be purchased in very small sizes. The trick is finding a way to reduce the energy consumption of a device as much as possible so that the operational time of the battery can be extended.
Energy is energy and totally independent of time
The first rule that designers need to understand is that energy has nothing to do with time. A device can use a large amount of power for a short amount of time, or a very small amount of power for a long time, and overall, the device could use the exact same amount of power.
This concept can be used to seriously reduce the energy consumption of devices by recognizing when a device needs to operate and when it does not. For example, an IoT device that operates on batteries with a Wi-Fi connection may need to send data only once every 10 seconds. If this is the case, the device can spend 9.9 seconds of its operational time in a deep-sleep mode, whereby energy consumption is minimized. Thus, the large amount of energy consumption during transmissions is averaged over the times when the device is not in use; this, in turn, can reduce energy consumption dramatically.
Reduce clock speeds
Many designers who work with microcontrollers and microprocessors tend to try and squeeze every last megahertz they can possibly get. While this may be practical for data-heavy operations, it is far from ideal in portable designs, and reducing the clock speed can lead to major energy savings.
The reason why reducing clock speeds on CMOS-based logic helps to reduce energy consumption results from how CMOS logic works. During a logic 1 or 0, CMOS consumes next to no power, as the input to CMOS gates is a capacitor, and CMOS transistor pairs (P and N) operate in compliment. This means that during either a 1 or 0 state, there is no path between power and ground (hence, no energy is consumed). However, during a change in logical state, there is a brief path between ground and power (as the transistor goes through their linear region), and this provides a path between power and ground. This is when energy is consumed, and the more the transistor remains in this region, the more energy is consumed.
Therefore, reducing the clock speed of a system can help to reduce energy consumption, but understand that the reduction in clock speed means fewer instructions are executed per second, and so it may take more time to execute the same task. This comes back to the previous problem of energy consumption: You may consume less power, but you still require the same amount of energy to complete a task.
Could custom silicon change the game?
One technology that could provide future devices with major energy-saving options is custom silicon and chiplets. The biggest drawback of off-the-shelf electronics is the often-large number of peripherals on chips that go unused and consume energy unnecessarily. This could include instructions that are not needed, peripherals that support unused buses, and generic circuits that are not needed in the final design.
While many manufacturers of microcontrollers do offer the option to turn off these areas to reduce power consumption, valuable silicon space is still being occupied by unnecessary hardware, and, as such, reducing the efficiency of the design. Custom silicon devices, however, enable designers to pick and choose the exact hardware that they need in their design and, by doing so, can either reduce the size of their chips significantly or fully utilize the silicon space.
Creating custom designs can be done in a number of ways, including designing the actual silicon itself (ASIC) or picking and choosing pre-designed silicon dies and mounting them into a single package. It is more likely that the use of pre-made silicon dies will be adopted, as these are far easier to manufacture and assemble and are also more economical.
Conclusion
There are many power options to portable devices, but only a few provide credible options. As electronics get smaller, the power requirements also fall (smaller transistors consume less energy); however, the increased number of transistors increases the overall energy consumption.
As such, designers may be more reliant on power-saving techniques such as under-clocking and removal of unneeded hardware. Furthermore, the future could see electronics move toward custom devices, and entire circuits can be constructed onto a single package without the need for external components or parts, and such devices could be ordered like PCBs.
