Synthetic jet cooling for small form factor computing
To say that Moore’s Law is now in full effect would be an understatement; electronics are becoming exponentially more fast and powerful at an ever-increasing rate. Meanwhile, the sizes of the devices are becoming smaller and smaller, largely due to the needs and wants of consumers. While there seems to be no limit as to how powerful chip designers can make speed and storage, there are several barriers to how small we can continue to make the products that utilize these chips. One of the most critical of these barriers is proper thermal management. Synthetic jet cooling offers very high heat transfer rates in very compact, highly reliable packages. This enables products to achieve the small size and power density associated with active cooling, without the low system Mean Time Between Failure (MTBF) of fans.
As products become smaller and more feature rich, the heat that builds up can often cause them to become performance limited or fail altogether. In the past, small form factors have relied on large passive heat sinks, unreliable fans, or conduction cooling to transfer heat to a different part of the product. Each of these techniques is becoming increasingly problematic as engineers continue to create smaller designs with more power and higher levels of integration.
Heat sinks take up space within products and also add weight and cost. The thermal dissipation of a passive heat sink alone will not meet the needs of engineers challenged with cutting-edge designs, as there is just too much power packaged into too small a space. Traditionally, when engineers are faced with a critical thermal problem they turn to fans. Unfortunately, fans’ rotating machinery contains bearings, and bearings have friction. This means that when you calculate the reliability of your product, you do not anticipate if your fan bearing will fail, but when your fan bearing will fail. Often the tradeoff between better cooling and low reliability is not worth using fans, which results in defeaturing, reduced performance, and specmanship.
Conduction cooling is no longer an option for most computing devices, and may soon be ineffectual for almost all small form factor devices due to the continual rise in heat generated by processors. An active and reliable mode of cooling is needed that can be built into thinner embedded systems that does not hinder design or functionality. To this end, synthetic jets are being applied by electronics developers and manufacturers to address the growing thermal problem.
How synthetic jets work
Originally developed for broad applications, such as elimination of flow separation on active flight control surfaces and microfluidic devices for heat transfer and cooling, synthetic jet technology allows for an unsteady, pulsating jet to sweep quickly over the heated surface. The vortices inherent to the flow create a high level of mixing.
SynJets use flexible diaphragms to create a turbulent, pulsating airflow that can be directed at precise locations for hot-spot cooling. The oscillating diaphragms create high-velocity pulses of air. This high-velocity pulse of air removes the heat being conducted by the heat sink, and at the same time pulls entrained air from the area in its wake. This process continues so quickly that the pulses of air aggregate to form what appears to be a steady, unidirectional flow.
Figure 1 shows a simplified example of a SynJet cooler. As the diaphragm moves up and down, the high-velocity jet flow is ejected from the nozzles. The local low pressure created by this high velocity results in the secondary flow – the entrained air – being drawn into the flow field and over the heat sink. This secondary flow provides additional fluid mass that can be used to carry the heat away from the heat sink.
Key benefits of synthetic jet technology
Typically, the addition of a SynJet to designs previously cooled by natural convection results in heat sinks less than one-third of the original size and a cooling and heat transfer improvement of up to 60 percent. Product designers use this feature in a number of ways depending on the needs of their customer. Some choose to make their product as small as possible, while others take advantage of the reduced weight to differentiate their product. Regardless of the final configuration, the simple effect of increasing the thermal design envelope opens up new options for the system design.
Additionally, SynJets are a scalable technology. Solutions such as the Nuventix RazorJet are as small as 3 mm in height and used to cool low-wattage electronics such as handheld devices (Figure 2). Higher capacity products have been designed with the ability to cool hundreds of watts of power and are used to cool high-power electronic devices as well as LED lights and datacenters.
SynJets are inherently reliable, whereas rotating machinery like fans and pumps are inherently unreliable and therefore typically require safeguards and additional technology to increase their reliability. Fans require rotating blades, which create loads and require the use of bearings to extend the life of the fan’s motor. These bearings are failure prone so they have to have lubricants added to them. Lubricants can leak out or become contaminated by particulate so they require seals; therefore, the lifetime of your fan, and thereby your product, is determined by how well bearings are sealed. This is just too complicated and uncertain.
Because SynJets have no need for bearings and thus no friction, they have incredibly high reliability and do not have the complicated technological band-aid stack that fans do. The typical L10 lifetime of a SynJet is 100,000 hours at 60 ºC with a 90 percent confidence factor; special classes of SynJets have been developed that have 2.5x that lifetime. On top of the lack of bearings and frictional parts, the diaphragms are designed with ample clearance so they are tolerant to the ingress of particulates such as dust or sand. A 30-year dust test for SynJet coolers has shown virtually no effect on cooling performance.
A number of customers have found that by adding a SynJet, they can increase their system MTBF. This is because having a high-reliability active cooling element in their product lowers the junction temperatures of a number of key components, providing a net system benefit.
SynJet designs also result in low acoustics. The lack of frictional parts such as bearings or brushes eliminates typical acoustic issues associated with interfaces used in active cooling. Unlike fans or blowers, synthetic jets like SynJets maintain low acoustic levels over their lifespan, typically maintaining a dBA of approximately 22, which is about the sound of a low whisper.
Low power consumption
Through the development of very efficient actuators, synthetic jets such as the SynJet require very low power to operate. The reason for this is simple and straightforward – the SynJet is a spring mass system and has a resonance. By operating at a resonant frequency that is typically below 80 Hz, SynJets sip small amounts of power while maintaining their cooling efficacy. Different field configurations and system backpressures can result in different system resonances. Intelligent electronics in the system guarantee operation at the optimal frequency and low power consumption over the life of the product.
Fundamentally, to create a synthetic jet you need to change the volume of a cavity and provide an orifice for fluid to be ejected during the exhalation phase and recharged with fresh air during the inhalation phase. This means that the orifices (or nozzles) can be placed anywhere on the boundary of the volume and flow will be generated (Figure 3). It is possible to make a spherical cavity that generates airflow in all directions as the diaphragms displace the volume.
The implications for product design are significant. The ability to place cooling where you want it and direct airflow where you need it is unprecedented. What this means for product designers is that they can let the cooling system design follow the design requirements of the product, instead of the product design following the design requirements of the cooling.
Due to the size, shape, and technology it is exceptionally simple to adapt synthetic jets into a design. Rather than creating a design to manage heat transfer needs, this cooling solution can be made to fit around the design. The RazorJet is the best representative of the remarkable flexibility of the synthetic jet for hot-spot cooling and small form factors, due to the ability to use multiple jets in various formations without sacrificing size and maximizing heat transfer (Figure 4).
Synthetic cooling solutions for small form factor designs
The future of small form factor devices and mobile and portable electronics relies on a solution to provide efficient cooling and thermal management. Synthetic jet technology may very well be the key.