Where did the idea of the Digital Micro Switch technology come from?
Our lineage of innovation & entrepreneurship starts with the “Wizard of Menlo Park” himself, Thomas A. Edison. Born out of a decade-long effort at General Electric’s Global Research Center, Menlo Micro is bringing to the market a series of technology innovations that have enabled us to reinvent one of the most essential and fundamental building blocks required for nearly every electronic system imaginable, the electronic switch. This was born from a vision to provide programmable, remotely controllable relays to substantially enhance energy management and electrification in a great number of applications that have not seen any advancement in 50-150 years. These applications still use classic electromechanical and mechanical architectures and even fuses. The vision made possible by our introduction of as close to an ideal switch as has ever been introduced.
Switches are ubiquitous and perform various functions in electronic systems. Some switch AC or DC signals, and others switch radio frequency (RF) signals. Some are mechanical switches that open and close a circuit with physical contacts, and others are solid-state switches that use a semiconductor-based device to perform that function. All switches are designed for specific application requirements, and all come with trade-offs.
What are the features and benefits that it brings to the industry?
Driven by the extreme operating conditions required by General Electric’s industrial businesses, the GE Global Research team was led to look at the micro-mechanical switch from an entirely different perspective. That different approach led to a deep understanding of failure modes, and in turn, the development of a unique set of patented metal alloys and processing techniques that enable reliable operation without sacrificing performance while substantially enhancing both frequency performance – DC to mmWave and Power Handling mW to KW.
With the core technology development completed and qualified for production in GE systems, the Menlo Micro team is bringing high-performance switch products to multiple end markets, with a goal of revolutionizing the way electronic systems are designed. Some of the key attributes of the Menlo Micro switch technology include:
Size. Board space and weight are at a premium in many applications. Traditional mechanical switches take up lots of space, have limited number of channels, and in some cases, need to be manually assembled. Menlo Micro switching elements are smaller than the width of a human hair and are architected to be scalable, depending on the power ratings required. All switches are manufactured using automated wafer-level processes and tools. They are so small we can fit hundreds of them in a space smaller than 10mm2.
Speed. Making mechanical structures small also means you can make them move fast. Make them real small, and you can make them open and close real fast. A typical mechanical switch might operate in a few milliseconds, whereas Menlo Micro switches can operate 1000x faster, in only a few microseconds. This can have enormous implications on systems that were previously limited in performance by how fast they could reconfigure, or open and close critical circuits.
Power Handling. This is an area where Menlo Micro completely throws conventional wisdom out the window. When faced with the prospect of handling higher power levels, most people think bigger. More mass, more metal, larger air gaps between conductors. We took a different approach. We make everything smaller and move the electrical contacts closer and closer together. Our miniaturized switches and scalable architecture allows us to handle 100s of volts and 10s of amps, KWs without arcing. Freidrich Paschen would be proud.
Power Efficiency. In an increasing number of systems, power is getting more and more expensive. When you’re working off a battery, every 0.1dB and every microamp counts. Both RF and AC/DC losses need to be balanced with amplification and in some cases extra power supplies. This is where the Menlo Micro technology really shines. We can scale our switches to have ultra-low losses, from 1 ohm down to a few milliohms. And not only that, but our electrostatic-driven actuator means that a single switch only needs a few pA to function, making it ideal for industrial IoT actuation applications. That’s 0.000000000001 Amps for those who are counting.
Reliability. When you are developing products to meet the needs of businesses that serve markets like healthcare, aviation, and other mission-critical industrial applications, reliability is not an after-thought, it’s the primary design criteria. In the end, that’s why we are here. Our mechanical switching device has lifetimes more than 1000x longer than traditional mechanical switches; not millions of cycles, but 10s of billions of cycles without degrading performance. Even more important than the performance that’s been demonstrated to date, is the deep understanding in material science, reliability and failure analysis that enables us to model and predict failures so we can push the technology even further. We have lots of performance enhancements planned in the very near future.
Product Development and Scalability. Cost and scalability have long since been big issues for MEMS – especially when it has come to contact switches. What Menlo has done is simplify the cantilever beam to a simple, characterized unit cell. We are able to do this because of the robustness of the materials. Subsequently, instead of designing a new process for every product, we develop multiple products with the same unit cell, connected differently in each product circuit, built on the same process. You can think of our unit cell switch as our “transistor”. How we connect up those “transistors” or switches makes up our product. This gives us scalability and in-turn drives cost structure.
These are just some of the features that have been demonstrated so far…and the fun part is that we’re just getting started. As we continue to build up the Menlo Micro team, and engage with dozens and dozens of innovative customers, our product capability will expand and our innovation will never stop. Innovation and entrepreneurship: it is in our lineage and part of our DNA.
Which markets and applications are you targeting and you think can benefit the most with the DMS products?
Menlo Micro’s DMS platform is a game changer for those who design electronic systems, with a market opportunity of more than $5 billion. It has the potential to serve multiple industries, including: next generation 5G mobile networks, industrial IOT markets, battery management, home-automation, electronic vehicles and medical instrumentation.
What are the challenges you face to penetrate the current market for electromagnetic switches?
Many of the applications that we are going after already use electromechanical and mechanical switches. Because of our high reliability MEMS structure, we are able to outperform the “classic” mechanical switch by orders of magnitude on reliability, size and power dissipation. Not to mention that overall frequency and power handling performance are enhanced as well. The biggest challenge is the negative perception that so many have of MEMS switches now because of the market failures of the past. Our existence proof in high end MRI applications that have been shipping with our solution for 2 years now helps us immensely with this issue – customers rapidly see that we have solved the reliability issue for MEMS.
You mention that the wireless telecom applications is one of your key markets – how does the Digital Micro Switch fit into the next wave of 5G infrastructure?
Menlo Micro’s next-generation wafer-level process enables the fabrication of micro-mechanical switches for both power and RF applications. With the DMS process, Menlo has demonstrated an industry leading RonCoff performance below 50 fs, and will sample devices at 10 fs, a 10x improvement compared to traditional switches, in the next 12-18 months. RonCoff is the key figure-of-merit used by the semiconductor industry to quantify the quality or performance of an RF switch. With the proliferation of new frequency bands, including much higher frequencies, the performance of the RF switch will become an increasingly important factor in enabling higher data-rates, longer battery life, and much more flexible architectures for 5G applications. Menlo’s switch technology has already been modeled and shows outstanding performance in the mmWave frequency ranges such as the much discussed 28GHz. Menlo Micro will demonstrate this technology at Mobile World Congress this week in Barcelona, Spain, Hall
Meeting Room, 6O24MR.
Using proprietary materials, designs, and wafer-level processing techniques licensed exclusively from General Electric, Menlo’s DMS technology has proven reliability and can not only handle kW of power, but can be scaled down in size and cost to making it cost and size -competitive with silicon-on-insulator (SOI), making it a viable technology not only for wireless infrastructure, but a leading technology for handset applications. The switches are made from proprietary metal alloys, plated onto glass wafers with through glass vias (TGV) from supplier and investor Corning Incorporated (NYSE: GLW), resulting in extremely high reliability and improved RF performance.
In addition to the key figure-of-merit RonCoff, Menlo has demonstrated linearity above +90dBm, which is 15-20dB better than what is traditionally considered best-in-class performance. Switches that the company now has in development will have insertion loss < 0.3dB at 12GHz and cover bandwidths from DC-18GHz, with the ability to extend up to even higher frequencies, including mmWave.
Cost competitive to existing semiconductor technologies with performance to mmWave capability makes Menlo’s DMS technology a leadership technology for 5G and beyond.
Menlo and Corning are working together to establish a supply chain for high-performance semiconductor glass wafers with TGV technology, and scale-up of the manufacturing processes.
In addition to the performance advantages, the Menlo DMS process is typically 12 metal layers, thanks to advancements in high-temperature, high-reliability alloys. With only half of the manufacturing steps required of a typical CMOS wafer, the process can be scaled to be cost-competitive with SOI, enabling high-volume applications such as antenna switch modules (ASM), filters, tuning networks, and band-switching for next-generation 5G smartphones and communications systems. Menlo is actively engaged with a number of technology partners to license the DMS technology into the consumer mobile handset market.
Where are you in terms of new product development and product availability?
Menlo has a robust line of technology innovation and we are on-track with our product development goals. We are sampling 3 products today with 1 in production. 3 additional products will be in production during 2018. Shipping today is the 7100 HV switch for MRI coil applications. Additional products in production will be a cost reduction of the 7100 as well as our 3GHz tuning product and our 18GHz RF switch. We are also developing products for several semi-custom engagements in the wireless handset and the test instrumentation markets. We will be in full production in our commercial 8in fab by Q32018, with engineering samples from that facility throughout the 1st half of 2018.
With power loss and handling becoming more critical to the performance of end applications, how do the DMS products perform in regards to power?
This is the next biggest differentiator for our technology besides reliability. We are able to handle kW of power – 10s of Amps and 100s of Volts, but because we are an electrostatic device we only draw pico Amperes to stay turned on. Additionally, because of our extremely low contact resistance we dissipate very little power in the device itself. Compared to other electromechanically devices or even semiconductor devices such as IGBTs, MOSFETs or SiC, we can switch 240V at 10A without a heatsink, drastically reducing size and cost.
What is the next “big” thing you see for the semiconductor industry?
I believe that we are going to see innovation on 2 levels – IoT driving sensors to collect data and actuation to take action on the data (i.e. – turn stuff on and off). Both of these applications are going to require basic materials and components invention to truly revolutionize then and make them commercially viable to have everywhere. As you can see, Menlo’s play is about fundamental materials to revolutionize components that haven’t changed in 150 years, but need to for the changes in the markets that are developing but cannot progress without these revolutions.
There has been a resurgence of “startups” in the technology industry. What do you think is driving this?
I believe that with the huge growth in technologies required for the edge and the electrification of everything, there is a resurgence of invention. I also believe that we are beginning to see some more fundamental materials and components inventions because of this. Coupling that invention with the availability of cash is spurring on more start-ups. There are some interesting changes to the dynamics however – we are seeing more strategic and “family office” investment than before. Additionally, entrepreneurs are being called to leverage and de-risk better, especially in hardware plays. It looks like the next 3-5 years should be a robust timeframe for start-ups.