From student research in the visibility of radio waves and high frequency switching, to Hitachi
Ohmae: When I was in third year junior-high school, I chose to progress to engineering. My parents both worked, and I was keen to start earning, so I went on to an institute of technology (a specialized vocational high school). Once I was enrolled, though, I found that my classmates were all very highly skilled, and I realized that if my classmates were going to be entering employment in five years time, in that five years I’d never beat them, so I entered an engineering university. There weren’t enough hours for research in my two years at university, so I stayed on for another two years in the postgraduate program.
The research office I was in at university was working on antennas, and I researched estimation of the direction of the origin of radio waves, which involved making the direction from which they came visible. It’s probably easiest to understand if you think of it as a technology used in radar. I first became aware of Hitachi when I presented at an international forum in the first year of my master’s and someone from Hitachi approached me. Ultimately, that person got me into working for Hitachi.
Nishimoto: After I graduated junior high school, I went on to an information technology senior high school. The school specialized in programing and systems engineering and for three years I studied information processing and software. While I was studying software it occurred to me that if you want to run software, it’s important to understand hardware [laughs], so I went on to university and into an electronics and electrical faculty, which is where I really got stuck into studying hardware. Then in my fourth year as an undergraduate I joined a research office specializing in integrated circuits, overseen by a teacher who was formerly from Hitachi.
I continued researching micro electro-mechanical systems (MEMS) for two years in a master’s program. My research topic was MEMS switches for switching high radio frequencies (RF) and I also presented academic papers. I wanted to keep working as a researcher after I finished my master’s, and it occurred to me that in a company with a research institute I might be able to engage in very diverse research. Also, the responsible teacher at the research office had been with Hitachi, so I got my teacher to introduce me to Hitachi and I joined the company.
A researcher pursuing radio wave visibility and a researcher responsible for developing electronic circuits for diverse fields
Ohmae: When I joined the company I stressed that I wanted to work on research into the visibility of electromagnetic waves. Radio waves aren’t visible to the eye and some people don’t believe in them, or don’t understand where they come from. I also wanted to work on visibility to better understand radio waves, and I’ve actually been allowed to research that topic the whole time, right up to the present day.
An outcome of my research was a news release in 2014 titled “Hitachi develops technology for making the direction of origin of faint electromagnetic waves in the 2.45 GHz band visible”. I conceived of the idea that it might be possible to detect radio waves with a mechanism that replicated the human eye, and using a spherical radio wave lens and a high-density sensor equivalent to the human retina, I succeeded in making radio waves visible without need of a complex algorithm.
But, radio wave lenses cease to function as a lens when the lens is less than the wave length, and even at 2.45 GHz, devices are large and difficult to carry. You can use radio wave lenses for higher frequency millimetric waves, but the area of focus had shifted to detection of low frequency noise, like from trains and motor vehicles, and little by little the objective diverged away from developing radio wave visibility using radio wave lenses.
After that, I got involved in R&D into small-scale sensors for identifying sources of electromagnetic noise common in malfunctions in trains and motor vehicles. In 2017, in collaboration with Kanazawa University, I developed a small-scale sensor to prevent malfunctions in automated equipment by identifying sources of electromagnetic noise. I’m currently involved in research into technology in the power electronics (power control) field for analyzing how noise is caused and carried.
Nishimoto: It’s about 15 years since I joined the company, and if I think back, I’ve been involved with a whole range of devices. When I was a student, I thought that if I was in Hitachi I’d be able to engage in a wide range of research without being restricted to any one research field, and in effect, I’ve achieved that. For example, right after I joined the company it was at a time when Hitachi was manufacturing hard discs, so I researched signal processing in disc testing devices. After that, I got involved in a wide range of fields of development, including R&D of the signal sampling circuits of analytical devices for clinical testing and R&D of power amplifier circuits, signal detection circuits for semiconductor testing devices, high-voltage power circuits, and developing semiconductor integrated circuits to drive ultrasonic vibrators for ultrasonic diagnostic devices in the healthcare field.
Discovery that it is possible to make a filter with good characteristics with a staggered condenser connector
Ohmae: In 2022 I won the top prize in the Japan Electrical Manufacturer’s Association’s Electrical Manufacturer’s Engineering Achievement Awards for my ongoing research into radio waves and noise. The award was for technology to suppress normal mode noise in inverters for 800V electric vehicles.
Electrical and electronic equipment needs electromagnetic compatibility (EMC). Noise is unintentionally produced when electrical or electronic equipment activates. There are two conceptualizations of EMC: keeping noise at or below a standard level, and no malfunction on receival of a fixed radio wave. One of those is to keep noise at or below a level regulated by nations or regions.
The Hitachi Group manufactures power transformation inverters to drive AC motors from DC electric batteries. Noise is produced when the inverter switch is switched.
Noise suppressing technology is essential to ensure no noise leakage, but it’s very difficult to meet EMC standards in devices like motor vehicles or trains that handle large amounts of electrical power. Irrespective of the power handled, regulated noise levels are in the order of nano or micro watts, but trains and motor vehicles handle several hundred kilowatts of power. Noise has to be kept within a very low range of tolerance. For example, the idea is to combine filters to make sure there’s no noise leakage, but as power increases, filters have to get bigger.
Our mission is to make inverters for HEV/EVs lighter and higher performance. In Japan, 400V inverters are the norm, but in Europe the trend was for higher voltage 800V inverters, with a view to shortening EV charging time. The problem with 800V is that noise amplitude increases. On the other hand, be it 400V or 800V, the regulated noise level is the same, and it’s desirable that inverter size doesn’t change. There were big and thorny problems associated with noise suppression.
Condensers are introduced to the positive and negative electrodes of inverters to act as noise suppressing filters, but the problem was that a filter necessary to suppress noise from an 800V inverter was huge. We discovered that we could make a filter with good characteristics with a staggered condenser connector. If a reverse phase current is nearby, the inductance (when change in current becomes induced electromotive force) that corresponds to resistance in an AC circuit gets smaller. We used that characteristic to reduce inductance, improve filter characteristics and reduce size.
We have patented the structure and introduced it to a high-voltage 800V inverter that was mass produced and shipped in 2019. In performance terms, we’ve achieved 6.1dB greater attenuation than the conventional filter structure. That corresponds to roughly a doubling in filter performance, and so we’ve achieved a level of attenuation that would usually have required the addition of another component, without adding a component.
This led us to fit a high-performance filter for 800V and to achieve the world’s first inverter for installation in an 800V vehicle. Motor vehicles are used all over the world, in every country, so the fact that we achieved 800V while meeting the particularly stringent European EMC standard I think has resulted in a contribution to society, which is seeking to change to EVs. I think that result led to the external award. EMC is behind the scenes of EV and inverter development. It’s very rare for awards to shed light on “behind the scenes,” so the achievement was particularly sweet.
Improved performance in non-destructive testing of electronic devices by higher precision noise filtering
Nishimoto: I’m going to talk about research into eliminating noise from ultrasonic receival circuits which transmit and receive ultrasonic waves, and which, while still noise, is different from the electromagnetic wave noise that Ms. Ohmae is talking about. The circuits are one of the core components in the FinSAT7, which is an ultrasonic imaging device that enables non-destructive testing of electronic devices. The FinSAT7 won the 65th Best Ten New Products Award sponsored by the Nikkan Kogyo Shimbun. It was awarded for being a device that contributes to the need for improvement in both testing precision and productivity in the fields of R&D and manufacturing.
The FineSAT7 is a device that creates images and tests the interfaces of laminated wafers and the interior of semiconductor devices. Device interiors can’t be checked by eye or visible light, so testing is done using ultrasonic waves. Ultrasonic waves are transmitted, waves reflected by the device are received, and non-destructively, internal defects are detected and images created. In other words, the device developed that is capable of detecting minute defects is the FineSAT7.
What I developed is noise reducing technology for ultrasonic wave reception signals. You need to reduce noise when receiving reflected ultrasonic waves to detect minute structures and defects. The key is to keep the signals you want and to treat the ones you don’t want as noise, so my research had to start from thinking about the definition of what is a signal and what is noise.
For example, there’s thermal noise (noise from the Brownian motion of electrons), which is generated just by virtue of there being a temperature, and there’s noise that is unique to semiconductor devices used in circuits. A device also has noise. For example, where there is a circuit that sends an ultrasonic wave adjacent to a circuit that receives an ultrasonic wave, you also need to remove noise generated on the transmission side. In my view, the technology to reduce noise is a field of research in which you need to establish technology that reduces noise across an entire device by identifying the multiple causes of noise and understanding the mechanisms that give rise to the noise.
At the same time, while pursuing the objective of improving precision in detecting minute defects, we sought to achieve improvements in detection speed. When you’re seeking to capture a faint signal, you need a lot of data for the same device. We’ve substantially improved the FineSAT7 sampling rate, to 8 Gsps, and we’ve increased the number of bits per piece of data. We succeeded in doing it with the same calculation complexity as conventionally, while also reducing noise. Together with enabling one-time testing of 300 mm (12 inch) diameter wafers, we achieved advances in both the precision and speed of non-destructive testing.
There are all sorts of technologies being used to reduce noise. When the frequencies of a signal and noise are different, the solution is to apply a filter, but you need to investigate which would be more efficient, an analog or a digital filter. When you can’t separate the two by frequency, in the event the flow of current is different for each bit of noise, you can take the analog approach of adjusting the route of the noise to reduce it, or if the noise has a characteristic pattern, the approach is to pattern match and eliminate the noise digitally. If you’re able to reduce the noise of an ultrasonic wave and convert it to image data, after that you could also use image processing to achieve noise reduction.
The problem with these approaches to noise reduction is that you will likely end up at the mercy of one after another noises needing to be addressed, much like the whac-a-mole game. For that reason alone, I feel that it’s important to define and separate signals and noise and that it is possible to have a systematic methodology for achieving noise reduction.
The joy of knowing technology you have developed is generally useful
Ohmae: I think I probably continue to have a very simple motivation for technology, which is just wanting to understand radio waves. I want to see radio waves; I want to experience them. I’m in daily conversation with radio waves, just begging the noise to get out of the way.
Nishimoto: At Hitachi, you can really see that the researchers themselves are engaged in making products. We achieve new performance that’s not yet out there in the world as the result of our research, and that gets delivered to customers. There are all sorts of challenges, not just in the research itself, of course, but in areas other than the research, but when you overcome those and you see a product reach customers that they then use, you really get the sense that your research deliverables are making a contribution to society and that’s a great feeling of achievement. The experience of making a product and handing it over to society isn’t something you can achieve in the academic research field, so probably that sense of achievement is only available because of being a corporate researcher.
Another thing is that we are ourselves users of the FineSAT series of ultrasonic imaging devices that we’re talking about now. The greatest delight in doing research in a company like Hitachi, which has a big range of products, may be that you get to use the products you create through R&D as tools for your further R&D.
Ohmae: Staff at the client motor vehicle manufacturer were just as happy as we were with the electromagnetic noise reduction in the inverter, and they were very complementary. It made me very happy. I was happy to have been able to reduce the noise to at or below the regulated standard, which is my research field, but the fact that ultimately a product was produced is an even greater source of happiness. The highlight was the launch of a motor vehicle actually fitted with the Hitachi inverter and seeing that long line of cars. I wanted to just shout right there that our Hitachi technology was in those cars. I found myself thinking how blessed I am to be a researcher.
OHMAE Aya
Unit Manager
Electronic Systems Research Dept.
Production Engineering and MONOZUKURI Innovation Center
Center for Sustainability
Research and Development Group Hitachi Ltd.
Two books that have supported turning points to change
Several years after I joined the company, as I was coming to feel that the object of electromagnetic noise was changing from analog communications to digital communications, I found a book in a bookshop in the city that I happened to go into, and that was Digital Communications – Fundamentals and Applications (written by Bernard Sklar et al, translated by Morinaga Norihiko et al, Pearson Education). It very kindly and meticulously explains digital communications, from the fundamentals to application, and has been very, very useful in my work since. I feel it was a watershed book for me. The Japanese translation of the original English book is four to five centimeters thick and I’d like to praise the self that I was back then for buying it on the spot, without so much as a blink, when it cost 12,000 yen. The other book is Chaseki kara hirogaru kanshi no sekai (Chinese poem at tea table) (written by Morota Tatsumi, Tankosha), which is a book that gives you a taste of the Chinese poetry that is the basis for language used in the world of tea. I entered the world of engineering after graduating from junior high school and have spent my life without becoming very involved in cultural studies Lately I’ve discovered the pleasures of Chinese tea and I’ve a growing interest in knowing more about culture. I feel that by studying the culture that contextualizes the tea, it makes it even more interesting, and I think that’s common to study generally, including engineering.
NISHIMOTO Takuma
Unit Manager
Electronic Systems Research Dept.
Production Engineering and MONOZUKURI Innovation Center
Center for Sustainability
Research and Development Group Hitachi Ltd.
Staple volume on circuit technology and problem solving classic to accompany research
I’d like to talk about Analysis and Design of Analog Integrated Circuits (written by P.R. Gray and R.G. Meyer et al, translated by Asada Kunihiro et al, Japanese edition published by Baifukan). It’s a book that everyone who gets involved in analog circuit design casts an eye over. One of the authors, R.G. Meyer, is the biggest of the big in analog circuits, and I’ve been studying using this book since I was in the research office at university. The first copy I bought fell apart, and I’m on my second copy. The other book I want to mention is one that has been most useful to me in thinking about how I do my work, How to Solve It (written by G. Polya, Japanese edition published by Maruzen). It’s a book that talks about the process of solving things. It’s an old book, but to the core of my being as a researcher, I don’t think that the way of thinking involved in organizing and solving issues has ever changed.
