MUNICH, Germany – An electric potential sensor that promises to be highly disruptive in multiple markets and that has been in development at the University of Sussex in England for more than eight years, is set to be manufactured and commercialized by Plessey Semiconductors Ltd.
Plessey (Roborough, England) has agreed technology license terms with the University of Sussex, although Michael LeGoff, CEO of Plessey, declined to reveal the details or the nature of its exclusivity at a press conference a the Electronica exhibition here. Plessey is a recently-formed company that has taken over the products, IP and a wafer fabrication facility formerly associated with the name.
The sensor can be used to measure the electric potential without drawing current according to the developer Professor Robert Prance of the university's Centre for Physical Electronics and Quantum Technology. "It's the almost perfect voltmeter. Electrically non-invasive and with minimal field disturbance," said Prance.
The main feature of the sensor is the electronically-enhanced input impedance achieved by the use of feedback techniques leading to input impedances as high as 10^17 ohms at 1-Hz. These sensors can then function as voltmeters for ac signals from various sources provided the input impedance is much larger than the sensor-source impedance.
Sensors made by the university research team have demonstrated a sensitivity of microvolts per meter and an accuracy of 2 percent.
I agree with "Sharps", and I would offer that in addition to sounding quite a bit like the analog "bootstrapping" technique, which I saw explained in the late 1960's as a means of improving transistor circuit input impedances, charge amplifier circuits have also been around for quite a while.
Of course, if they have come up with a means of rejecting the ambient background field well enough to deliver the described performance, that is quite an accomplishment indeed, and very worthy of acclaim.
But I am more likely to consider the development in light of the Heisenberg uncertainty principle, which reduces to the assertion that "measuring a quantity has an effect on that quantity", and this usually introduces some error.
Fundamental noise limitations from a small-geometry FET will limit the effectiveness, feedback techniques or not. In particular higher bandwidths will be difficult. However, that's not to say it couldn't be useful, particularly in arrays with plenty of signal processing. And the reproducibility as an integrated part will be greatly advantageous compared to discrete realizations.
I had a recent occasion to be reminded of the prevalence of electric fields in the local environment, when the aluminum can housing an electret "condenser" mic capsule went open-circuit and ceased to shield the input circuit, which typically has a local gate resistor of a few hundred megohms and a small JFET, which pretty much sets the low-frequency electrical response cutoff. The result was a loud hum/buzz in the audio. I use the mic to remotely monitor the sounds at the front door, so that I don't miss a mail delivery or similar.
Hi Dr Quine
I was fortunate enough to see an impressive demonstration of the technology at a press event at Electronica. It would seem that in "remote" sensing it might be possible to "confuse" the sensor....so muscles firing could mask the presence of heart beat for example. Professor Prance actually demonstrated that while holding a sensor in each hand.
But the ability to create 1- and 2-D arrays of sensors and intelligent application-specific systems architectures should get round most problems.
As to price; as Plessey has yet to do any of the monolithic or monopackage integration I am sure their executives would say it is too early to give an indication on price. But fundamentally i don't see any reason why the sensor itself should cost more than a few dollars/pounds/euros.
But if you are replacing expensive medical monitoring equipment it would be natural to want to price to value created rather than to the cost of manufacture!
In sports and human-machine interface applications the prices may have to be lower anyway.
This sounds like a refinement of analog 'bootstrapping' to raise apparent input impedance. Effectiveness depends on how accurately the incoming signal can be modelled in the feedback circuit (so as to match it and oppose it) and also how well interfering signals can be nulled out. No doubt the research work has worthwhile applications, but I hope this isn't a lot of marketing types getting over-excited about a relatively incremental development?
This new technology sounds like it has "electrical potential". How will the sensors distinguish the voltage of interest from other ambient voltages in remote sensing applications? It sounds like these sensors are inexpensive enough to be used to monitor multiple subsystems. What is the expected price point?
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.