Redefining electrical current law with the transistor laser

Redefining electrical current law with the transistor laser

http://news.illinois.edu/WebsandThum...holonyak_x.jpg
Photo by
L. Brian Stauffer

---

A major current law has been rewritten thanks to the three-port transistor laser, developed by Milton Feng and Nick Holonyak Jr. at the University of Illinois.




5/12/10 | Liz Ahlberg, Physical Sciences Editor | 217-244-1073; eahlberg@illinois.edu

CHAMPAIGN, Ill. — While the laws of physics weren’t made to be broken, sometimes they need revision. A major current law has been rewritten thanks to the three-port transistor laser, developed by Milton Feng and Nick Holonyak Jr. at the University of Illinois.


With the transistor laser, researchers can explore the behavior of photons, electrons and semiconductors. The device could shape the future of high-speed signal processing, integrated circuits, optical communications, supercomputing and other applications. However, harnessing these capabilities hinges on a clear understanding of the physics of the device, and data the transistor laser generated did not fit neatly within established circuit laws governing electrical currents.


“We were puzzled,” said Feng, the Holonyak Chair Professor of Electrical and Computer Engineering. “How did that work? Is it violating Kirchhoff’s law? How can the law accommodate a further output signal, a photon or optical signal?”


Kirchhoff’s current law, described by Gustav Kirchhoff in 1845, states charge input at a node is equal to the charge output. In other words, all the electrical energy going in must go out again. On a basic bipolar transistor, with ports for electrical input and output, the law applies straightforwardly. The transistor laser adds a third port for optical output, emitting light.


This posed a conundrum for researchers working with the laser: How were they to apply the laws of conservation of charge and conservation of energy with two forms of energy output?


“The optical signal is connected and related to the electrical signals, but until now it’s been dismissed in a transistor,” said Holonyak, the John Bardeen Chair Professor of Electrical and Computer Engineering and Physics at the U. of I. “Kirchhoff’s law takes care of balancing the charge, but it doesn’t take care of balancing the energies. The question is, how do you put it all together, and represent it in circuit language?”


The unique properties of the transistor laser required Holonyak, Feng and graduate student Han Wui Then to re-examine and modify the law to account for photon particles as well as electrons, effectively expanding it from a current law to a current-energy law. They published their model and supporting data in the Journal of Applied Physics, available online May 10.


“The previous law had to do with the particles – electrons coming out at a given point. But it was never about energy conservation as it was normally known and used,” Feng said. “This is the first time we see how energy is involved in the conservation process.”


Simulations based on the modified law fit data collected from the transistor laser, allowing researchers to predict the bandwidth, speed and other properties for integrated circuits, according to Feng. With accurate simulations, the team can continue exploring applications in integrated circuits and supercomputing.


“This fits so well, it’s amazing,” Feng said. “The microwave transistor laser model is very accurate for predicting frequency-dependent electrical and optical properties. The experimental data are very convincing.”
The Army Research Office supported this work.

As a matter of fact, I've read quite a bit of his stuff, as well has communicated with him personally several times in the past. Most of my communication was related to something called "Tipler cylinders" and related also to the alleged time traveler, John Titor.

Actually, I predicted in classes I used to teach many years ago that a new sort of computer would be 'invented' by the 10-12 year of the new century and it would use laser technology in conjunction with a new type of transistor logic which will have more than 2 logical states....

Appears I hit the nail on the head.

i adore physics. love it.

i have to ask though, what was the hard part in this discovery? figuring out a way to physically put laser technology into one of these nodes?

if you are curious as to why quantum physics is such an interesting thing, watch these presentations:

http://quantumiscool1.ytmnd.com/
http://quantamiscool2.ytmnd.com/
http://yqpic3.ytmnd.com/
http://wqpic4.ytmnd.com/

don't let that site/domain fool you. the guy who created the presentations was a physics student at texas a&m.

Lasers aren't that difficult to make. They are damned difficult to shrink!

I can build a laser from crap in my junk boxes. Wouldn't be a great one, but, it would work. I just can't turn it into a tiny transistor that is a few microns across.

so it's the laser technology that made this possible? see where i'm getting at? what was the catalyst of this discovery? the laser technology or the idea to use lasers?

Quote:

---

so it's the laser technology that made this possible? see where i'm getting at? what was the catalyst of this discovery? the laser technology or the idea to use lasers?

---

Speed.

That's the catalyst.

Students asked me about the upper limit of speed for computer systems when they were hitting about 200 Mhz. I said the upper limit was currently based on size of the technology at the time.

Basically a computer is a high frequency radio system. There is a clock built in that runs, usually somewhere in the Megahertz range, then there are step up devices to speed the clock. The clocking system in a computer is what makes the digits count, the flip-flops flip, registers push out or hold their data.

Without a way to make the counter click, the computer is worthless.

The clock is the heart and soul of a souless machine and without it, completely dysfunctional.

So back in those days, given the size of the microprocessors and the chip sizes, and even transistor sizes (on the micron scale inside a processor and memory chips) you have an issue with propagation within a system.

The BIGGER the system, the SLOWER it works (even on a semi-quantum scale).

My students thought that there could be no limit on the top speed for a computer (clock) because you can do radios that are (i nthose days) in the gigahertz range.

I explained that because computers use current, electron movement and radio frequecy clocking speeds the upper limit for (the current day's technology) was not much higher that 200, perhaps 300 Mhz. It took some time, if I recall to break from 200 to 300 Mhz.

Even today, CLOCK speeds are not all that high and computers are running around 2 gigahertz processors (unless they are what is called "over clocking". Did the same thing with a 6800 microprocessor back in the day by changing out the crystal from 3.5mhz to a 6.2 mhz crystal. The actual clock speed was significantly less than either of those though). Point being, they still have a barrier to break and they can't do it on a current/voltage level.

LIGHT waves however, are not exactly the same as electrical current. They are similar, but different.

So, breaking the barrier of speed is based on necessity - and thus the idea to use laser systems. I didn't come up with the idea (well, I did independently of anyone else, but others thought of it too) of using a laser as a method to transmit data inside of processors and memory chips... however, it's been a long considered idea.

They simply have taken the better part of twenty years to come up with a way to do it.

i am definitely familiar with pc's and clock speeds, however, i was under the impression that clock speed had 110% everything to do with how "cold" a processor functioned when electricity was passed through it. being how small cpu's transistors are, there will always be microscopic imperfections in it's making. therefore if Intel or AMD is making 3ghz chips, they will end up with 2ghz, 2.5ghz, and 3ghz's. this is because in order for the chip to operate without errors, it produces a certain amount of heat. and if the cpu cooler cannot keep the chip cool, then the bios will downclock the CPU so the fan can keep the cpu at a certain temp. and since electricity travels in waves, it's measured in hz.

am i off on anything here?

the heat is generated due to the number of devices operating within.

A simple processor has thousands and thousands of transistor devices.

More modern, more complex processors have millions, if not billions of them.

The more devices internally in a chip the more power it takes to run them. Those devices use current, and heat energy (just like a light bulb) is generated.

the more heat there is the harder for the device to operate due to internal impedance changes.

All things have either a positive or negative temperature coefficient. The hotter something gets usually, the higher its resistance (in general, but not always) gets, thus causing higher voltages to push the same amount of current... and so on.

Semiconductors devices (diodes, transistors and other devices used internally in processors) can get runaway problems, overheat and burn themselves out.

Also, heat causes drifts in frequencies in radios, and even in the clocking systems of computers. Thus more heat, causes drift, when can cause a clock to out of tolerance and beyond or below what the processor needs to actually work.

There are limits in both directions. And heat can cause shut down to total failure.

that was my thinking.

however, i'm not making the connection between that discussion and this thread (possibly nothing lol).

It's simpler than this.

Just passing data more quickly using a photon instead of an electron is the key. Nothing to do with the "state" (like 1 or 0)