IBM Creates Ring Oscillator on a Single Nanotube

deeptrace writes "IBM has combined CMOS circuitry and a single carbon nanotube to implement a 5 stage ring oscillator. Even though the oscillator runs at just 52 MHz, they expect that it could reach the GHz range with improvements. The frequency of the current oscillator was higher than previous circuits using multiple nanotubes. IBM describes the achievement in the paper "Integrated Logic Circuit Assembled on a Single Carbon Nanotube" to be published this week in the journal Science."

A what?

[Eightyford]

What the hell is a ring oscillator, you ask? Well, wikipedia says:

A ring oscillator is a device composed of an odd number of NOT gates whose output oscillates between two voltage levels, representing true and false. The NOT gates, or inverters, are attached in a chain; the output of the last inverter is fed back into the first. The simplest ring oscillator, then, is a single inverter whose output is fed back to itself. Because a single inverter computes the logical NOT of its input, it can be shown that the last output of a chain of an odd number of inverters is the logical NOT of the first input. This final output is asserted a finite amount of time after the first input is asserted; the feedback of this last output to the input causes oscillation.

A circular chain composed of an even number of inverters cannot be used as a ring oscillator; the last output in this case is the same as the input. However, this configuration of inverter feedback can be used as a storage element; it is the basic building block of static random access memory, or SRAM.

A real ring oscillator only requires power to operate; above a certain threshold voltage, oscillations begin spontaneously. To increase the frequency of oscillation, two methods may be used. Firstly, the applied voltage may be increased; this increases both the frequency of the oscillation and the power consumed, which is dissipated as heat. The heat dissipated limits the speed of a given oscillator. Secondly, a smaller ring oscillator may be fabricated; this results in a higher frequency of oscillation given a certain power consumption.

To understand the operation of a ring oscillator, one must first understand gate delay. In a physical device, no gate can switch instantaneously; in a device fabricated with MOSFETs, for example, the gate capacitance must be charged before current can flow between the source and the drain. Thus, the output of every inverter of a ring oscillator changes a finite amount of time after the input has changed. From here, it can be easily seen that adding more inverters to the chain increases the total gate delay, reducing the frequency of oscillation.

Re:Nanotubes..

[Mister Transistor]

Most semiconductors only turn on at a certain voltage level. For example, most silicon transistors turn on at about positive 0.7 volts. Any less than that and the trasistor won't conduct, even if you go below 0 volts to a negative voltage.

What the person was saying about nanotubes is they will "turn on" or begin to conduct again after the voltage drops below 0 to a certain negative level. Kind of like a device that takes the absolute value of the voltage, and if it's above a certain value it conducts or switches "on".

Re:Can you please explain why this is significant?

[eurowombat]

Ring oscillators are simple circuits with which you can easily compare different circuit technolgoies. You simply scale the circuit to whatever your new design rules are, say 90 nm -> 65 nm, soi, etc. and measure the new frequency of the oscillator. This gives you a good base point for measuring and comparing the performance of the new technology.

Re:Explanation?

[jackstack]

FYI - a ring oscillator is just a proof of concept and there is no practical application, per se. It shows that their carbon nanotube transistor technology is well understood enough so that they can make simple logic devices (an oscillator is a bunch of inverters (NOT gate) strung together. Not long ago, slashdot had an article about a transparent ring oscillator from Oregon State Univ. Again, this was done as a stepping stone from discovering an entirely new semiconductor (this is NOT silicon, people) to making a useful device.

Re:Small, and fragile

[Quantum Fizz]

Since nanotubes carry current along the outer surface of the tube, could it be that multiple nanotubes cause the electrical quanta along the surface of each tube to interfere and degrade the signal?

A carbon nanotube (CNT) is a rolled graphene plane (ie, carbon atoms in a hexagonal structure). So of course all current will be on the 'outside' of the tube, as the tube itself really only consists of the outside.

IBM was probaby comparing single-wall nanotubes to multi-wall nanotubes. Multiwall nanotubes are composites of a bunch of concentric single-wall nanotubes. Their better results in the single-wall variety are probably due to less scattering between the graphene planes. A single CNT has a well-defined crystal structure, and is actually quite interesting. The graphene plane itself is sometimes referred to as a 'zero-bandgap insulator', where the density of states linearly goes to zero at the fermi energy (unlike an insulator or semiconductor which has a energy gap at the fermi energy, and hence cannot conduct decently like a metal).

However through changes to the nanotube material, the performance of the nanotube may be impreved.

They probably can get to higher frequencies. I mean, even the vibrational phonon modes of a single nanotube can be in the GHz range or higher (ie, these are the various modes of vibration that the nanotube would exhibit if you struck it, kind of like a wind chime). I don't know specifics, but I don't see why the nanotube couldn't support electronic channels with bandwidths into the GHz or even higher as well.

Although nanotubes do have interesting characteristics different from typical metals and semiconductors. Ie, the electron-phonon interaction goes as 1/T, instead of 1/T^5 (where T is temperature). So at low temperatures there might be useful ways to couple electronic channels to vibrational modes not possible in conventional materials. Or vice versa, the phonon modes might more easily kill off electronic signals. There's alot of interesting work being done with nanotubes, and I'm sure some clever physicists and engineers will exploit these characteristics well in the near future.

Re:Explanation?

[jedZ]

A 5-stage ring oscillator is the hardware equivalent of a program that displays 'Hello World!'

Re:A what?

[suchire]

It's just a negative feedback circuit. If you inhibit yourself, then you stop producing, which stops inhibiting your own inhibition, etc. and this causes oscillation.

This is significant because...

[Spy der Mann]

From TFA:

IBM succeeded in creating a ring oscillator, a test circuit used to evaluate the performance of new materials and semiconductor manufacturing techniques, out of a combination of the CMOS circuitry used by the majority of today's chips and a single carbon nanotube.

OK here's the explanation in 1337:

Carbon nanotubes = t3h w00t
CMOS = reality
Ring oscillator = first tests to integrate t3h w00t into reality

It means that before this, nanotubes and nanotube transistors were only tested in the lab, using microscopic clamps, cables, probes, etc. But this is the first time that a carbon nanotube can be integrated into a working CMOS chip (a small step for chips, a giant leap for mankind). Once CMOS manufacturing can be adjusted for carbon nanotubes, we'll be able to manufacture nanotube memory, nanotube chipsets, and finally, nanotube CPU's!

This is what i've been waiting for since i ever heard about nanotube transistors (however, i think that using graphene sheets instead of nanotubes will be much more effective).

Odd... just did this in class today...

[jpardey]

Lets see if this helps. Some people were confused...

A ring oscillator is a device for making square waves. It uses a common component, a NOT gate. In digital logic, there are two levels, high and low (or 1 and 0, respectivly). High is usually, as far as I have seen, +5 volts, while low is 0 volts (ground).

A NOT gate simply inverts the input. If the value is 1, it outputs 0. If the value is 0, it outputs 1. If the value is somewhere between the two, it will choose one state or the other based on some threshold voltage.

Changing output is not instantaneous. How much time it takes, I don't know. However, it is very fast.

I was going to draw a schematic, but I gave up on appeasing the lameness filter. So, we will use the power of imagination! Imagine one of these NOT gates hooked up to itself. It will switch on and off at a terrific rate. Put a wire on the output, and you have a square wave! Want it slower? Take another two NOT gates, and put them in the loop, so that the first one goes to the second goes to the third. Slower? Another two. If the number of NOT gates was even, the inverted signal would be uninverted by the next NOT gate, which is not what we want.

For more control, one can use a capacitor in a certain arrangment (I'm not looking through my notes). It will take a while to charge and discharge, acting as a delay. Just don't read its voltage as the signal, or you will get a dropping bit, then a rising bit, rather than a nice clean square wave.

Quite useful devices. I hope this clarifies things.

Re:A what?

[Planesdragon]

"It's a type of computer circuit, important to modern PCs."

How's that?

Re:A what?

[Ucklak]

I think it's a type of clock that doesn't need a crystal for the oscillation.

Re:Someones gettin laid tonight...

[andersa]

It's the same at the Niels Bohr Institute. 'Nano' definitely has the hottest chicks. They are almost in the same league as the biochemistry department!

Re:Small, and fragile

[PhysSurfer]

All nanotubes are made up of this graphene plane.

Actually, you can make nanotubes out of other materials besides carbon. Metallic nanotubes, for example, will have different crystal structures than the graphene hexagon.

A tube with 100 atoms will have 100 distinct oscillating modes.

No, it will have 300, one for each degree of freedom. However, three of these will be translational modes, which are not phonon modes, so really there will be 297 distinct phonon branches. In addition you should distinguish between the number of atoms in a Carbon Nanotube, and the number of atoms in its unit cell. A unit cell may have 100 atoms, but the entire nanotube can be made of 1000s of unit cells. The number of atoms in the unit cell is the important number for calculating phonons.