Light Emitting Transistor

From the University of Illinois at Urbana-Champaign, the light emitting transistor... /ubbthreads/images/graemlins/cool.gif

Here's another article with a better picture...

Does that mean that the LED can be it's own output transistor in a boost circuit now? /ubbthreads/images/graemlins/grin.gif

Seems I remember reading about that in Compound Semiconductor magazine (the paper version) not too long ago.

Actually Harvard was before them back in October 2003:

http://adsabs.harvard.edu/cgi-bin/nph-bib_query?2003PhRvL..91o7406H

Abstract
We report the first organic light-emitting field-effect transistor. The device structure comprises interdigitated gold source and drain electrodes on a Si/SiO2 substrate. A polycrystalline tetracene thin film is vacuum sublimated on the substrate forming the active layer of the device. Both holes and electrons are injected from the gold contacts into this layer leading to electroluminescence from the tetracene. The output characteristics, transfer characteristics, and the optical emission properties of the device are reported. A possible mechanism for electron injection is suggested.

Also see the light emitting transistor from 12 years ago:
Applied Physics Letters Vol 61(9) pp. 1051-1053. August 31, 1992

http://content.aip.org/APPLAB/v61/i9/1051_1.html

Or this 12 year old one from AT&T:
http://www.ee.sunysb.edu/~serge/104.pdf


http://www.khlim.be/~jgenoe/PUBL/APL01051.pdf


Dream on James.

Anyhow, saw thaT in EETimes a nearly two months ago, interesting, I know that :

http://www.eetimes.com/at/oe/news/OEG20040106S0017

http://www.eetimes.com/story/OEG20040106S0017

Hard copy had a nice photo

Great description from that abstract; thanks. So Harvard's was an organic FET (FELET?) and I gather that the unit from Urbana was a bipolar and non-organic construction...? I would guess that the FET version would have a lower maximum frequency due to capacitive effects across the high impedance input.

None of these are intended for general illumination, as far as I can tell;I think they're primarily aimed at optoelectronic interconnects within and between boards, where the maximum frequency is more important than the brightness.

No, MOSFETs are quite quick.

These are very common parts, go to 1GHz no problem.

http://www.macom.com/data/datasheet/LF2810A.pdf
http://www.semiconductors.philips.com/acrobat/datasheets/BLL1214-250_3.pdf

Also pushing the high-speed possibilities of CMOS will be the FinFET, a quasi-planar double-gate MOSFET. The FinFET structure can be fabricated in a conventional CMOS process and can be inserted wherever needed with ordinary design tools. Because of its novel design, the Omega FinFET operating at 0.7 volts nevertheless achieves a gate delay of 0.39 ps for the N-ch MOSFET making it among the fastest devices ever reported at this gate length.

You may start seeing the FinFET in your processor soon, to replace the slow and leaky CMOS MOSFET transistors inside of it, they are looking to employ them when they get to the 65-45nm node:

http://www.semiconductorfabtech.com/industry.news/2002/10902.shtml

CMOS is Complementary Metal Oxide Semiconductor, which means you have both n-ch and p-ch MOSFETs.

More on FinFETs: http://www.eetimes.com/semi/news/OEG20021126S0035
http://www.geek.com/news/geeknews/2003Jul/bch20030718020888.htm

If your processor was full of bipolars, you'd really be sucking some serious juice!

Thanks for the clarification on MOSFETs and speed. My original opinion was biased (reverse-biased?) by my experiences with the old MPF102.

[ QUOTE ]
If your processor was full of bipolars, you'd really be sucking some serious juice!

[/ QUOTE ]

Ohhhhhh, yeah! I remember an old character generator that was the size of a small refrigerator, built around TTL, and required something like 20 amps just for the +5v logic! Going further back I also remember RTL, which really pulled some serious current! /ubbthreads/images/graemlins/eek.gif