Maker Pro
Maker Pro

Advantages of photonics -- other than bandwidth -- over electronics

L

Lostgallifreyan

Jan 1, 1970
0
Kai-Martin said:
How does "Light Amplification by Stimulated Emission of Light" exclude
modulation?



Sure.
Also, diode lasers are notorious for modulating all by them self.
Most TiSa Lasers can be modulated by mirrors sitting on piezo actuators.
Knock on the housing of a running gas laser and its output frequency jerks.
Makes me wonder if there is any type of laser that can not be modulated.

You can even keep the laser untouched and modulate the beam itself. The
gadgets to do this are called EOM and AOM. These are acronyms for "electro
optic modulator" and "acousto optic modulator".


I'd thought to mention AOM (PCAOM) but didn't because I was out of my depth
there. :) Although they are frequency controlled, do they not select or
attenuate existing wavelengths of the light sent into them, rather than
modulate its frequency? If they really do frequency modulation, they would be
maybe the most interesting devices I've heard of in the last ten years or more,
and I seriously need to do some reading...

About jumps of frequency in gas lasers, that might my a momentary Doppler
thing? I'm thinking that would readjust as soon as it could, to whatever the
lasing medium wants to put out.

Agreed about diode lasers. Another poster mentioned that too. I'd considered
their temperature dependence before I saw either of these posts, but I decided
that this was so slow I couldn't call it modulation, as such.
 
J

John Larkin

Jan 1, 1970
0
Just play with a fiberoptics link for a while and you may want to
reconsider those statements. A modulated-laser fiber link is a
nightmare of mode jumps, unpredictable connector behavior, reflections
and interferance all over the place, inconsistant part behavior, nasty
temperature effects, losses everywhere, and big bucks. And LED fiber
links are pigs. The emi resistance part is right, but the rest is
mostly speculation.

Besides, any photonic system is electronic on both ends.

John
 
D

David

Jan 1, 1970
0
Let me know when they can be ordered from the local electronics
distributor.

Until then, the statement stands.

Jim

what is being refered to is an optical isolator, and it is pretty much
a catalogue/off the shelf item, available from the likes of Thorlabs,
Newport, New Focus, Melles Griott, etc.as well as numerous other
distributors,
Just because Radium doesn't really know what he is talking about is no
reason to assume the opposite point of view.

Photonic integrated circuits are also a pretty standard term and class
of component in the industry, they just have a slightly different
definition to what you may associate with it. Any integration of, for
example, coupler -attenuator- WDM, on the same substrate is often
refered to in this manner.

cheers

David
 
P

Phil Hobbs

Jan 1, 1970
0
Let me know when they can be ordered from the local electronics
distributor.

Until then, the statement stands.

IBM, Luxtera and British Aerospace, as well as a number of other
companies and many universities, are working on "photonic circuits".
That's my principal research area at the moment. It's important to
distinguish between optical logic and optical communication. Optical
logic will IMO never be competitive with electronics. You can control
many electrons with one electron in a transistor, but you can't control
many photons with one photon in any photonic device whatsoever.
Transistor speeds are advancing steadily, and although the technology
gets hairier and hairier, there seem to be no fundamental limits until
you get into the low tens of terahertz, where semiconductors stop
looking like electrical conductors. (This is due to the plasma
resonance--at any finite carrier density, you can only propagate fields
up to a certain frequency.) Other limits may obtrude sooner, e.g. the
statistical distribution of dopant atoms in very small devices means
that threshold voltages become very hard to control, but ways are being
found to live with these.

It's true that wire has a lot of advantages over on-chip optics at
ordinary speeds--routability, ease of fabrication, loose temperature
tolerances, you name it. Wire is great, and there are a lot of smart
people working very hard (with lots of money behind them) to make it
even better, mainly by reducing the signalling voltage. People know
about wire, wire doesn't generate many midnight phone calls, and you can
sleep pretty well after betting your $3bn fab investment on the
proposition that wire will continue to work.

The problem is that there appear to be hard limits to what you can do
with wire, in terms of bandwidth vs distance, signal integrity, and
power consumption. Optical communication can probably be brought down
to the level of 100 fJ per bit (100 uW/(Gb/s)) and possibly quite a bit
lower, whereas the 2 pF/cm capacitance of wire isn't going to improve by
orders of magnitude, ever. Putting dense optical communications and
optical switching on a chip is what's usually meant by a "photonic
circuit", and it appears that they'll be a mainstream technology in 10
of 15 years.


Cheers,

Phil Hobbs
 
L

Lostgallifreyan

Jan 1, 1970
0
Phil Hobbs said:
IBM, Luxtera and British Aerospace, as well as a number of other
companies and many universities, are working on "photonic circuits".
That's my principal research area at the moment. It's important to
distinguish between optical logic and optical communication. Optical
logic will IMO never be competitive with electronics. You can control
many electrons with one electron in a transistor, but you can't control
many photons with one photon in any photonic device whatsoever.

<snipped>

As far as I know, this is because the electric feilds of passing photons don't
interact as they'd have to, though maybe I'm not thinking right there..

Anyway, I'm wondering if it might be found to work if a suitable solid (or gel,
liquid, whatever..) were devised or found. If a very small beam of light at one
wavelength could change a material in a way that makes it block the passage of
a much stronger beam at another wavelength, you could do switching, and maybe
even fast proportional (linear) modulation of amplitude, with no moving parts.
A kind of photonic field effect transistor. If any kind of phtonic transistor
were made, and if I was a betting man at all, that would be where I'd put my
money.
 
C

Charles Manoras

Jan 1, 1970
0
Lostgallifreyan said:
Anyway, I'm wondering if it might be found to work if a suitable solid
(or gel, liquid, whatever..) were devised or found. If a very small beam
of light at one wavelength could change a material in a way that makes
it block the passage of a much stronger beam at another wavelength,
you could do switching, and maybe even fast proportional (linear)
modulation of amplitude, with no moving parts.
A kind of photonic field effect transistor. If any kind of phtonic
transistor were made, and if I was a betting man at all, that would
be where I'd put my money.

Don't. Very speculative to say the least ;-)

The features of the best photolithographically produced electronic
components are now well below the lambda of the light used to
produce them and getting even smaller all the time.

That's a true miracle and yet the "optical computer", this nearly fifty
year old bugaboo is not yet dead, in spite of the overwhelming evidence
that it is unfeasible.

(not unfeasible if you are a DoD grant distributing bureaucrat or a
grant proposal writing academic of course, however).

Do you think you get your highly hypothetical photonic gadgets to work
similarly with features as minute (and therefore be actually competitive)?

Even with a very "small beam of light"?
 
Radium said:
One advantage of photonics over electronics is more bandwidth but that
is only one of several advantages. Others are greater clarity, less
overheating, more resistance to EMI/ RFI, greater S/N ratio.

So its not just bandwidth.

People in the groups I post to keep telling me that the only advantage
of photonics over electronics is bandwidth. That is so wrong.

Sufficiently advanced photonics will be able to use much smaller
structures for logic and storage. Electronics has an absolute
structural limit defined by the Pauli exclusion principle. Though
we're a couple of paradigm shifts away from reaching it, and not really
close to the first one.

--Blair
 
R

Radium

Jan 1, 1970
0
Sufficiently advanced photonics will be able to use much smaller
structures for logic and storage.

Will they still need to be powered by electricity?
 
Radium said:
Hi:

One advantage of photonics over electronics is more bandwidth but that
is only one of several advantages. Others are greater clarity, less
overheating, more resistance to EMI/ RFI, greater S/N ratio.

Even a small radar dish has more bandwidth, EMI resistance
and SNR than fiber. And it's getting to the proce rannge now
where you can open up your own studio foe the price
of one fiber tech.
So its not just bandwidth.

People in the groups I post to keep telling me that the only advantage
of photonics over electronics is bandwidth. That is so wrong.

bandwidth is the stanard strawman in all signal processing.
BT is the relative parameter. And electronics
has so much BT, it' will be only be another twenty years and
it'll put half of the US Navy and Air Force out of buisnees
 
L

Lostgallifreyan

Jan 1, 1970
0
Charles Manoras said:
Do you think you get your highly hypothetical photonic gadgets to work
similarly with features as minute (and therefore be actually competitive)?

Even with a very "small beam of light"?

No. But that's not what I had in mind. My thinking is, that it points to a
door. We can't expect a new door to find the same path, certainly not a better
one on the same terms. My point is that the asking of crazy, ill-informed
questions oftens points at doors to truly new things. The effort to make
something in which one small quantity of light directly affects a material in
such a way that it rapidly and proportionally modulates a larger beam's
amplitude isn't (shouldn't be) a bid to emulate a successful technology with an
inferior method, it should be an interesting experiment. First, you make the
thing, if it can be done, then you play with it to see what else it might do,
and look around for problems that might be solved by its behaviour, once you
have it built and understood. Most truly new thiongs happened this way, so
instead of betting on it, maybe we should be putting serious investment into it.

If nothing else, the effort might bring a bit of linear behaviour to photonics,
which as other posters have indicated, could sorely use it.


PS. Meant 'weak', not small. :) Usenet is not renowned for allowing edits..
 
Let me know when they can be ordered from the local electronics
distributor.

You mean local photonics distributor.
Until then, the statement stands.

Opto-isolators are photonics. They do something electronics can't.

Solar cells, CdS sensors...etc.

Fiber optics, CD/DVD discs, LEDs (e.g., IR controllers) etc.

Lenses, etc.

I don't know why anyone's arguing this stuff. Photonics are
everywhere.

--Blair
 
B

Boris Mohar

Jan 1, 1970
0
There are no practical optics capable of being frequency modulated.

Lasers by defininition can't be frequency modulated.
Laser diode frequency varies with current. So does the amplitude. When
you modulate a laser diode by modulating the current you get bot FM and AM.
Look up two tone FM spectroscopy.



Regards,

Boris Mohar

Got Knock? - see:
Viatrack Printed Circuit Designs (among other things) http://www.viatrack.ca

void _-void-_ in the obvious place
 
B

Boris Mohar

Jan 1, 1970
0
You mean local photonics distributor.


Opto-isolators are photonics. They do something electronics can't.

Solar cells, CdS sensors...etc.

Fiber optics, CD/DVD discs, LEDs (e.g., IR controllers) etc.

Lenses, etc.

I don't know why anyone's arguing this stuff. Photonics are
everywhere.

--Blair

Every month I get Photonic Spectra magazine which is a free subscription.
This month had an interesting article on photonic clocking.
Link to abstract:
http://www.photonics.com/content/spectra/2006/June/features/82894.aspx



Regards,

Boris Mohar

Got Knock? - see:
Viatrack Printed Circuit Designs (among other things) http://www.viatrack.ca

void _-void-_ in the obvious place
 
Top