| Author | Message | |||
     Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 453 Registered: 9-2001 |
Busan - 100 nm is more than an order of magnitude smaller than the wavelengths we use in the NIR. I don't really know how it will behave optically but I suspect you'll have a REALLY hard time getting any light into the fiber. To my best knowledge, fibers with such small diameters are what are known as "single-mode" fibers, that is, they will only transmit one mode of light through the fiber. The advantage of that is that the light, which is invariably from a laser, can be modulated at very high speeds, so high that the fiber may be longer than the length of even one single laser pulse in it. These are therefore used for very high-speed (e.g., gigabytes per second and higher) data transmission, the sort of fibers used by the Internet and the telephone companies for cross-continental transmissions. In this sort of application, the multiple laser pulses, representing the data bits, follow each other like trains on a track. Larger fibers would allow the pulses to spread out within the fiber, eventually overlap and you would lose the ability to tell one pulse from another. This would limit the rate at which the pulses could be sent. I would certainly be interested to hear what results you can obtain when using it for spectroscopic application. Howard \o/ /_\ | |||
| Dusan Kojic (dkojic) New member Username: dkojic Post Number: 2 Registered: 7-2011 |
I heard that hamamatsu managed to make an optical fiber with a diameter of 100 nm. My boss went to their factory and got one of those. I still haven't seen it so I haven't worked with it in order to be able to provide you with a more precise info. I also don't know the details of hamamatsu product but for purpose of collimating light that sounds like a pretty good solution. If Jerry could go to their site I think he should be able to dig out something. Regards, Dusan | |||
| Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 452 Registered: 9-2001 |
Jerry - A long time (35-40 years) ago I once had a chance to see and handle a FO plate like that. At the time, I was looking at it solely for its "gee-whiz" properties and had no real "use" for it then. The one I had was about 1 inch thick, as I recall. The data sheet doesn't say much about the thickness, although they mention that the thickness can be from 1 - 15 mm. Note the statements on the data sheet saying that light beyond the numerical aperture (the maximum angle the fiber will accept and transmit) of the fiber is lost. This is why energy is lost when using these plates, and is also the mechanism for increasing the collimation. As shown on the first page of the data sheet, a fiber plate can transfer an image from one face to the other. But if you don't need an image, you can get the same effect simply by passing light through a straight, narrow tube. Light entering the tube at an angle will be absorbed and, since the tube is not a fiber optic, any residual reflected light will be completely absorbed after multiple reflections down the tube, leaving only the light poassing straight through to exit the other end. You can adjust the degree of collimation by varying the length and diameter of the tube. \o/ /_\ | |||
| Jerry Jin (jcg2000) Senior Member Username: jcg2000 Post Number: 45 Registered: 1-2009 |
Howard and Andrew: I am grateful of your advice. There is commercial product of the optical part you mentioned "Fiber Optic Plate (FOP) is used to transfer an image without using a lens. They consist of a large number of optical fibers bundled together." Here is one from Hamamatsu: http://sales.hamamatsu.com/assets/pdf/catsandguides/FOP_TMCP1005E03.pdf It looks this part can make diffusive light becomes "bundled together". It is not perfect collimation, but at least a step closer to collimation. Anyone used this fiber optical plate before? Thanks. Jerry Jin | |||
| Andrew McGlone (mcglone) Advanced Member Username: mcglone Post Number: 25 Registered: 2-2001 |
After a very quick search on the internet, I find references to the use of such devices for coupling small fibres together, for instance. In fact, a little embarrasingly, I'm also directed back to first book I ever read on NIR, Brian Osborne's Practical NIR Spectroscopy (1986) where some examples of use are given. I'm obviously 25 years behind the game with this notion... I also spot some japanese researchers in 2002 exploring them for coupling light out of a 1mm fibre into micron sized pin diodes. Evidently far more accomodating than a ball lens, they say. Run that idea in reverse and its interesting, don't you think? Just like your suggestion of a parallel array of fibres, you lose light, yes, but increase directionality for what you do get through. I don't know, because my scheme must expands the area on exit, I'm tempted to think your scheme of parallel fibres might be better. Hmmm, its all rather irrelevant to Jerry if he wants collimated light... | |||
| Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 451 Registered: 9-2001 |
Andrew - those concentrators are designed to maximize the total energy impinging on the target, not to give the beam any special characteristics, such as being focussed or collimated. Therefore the design criteria are different. Nevertheless, all physical systems are constrained by the laws of thermodynamics. In the case of optics, that constraint is manifested in what is known as "The Equation of Radiative Transfer". The Equation of Radiative Transfer ultimately comes from the First Law, and is the result of expressing mathematically the fact that in the absence of scattering or absorption of the light, the optical power in the beam must remain constant regardless of how much it's reflected or refracted. It's actually fairly simple to write down: P = e * A * w * d<lambda> where: P is the power transmitted e is the optical intensity A is the area of the beam w is the solid angle of the beam d<lamba> is the bandwidth P, as I said, is constant by virtue of the First Law. e is also constant in any given optical beam. A and w are the geometric factors, and in an optical system can be traded off by the use of mirrors and lenses, to make the light do what you want. In general, however, when you focus light, you're making the light fall on a small spot, and at that spot the light subtends a larger solid angle than before it was focussed. None of the terms in the equation of radiative transfer can be zero, or the energy of the beam will also be zero. Therefore even a laser beam, which we normally think of as being collimated, still have to have some small amount of divergence, and the intensity is very high because of that. \o/ /_\ | |||
| Andrew McGlone (mcglone) Advanced Member Username: mcglone Post Number: 24 Registered: 2-2001 |
Yeah, my scheme certainly can't collimate, just increases directionality for some of it. That means you certainly couldn't focus it, theoretically only back play it at best, so your gedankenexperiment doesn't get a play at all. (and Howard, it was your mention of reducing solid angle with fibres that made me think about it). Still I see those sorts of concentrators on solar cells, accepting light from a wide sky angle and think the reverse would be interesting, might be useful somewhere... However its not like a lens at all, you can't image with them or similar. | |||
| Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 450 Registered: 9-2001 |
Andrew - your scheme would run afoul of the Second Law for the same reason Jerry can't collimate the light. Here's the gedankenexperiment that illustrates the problem (as in all gedankenexperiments, everything is assumed to be ideal): Assume you can do that. Then you could take a (ideal) mirror and arrange the optics so that the point where the light is concentrated falls on (or in Jerry's case, you could focus the collimated light onto) a spot on the source of the diffuse light. Then that spot would be receiving more energy than it was emitting, and would therefore experience an increase in temperature over the rest of the source. You would therefore have created a temperature difference without the expenditure of any work. This violates the Second Law. Therefore, since we can't violate the Second Law, we can't concentrate the light the way you describe, nor can we collimate the diffuse light. \o/ /_\ | |||
| Andrew McGlone (mcglone) Advanced Member Username: mcglone Post Number: 23 Registered: 2-2001 |
I've wondered in the past, for capturing scattered light into a fibre, about running a compound parabolic concentrator in reverse. Would increase directionality although it spreads the beam spatially too. So pretty doubtful for your 'tiny' setup but might spur your thinking onwards. | |||
| Howard Mark (hlmark) Senior Member Username: hlmark Post Number: 449 Registered: 9-2001 |
Jerry - what you want to do flies in the face of the second law of thermodynamics, I'm afraid. Diffuse light cannot be collimated unless the source is so small that it can be considered a point source, and then the collimating lens would have to be large compared to that "point source". To some extent, however, you can reduce the solid angle of the output light by running it through an optical fiber. It's possible to make a plate of parallel optical fibers close-packed and glued together (I think such a device might even be available commercially), for which the spread of the output beam is limited by the angular acceptance of the fibers used. The output beam will then not be as energetic as the input beam, also limited by the second law. I see no way to do better than that. \o/ /_\ | |||
| Jerry Jin (jcg2000) Senior Member Username: jcg2000 Post Number: 44 Registered: 1-2009 |
Dear All, I am trying to make highly diffuse light become directional.The idea is illustrated in the attached figure. I also want the part to be tiny,so a lens won't work here. I am not sure a fiber optical plate would be my solution. Please advise if you have any idea. Thanks Jerry Jin
|
collimating plate