Killing a laser's temporal coherence?
Joe Fikes
gas laser? I'm trying to image a laser-illuminated target with a line-scan CCD
camera, and the speckle is driving us nuts.
Thanks.
Joe
--
My return address contains two minus signs (-) in a
(so far ineffective) attempt to thwart spammers. To get me
for real, delete both minus signs in my return address.
Alan Shinn
--
I have used rotating or vibrating plastic ziplock bag material
(slightly foggy) in the beam. Always worked well for me.
Looking forward:
Alan Shinn
Experience the
beginnings of microscopy.
Make or buy your own replica
of one of Antony van Leeuwenhoek's microscopes.
visit http://www.sirius.com/~alshinn/
Don Stauffer
>
> Does anyone know of an effective way to kill the temporal coherence in a diode or
> gas laser? I'm trying to image a laser-illuminated target with a line-scan CCD
> camera, and the speckle is driving us nuts.
>
> Thanks.
>
> Joe
>
> --
> My return address contains two minus signs (-) in a
> (so far ineffective) attempt to thwart spammers. To get me
> for real, delete both minus signs in my return address.
Don't laugh, but some people send beam through a small tank of
milk/water mixture!
--
Don Stauffer in Minneapolis
home web site- http://home1.gte.net/stauffer/
home email- stau...@gte.net
work email- stau...@htc.honeywell.com
Andy Resnick
> Does anyone know of an effective way to kill the temporal coherence in a diode or
> gas laser? I'm trying to image a laser-illuminated target with a line-scan CCD
> camera, and the speckle is driving us nuts.
>
> Thanks.
>
> Joe
>
> --
> My return address contains two minus signs (-) in a
> (so far ineffective) attempt to thwart spammers. To get me
> for real, delete both minus signs in my return address.
A slight correction here: the speckle is caused by spatial coherence rather than
temporal coherence. That being said, a piece of ground glass would solve your
problem nicely. Some plastics have strains and structure present that can cause
unacceptable amounts of diffraction.
Andy Resnick
Gerald W. Gordon
optical fiber and shake the middle of the fiber.
Jerry
Gor...@med.unc.edu
Helge Nareid
<stau...@htc.honeywell.com> wrote:
>Don't laugh, but some people send beam through a small tank of
>milk/water mixture!
Done that - it works quite well, the main problem is finding the right
mixture ratio to prevent excessive diffusion. A drop of milk and a
decent cuvette is normally fairly easy to find if you need a quick way
to remove speckle. A moving diffuser is another good method - a
rotating ground glass screen works just fine. It is more suitable for
a more permanent set-up - you don't have to worry about the milk going
off.
--
- Helge Nareid
Nordmann i utlendighet, Aberdeen, Scotland
David L. Rosen
spatial coherence very much, use a depolarizer. However, I can swear to
its efficiency. A depolarizer uses microcrystals as optical retarders.
Diffuse scatterers will destroy both temporal and spatial coherence.
However, spatial ccoherence can be brought back with an f-stop type apeature.
I think that really good f-stops ones are called "spatial filters."
Use a diffuse scatterer. LabSphere sells Spectralon diffusering
reflectors. If you want to scramble the coherence, completely, use an
integrating sphere.
Opalized glass will also scramble coherence.
David L. Rosen
swear to its efficiency. A depolarizer may be able to get rid
of some temporal coherence.
Prof Harvey Rutt
David L. Rosen wrote:
> If you want to scramble temporal coherence without scrambling
> spatial coherence very much, use a depolarizer. However, I can swear to
> its efficiency. A depolarizer uses microcrystals as optical retarders.
A depolarizer has *no* effect on temporal coherence; as its name implies it converts
linear etc pol. light into randomly polarized. The best known are the Lyot & another
whose name escapes me (Cornu??) ; they rely on calcite wedges, & in one kind introduce
unresolvable rapid spatial variations in polarization, in the other unresolvably rapid
spectral variations. The latter is hence no use with a monochromatic source. A good
quality depolarizer has no effect on spatial coherence either. What is 'unresolvable'
depends on your system; true depolarizers are surprisingly complex.
> Diffuse scatterers will destroy both temporal and spatial coherence.
A diffuser destroys spatial coherence but not temporal.
> However, spatial ccoherence can be brought back with an f-stop type apeature.
Once destroyed by scattering spatial coherence can only be recovered with an extreme
loss of intensity - and you would be back where you started!
> I think that really good f-stops ones are called "spatial filters."
A spatial filter is normally a combination of two lenses with a pinhole at their common
focus; this does the opposite of destroying coherence; it is not an f stop which is an
aperture in the lense plane, loosely.
> Opalized glass will also scramble coherence.
It scambles spatial coherence & polarization, not spectral coherence.
Harvey Rutt
slu...@lw4u.com
I have a lot of experience in "killing" coherence exactly for the purpose of
using lasers as illumination sources for inspecting objects with line scan
cameras. May I presume this is a machine vision application?? Why one
should care whether the detector is a line scan camera is because they *can*
operate at much higher rates than area scan cameras. Thus, if you were to
use a technique such as current oscillation to reduce the speckle from a
laser diode (the oscillation forces more modes to operate, thus reducing
*temporal* coherence length), you would have to choose a higher frequency for
a fast line scan than you would for a slow area scan camera. But anyway,
there are *passive* *spatial* coherence reduction techniques that can be
useful:
The first way to do it is to break the laser. :)
As someone has suggested, running the laser through a multimode fiber and
wiggling it works rather nicely, but it is still active. You can also do the
same trick with a cheap, mega-multi-mode fiber (if such a term exists!) -- by
this I mean a long solid glass or plastic rod, say several millimeters
diameter -- read light pipe. What can work nicely is to take an acrylic rod
and bend it into a few loops, but not so tight a bend as to lose too much
light by killing TIR. The light can take many paths through the pipe with
different optical path lengths. This reduces spatial coherence and thus
speckle, and you don't have to wiggle it.
Speaking of this, two other techniques have worked quite well for me. The
first is to inject the light into many different fibers, each of which has a
different length than all the others. I have made bundles of about 30 with
1" length difference. So, for example, you might have a bundle with the
lengths being 21", 22", 23", 24", etc. At each end you bring the fibers
together, just as if it were a normal fiber bundle. Inject the light at one
end, and see a pretty good speckle reduction at the other end because of the
pretty good reduction in spatial coherence. This is also pretty efficient!
The other technique that worked well for me is to inject the light from a
diode into a small integrating sphere, then use the light coming out of a
second port to illuminate your target directly or to inject it into a fiber
bundle. Once more the light takes many different paths of different lengths
inside the integrating sphere. You should probably stay with a smaller
sphere for a given sized port -- the efficiency is in part a function of the
ratio of port area to surface area. I used Spectralon (great material!) from
Labsphere to make the final units, but a ping pong ball with a couple 1/8"
holes drilled in it is a nice start. Keep the holes about 90 degrees apart.
This method is much less efficient than the above, but is easier to try
quickly with a ping pong ball.
Both of the above were techniques I "invented" when I worked for Automatic
Inspection Devices, Inc. (I use quotes because I probably wasn't the first,
but I hadn't heard of them before.) As far as I know, they did not pursue
patents on the ideas. At least my invention records never saw an attorney's
desk that I know of, and both techniques were incorporated into commercial
inspection systems that were first sold many years ago and thus can no longer
be protected.
The milk/water in a cuvette idea can be enhanced by adding a little piece of
resistance wire at the bottom of the mix. Of course the ends are connected
to better conductive wire and a small current is sent through it. The heat
from the resisitance wire can set up a flow in the solution (my vocabular
fails me -- I can't think of what that kind of flow is called!??!), making it
more active, but without any moving mechanical parts except the solution. A
little white glue and water works well too. Both are limited to the shelf
life of the glue or milk. (Pew!) Colloidal gold such as that used in certain
inks is probably a better choice that stays stable. I've used it, but the
cuvette idea was not as attractive for the particular application at the
time.
For all of these techniques, a laser diode is usually a better choice than is
a gas laser. Typically a laser diode has a much shorter coherence length to
start with than does a gas tube. My efforts to use passive techniques to
de-speckle a HeNe did not produce good results except with a large
integrating sphere with small ports -- very inefficient. Also, you can still
use the current oscillation technique with a diode with any of the other
methods to further reduce coherence length.
You will also see different results depending on the numeric aperture of your
imaging optics and the apparent size of your illumination source.
Hope this helps,
======================================
Spencer Luster, Owner
Light Works, Creative Optical Devices
"Making Light Work for You"
1700 N. Westwood Ave.
Toledo, OH 43607
ph: 419-534-3718 FAX:419-534-3717
www.LW4U.com
======================================
In article <6pvdha$bs1$2...@news.ro.com>,
Joe Fikes <j-f...@r-o.com> wrote:
> Does anyone know of an effective way to kill the temporal coherence in a diode
or
> gas laser? I'm trying to image a laser-illuminated target with a line-scan
CCD
> camera, and the speckle is driving us nuts.
>
> Thanks.
>
> Joe
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
Mark Wilson
>
>Does anyone know of an effective way to kill the temporal coherence in a diode laser?
Modulate the laser with an electrical signal so that the laser has a 90% or higher
modulation depth. The modulation frequency should be swept from 100 to 900 MHz until
you find a frequemcy where the coherence of the laser is reduced. Not all frequencies]
are as effective as other ones. You can use a microwave bias tee to inject the AC
signal on top of the DC signal used to bias the laser. Typically, the laser diode
is biased at 50% power, and a sine wave signal is then applied to the laser to modulate
it up to 100% and down to 10% (near threshold). A simple bias tee can be made using
a 500pF ceramic capacitor and a 100mH choke both connected to the laser. The DC bias
is applied through the choke and the AC is injected through the capacitor. Once you have
found the frequency that destroys the laser's coherence, then use a VCO to drive the laser
in your prototype. Some lasers can be purchased with a built in "self pulsation" circuit
which would eliminate a lot of this effort.
mikex
devised to eliminate speckle for CCD pix.
The scrambler was approx. 10-20 loops of the multi-mode, 600um, fiber-op
cable (from the illumination source) supported on a "flexible" aluminum
plate with a AC motor and a weighted disk attached to the motor shaft.
The weight was offset from the center of the disk to "vibrate" the plate
and fiber. The RPM of the motor was approx. 1000RPM.
This was a "quick and dirty" mode scrambler but since we HAD to use
fiber-op delivery into the wind tunnel, this method worked PERFECT!
We activated the motor when the laser shutter was opened.
We used a PhotoMetrics 14-bit cryo-cooled CCD camera for precise optical
measurements. (this was used during the development of the MDC pressure
sensitive paint program)
Mike