Building an IR lightsource

C

Captain Dondo

Guest
I've modified a webcam to see into the infrared. And I know where to
get IR LEDs.

Now I have to build the light source itself.

This will be going into a critter cage which is glass/acrylic on all
sides. So the light source has to look decent and be unobtrusive. It
also should be easy to handle and clean.

The larger cage is about 30" long. I want to make 2 fixtures, each about
24" long, which can be put into the cage along the top of both long sides,
and pointing down / center.

I've thought about making a daisy-chain of IR LEDs (say 4 @ 8" OC) and
pulling them through an acrylic tube. Besides the obvious problems of
getting them aligned, I don't know if ordinary acrylic is transparent to
IR in the 800 - 900 nm range.

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?

Thanks,

--Yan
 
"Captain Dondo" <yan@NsOeSiPnAeMr.com> wrote in message
news:13g0j82hl0dn88c@corp.supernews.com...
I've modified a webcam to see into the infrared. And I know where to
get IR LEDs.

Now I have to build the light source itself.

This will be going into a critter cage which is glass/acrylic on all
sides. So the light source has to look decent and be unobtrusive. It
also should be easy to handle and clean.

The larger cage is about 30" long. I want to make 2 fixtures, each about
24" long, which can be put into the cage along the top of both long sides,
and pointing down / center.

I've thought about making a daisy-chain of IR LEDs (say 4 @ 8" OC) and
pulling them through an acrylic tube. Besides the obvious problems of
getting them aligned, I don't know if ordinary acrylic is transparent to
IR in the 800 - 900 nm range.

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?
Click to expand...
This sounds fine. Don't forget to limit the current in each 3 LED chain
to about 10mA. If you apply 5V and drop 4.8V, you are left with .2V,
..2/.01=20 Ohms. Dissipation is 2mW so a quarter watt resistor will be fine
for each chain.
 
Captain Dondo wrote:
I've modified a webcam to see into the infrared. And I know where to
get IR LEDs.

Now I have to build the light source itself.

This will be going into a critter cage which is glass/acrylic on all
sides. So the light source has to look decent and be unobtrusive. It
also should be easy to handle and clean.

The larger cage is about 30" long. I want to make 2 fixtures, each about
24" long, which can be put into the cage along the top of both long sides,
and pointing down / center.

I've thought about making a daisy-chain of IR LEDs (say 4 @ 8" OC) and
pulling them through an acrylic tube. Besides the obvious problems of
getting them aligned, I don't know if ordinary acrylic is transparent to
IR in the 800 - 900 nm range.

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?

Thanks,

--Yan
I browsed CONRAD.NL with "camera" search line, found there at
Click to expand...
least 5 with IR leds, nicely cased, or OEM print, for about 50-150
euros.
Try one like it in your country, leds in line with the camera
works very well.
 
V Sun, 30 Sep 2007 21:29:37 -0500, Lord Garth napsal(a):

"Captain Dondo" <yan@NsOeSiPnAeMr.com> wrote in message

I've thought about making a daisy-chain of IR LEDs (say 4 @ 8" OC) and
pulling them through an acrylic tube. Besides the obvious problems of
getting them aligned, I don't know if ordinary acrylic is transparent to
IR in the 800 - 900 nm range.

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?

This sounds fine. Don't forget to limit the current in each 3 LED chain
to about 10mA. If you apply 5V and drop 4.8V, you are left with .2V,
.2/.01=20 Ohms. Dissipation is 2mW so a quarter watt resistor will be fine
for each chain.
Click to expand...
Do you know if common acrylic is transparent to IR? If I need to do one of
those "ships in a bottle" projects where I have to drill holes in the tube
and then push the LEDs through things might get a bit more interesting.... :)

--Yan
 
Captain Dondo wrote:
V Sun, 30 Sep 2007 21:29:37 -0500, Lord Garth napsal(a):

"Captain Dondo" <yan@NsOeSiPnAeMr.com> wrote in message

I've thought about making a daisy-chain of IR LEDs (say 4 @ 8" OC) and
pulling them through an acrylic tube. Besides the obvious problems of
getting them aligned, I don't know if ordinary acrylic is transparent to
IR in the 800 - 900 nm range.

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?
This sounds fine. Don't forget to limit the current in each 3 LED chain
to about 10mA. If you apply 5V and drop 4.8V, you are left with .2V,
.2/.01=20 Ohms. Dissipation is 2mW so a quarter watt resistor will be fine
for each chain.

Do you know if common acrylic is transparent to IR? If I need to do one of
those "ships in a bottle" projects where I have to drill holes in the tube
and then push the LEDs through things might get a bit more interesting.... :)

--Yan
Try with your TV remote. See if it works trough 5 or more layers
Click to expand...
of the plastic a few meters away from your TV.
 
In article <13g0j82hl0dn88c@corp.supernews.com>, Captain Dondo wrote:
I've modified a webcam to see into the infrared. And I know where to
get IR LEDs.

Now I have to build the light source itself.

This will be going into a critter cage which is glass/acrylic on all
sides. So the light source has to look decent and be unobtrusive. It
also should be easy to handle and clean.

The larger cage is about 30" long. I want to make 2 fixtures, each about
24" long, which can be put into the cage along the top of both long sides,
and pointing down / center.

I've thought about making a daisy-chain of IR LEDs (say 4 @ 8" OC) and
pulling them through an acrylic tube. Besides the obvious problems of
getting them aligned, I don't know if ordinary acrylic is transparent to
IR in the 800 - 900 nm range.
Click to expand...
It is. For that matter, anything that is "clear" to visible light also
passes pretty much everything between visible and maybe 1400 nm or so, and
usually even longer. (Water starts to block IR at about 1400 nm and
blocks IR a lot past 1500 nm, and is more IR-blocking than most other
transparent materials). The main exceptions are found mainly where their
use is blocking/reflecting infrared.

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?
Click to expand...
LEDs all-too-often don't do too reliably well when connected directly
to fixed voltage source.

Better would to make a chain of 2 LEDs and a resistor (39 ohms for just
under 50 mA, 33 ohms for just over 50 mA, 18 ohms for about 100 mA, 22
ohms for a little under 100 mA).

Experiment and measure voltage drop across the resistor, divide square
of that by resistor value assuming the resistor is at low end of tolerance
range. Pick a resistor wattage that exceeds this, preferably by a factor
of at least 5/3 for really good reliability.

I have some experience with military grade 1/4 watt resistors that
resembled commercial grade 1/2 watt resistors, also with a couple 1/2 watt
resistors failing while dssipating about .35 watt.

- Don Klipstein (don@misty.com)
 
In article <13g0pko77030b28@corp.supernews.com>, Captain Dondo wrote:

Do you know if common acrylic is transparent to IR? If I need to do one of
those "ships in a bottle" projects where I have to drill holes in the tube
and then push the LEDs through things might get a bit more
interesting.... :)
Click to expand...
All common transparent plastics are transparent to all of the usual
wavelengths of IR LEDs.
Same is true of all common grades of glass - cheaper ones of pale
blue-green or pale greenish tint will attenuate IR a little bit, worse
than for visible red, but still pass most of the output of the usual IR
LEDs of nominal wavelength up to 950 nm - probably usually block about
half at most any wavelength in the IR-A range (past visible up to 1500
nm).

Keep in mind that most transparent colored plastics largely pass most IR
in the 850-1500 nm range. So does cobalt blue glass and "Woods glass"
(UV-pass, visible-block, dark violet color, used for "blacklights").
The most-IR-blocking kinds of colored glass appear to me to be
turquoise-cyanish to greenish and I suspect are tinted by iron or copper
or both.

Transparent greenish plastic I expect to pass most 850-1500 nm IR unless
it is a plastic filter made for use in welding goggles or a welding mask.

- Don Klipstein (don@misty.com)
 
On Sun, 30 Sep 2007 21:29:37 -0500, Lord Garth wrote:

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?

This sounds fine. Don't forget to limit the current in each 3 LED chain
to about 10mA. If you apply 5V and drop 4.8V, you are left with .2V,
.2/.01=20 Ohms.
Click to expand...
That's fine so long as the LEDs really do drop exactly 1.6V each. You only
need a 2% increase to get 4.9V (half the current) or a 2% decrease for
4.7V (50% more current).

I'd suggest making each chain use 2 LEDs and a 180 Ohm resistor. That will
be far more robust against slight variations in LED voltage.
 
In article <pan.2007.10.01.09.32.05.812000@nowhere.com>, Nobody wrote:
On Sun, 30 Sep 2007 21:29:37 -0500, Lord Garth wrote:

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?

This sounds fine. Don't forget to limit the current in each 3 LED chain
to about 10mA. If you apply 5V and drop 4.8V, you are left with .2V,
.2/.01=20 Ohms.

That's fine so long as the LEDs really do drop exactly 1.6V each. You only
need a 2% increase to get 4.9V (half the current) or a 2% decrease for
4.7V (50% more current).

I'd suggest making each chain use 2 LEDs and a 180 Ohm resistor. That will
be far more robust against slight variations in LED voltage.
Click to expand...
The usual IR LEDs do well with a lot more than 10 mA. I would go for a
good 30 mA.
With 5 volts and 1.6 volts across each of 2 LEDs, 30 mA means a 60 ohm
resistor.
The nearest standard values are 56 and 68. Keep in mind that at 30 mA,
IR LEDs usually drop less than 1.6 volts. Then again, if you use a 56 ohm
resistor and get a little more than 30 mA, I see little chance for harm -
IR LEDs tend to be rated for at least 50 mA.

Watch for heat buildup in a cluster lamp however.

- Don Klipstein (don@misty.com)
 
"Captain Dondo" <yan@NsOeSiPnAeMr.com> wrote in message
news:13g0pko77030b28@corp.supernews.com...
V Sun, 30 Sep 2007 21:29:37 -0500, Lord Garth napsal(a):

"Captain Dondo" <yan@NsOeSiPnAeMr.com> wrote in message

I've thought about making a daisy-chain of IR LEDs (say 4 @ 8" OC) and
pulling them through an acrylic tube. Besides the obvious problems of
getting them aligned, I don't know if ordinary acrylic is transparent
to
IR in the 800 - 900 nm range.

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?

This sounds fine. Don't forget to limit the current in each 3 LED chain
to about 10mA. If you apply 5V and drop 4.8V, you are left with .2V,
.2/.01=20 Ohms. Dissipation is 2mW so a quarter watt resistor will be
fine
for each chain.

Do you know if common acrylic is transparent to IR? If I need to do one
of
those "ships in a bottle" projects where I have to drill holes in the tube
and then push the LEDs through things might get a bit more interesting....
:)
Click to expand...
This would be easy to test as many video cameras see in Ir. This is
especially true of B/W cameras. Just get one LED lit up and drop some
acrylic in the path. I suspect it will attenuate to some degree but much
will pass through.
 
so then why do thermal imagers have such a bad time imaging people (heat
signatures) thru glass? standard/non filmed or IR protected. both seem to
be barriers to heat signatures


"Don Klipstein" <don@manx.misty.com> wrote in message
news:slrnfg0v2n.s3m.don@manx.misty.com...
In article <13g0j82hl0dn88c@corp.supernews.com>, Captain Dondo wrote:
I've modified a webcam to see into the infrared. And I know where to
get IR LEDs.

Now I have to build the light source itself.

This will be going into a critter cage which is glass/acrylic on all
sides. So the light source has to look decent and be unobtrusive. It
also should be easy to handle and clean.

The larger cage is about 30" long. I want to make 2 fixtures, each about
24" long, which can be put into the cage along the top of both long
sides,
and pointing down / center.

I've thought about making a daisy-chain of IR LEDs (say 4 @ 8" OC) and
pulling them through an acrylic tube. Besides the obvious problems of
getting them aligned, I don't know if ordinary acrylic is transparent to
IR in the 800 - 900 nm range.

It is. For that matter, anything that is "clear" to visible light also
passes pretty much everything between visible and maybe 1400 nm or so, and
usually even longer. (Water starts to block IR at about 1400 nm and
blocks IR a lot past 1500 nm, and is more IR-blocking than most other
transparent materials). The main exceptions are found mainly where their
use is blocking/reflecting infrared.

I have 5 VDC to power the LEDs. IR LEDs seem to want ~ 1.6VDC so I can
thread 3 in a chain.

Any suggestions?

LEDs all-too-often don't do too reliably well when connected directly
to fixed voltage source.

Better would to make a chain of 2 LEDs and a resistor (39 ohms for just
under 50 mA, 33 ohms for just over 50 mA, 18 ohms for about 100 mA, 22
ohms for a little under 100 mA).

Experiment and measure voltage drop across the resistor, divide square
of that by resistor value assuming the resistor is at low end of tolerance
range. Pick a resistor wattage that exceeds this, preferably by a factor
of at least 5/3 for really good reliability.

I have some experience with military grade 1/4 watt resistors that
resembled commercial grade 1/2 watt resistors, also with a couple 1/2 watt
resistors failing while dssipating about .35 watt.

- Don Klipstein (don@misty.com)
Click to expand...
 
HapticZ wrote:
so then why do thermal imagers have such a bad time imaging people (heat
signatures) thru glass? standard/non filmed or IR protected. both seem to
be barriers to heat signatures
Click to expand...
You are talking about a completely different part of the IR
spectrum. IR LEDs generally produce light in the 800 to
1000 nm range, while thermal imagers work with wavelengths
about 10 times that long. This is the difference between
shining near IR light on stuff and seeing what they look
like in the light they reflect, and treating the objects
being viewed as light sources (by the energy they emit
because of their temperature).
 
pretty much apples/oranges then.

thanks for the clarification.

"John Popelish" <jpopelish@rica.net> wrote in message
news:0fGdnbPLz7nGQp_anZ2dnUVZ_ualnZ2d@comcast.com...
HapticZ wrote:
so then why do thermal imagers have such a bad time imaging people (heat
signatures) thru glass? standard/non filmed or IR protected. both seem
to
be barriers to heat signatures

You are talking about a completely different part of the IR
spectrum. IR LEDs generally produce light in the 800 to
1000 nm range, while thermal imagers work with wavelengths
about 10 times that long. This is the difference between
shining near IR light on stuff and seeing what they look
like in the light they reflect, and treating the objects
being viewed as light sources (by the energy they emit
because of their temperature).
Click to expand...
 
HapticZ wrote:
pretty much apples/oranges then.
Click to expand...
Especially with respect to the absorption of window glass.
The tiny amounts of copper and iron in window glass give it
slight green or yellow tints (look at the edge of a pane).
But in the medium IR range, the absorptions are dramatic.
But even pure fused silica blocks IR just beyond 4000 nm (4
microns). Some transmission spectra for IR window materials
are shown at:
http://infrared.als.lbl.gov/IRwindows.html
Exotic materials like CaF, Ge and ZnSe are used to make the
lenses for heat imaging cameras, to pass the 10 micron
wavelengths given off at modest temperatures. Window glass,
while quite transmissive at near IR (remote controls work
through them) is essentially black at 10 microns.

thanks for the clarification.

"John Popelish" <jpopelish@rica.net> wrote in message
news:0fGdnbPLz7nGQp_anZ2dnUVZ_ualnZ2d@comcast.com...
HapticZ wrote:
so then why do thermal imagers have such a bad time imaging people (heat
signatures) thru glass? standard/non filmed or IR protected. both seem
to
be barriers to heat signatures
You are talking about a completely different part of the IR
spectrum. IR LEDs generally produce light in the 800 to
1000 nm range, while thermal imagers work with wavelengths
about 10 times that long. This is the difference between
shining near IR light on stuff and seeing what they look
like in the light they reflect, and treating the objects
being viewed as light sources (by the energy they emit
because of their temperature).
Click to expand...
 
so then these "night vision" digital tape video cameras (Sony NightShot) are
not actually seeing thru peoples clothes and viewing the heat signature, but
actually are seeing the true medium length IR and just eliminating all the
human sight band wavelengths?

then all the night hype is merely a function of a bandpass filter (or lack
of it?) of what ends upon the CCD chip?


"John Popelish" <jpopelish@rica.net> wrote in message
news:cJCdnWvltqWImJjanZ2dnUVZ_rmjnZ2d@comcast.com...
HapticZ wrote:
pretty much apples/oranges then.

Especially with respect to the absorption of window glass.
The tiny amounts of copper and iron in window glass give it
slight green or yellow tints (look at the edge of a pane).
But in the medium IR range, the absorptions are dramatic.
But even pure fused silica blocks IR just beyond 4000 nm (4
microns). Some transmission spectra for IR window materials
are shown at:
http://infrared.als.lbl.gov/IRwindows.html
Exotic materials like CaF, Ge and ZnSe are used to make the
lenses for heat imaging cameras, to pass the 10 micron
wavelengths given off at modest temperatures. Window glass,
while quite transmissive at near IR (remote controls work
through them) is essentially black at 10 microns.

thanks for the clarification.

"John Popelish" <jpopelish@rica.net> wrote in message
news:0fGdnbPLz7nGQp_anZ2dnUVZ_ualnZ2d@comcast.com...
HapticZ wrote:
so then why do thermal imagers have such a bad time imaging people
(heat
signatures) thru glass? standard/non filmed or IR protected. both
seem
to
be barriers to heat signatures
You are talking about a completely different part of the IR
spectrum. IR LEDs generally produce light in the 800 to
1000 nm range, while thermal imagers work with wavelengths
about 10 times that long. This is the difference between
shining near IR light on stuff and seeing what they look
like in the light they reflect, and treating the objects
being viewed as light sources (by the energy they emit
because of their temperature).
Click to expand...
 
HapticZ wrote:
so then these "night vision" digital tape video cameras (Sony NightShot) are
not actually seeing thru peoples clothes and viewing the heat signature, but
actually are seeing the true medium length IR and just eliminating all the
human sight band wavelengths?

then all the night hype is merely a function of a bandpass filter (or lack
of it?) of what ends upon the CCD chip?
Click to expand...
That's right. The light photographed is reflected light,
not light sourced from the skin. The silicon detectors in
CCD cameras have sensitivity out to about 1100 nm, unless
the camera includes a short pass filter to cut off the
sensitivity beyond red light at about 700 nm. The two most
common "night vision" or "x-ray" filters are 850 nm or 1000
nm long pass filters that exclude all the normal visible
light from the camera. The 850 nm filter provides a lot
more signal, since all light between 850 nm and the silicon
cut off. The 1000 nm filter shows less clothing surface
reflection and more skin reflection (which is pretty
efficient at 1000 nm) through clothing, but the sensitivity
is low, since the total band of light sensed is only between
1000 nm and 1100 nm.
http://www.ph.tn.tudelft.nl/Courses/FIP/noframes/fip-Spectral.html
http://www.optical-filters.com/rg850.html
http://www.optical-filters.com/rg1000.html


Night vision cameras also use these two filters and even
some shorter wave ones but include IR light sources to flood
the area being watched. Some of these cameras are not
silicon detector based, but photomultiplier based, that have
lots of gain, but also poop out in the near IR.

Thermal cameras use exotic detectors and lenses to let them
see the actual longer wave IR radiation emitted by warm
objects.
 
In article <jMhNi.288$sm6.90@nlpi069.nbdc.sbc.com>, HapticZ wrote:
so then these "night vision" digital tape video cameras (Sony NightShot) are
not actually seeing thru peoples clothes and viewing the heat signature, but
actually are seeing the true medium length IR and just eliminating all the
human sight band wavelengths?

then all the night hype is merely a function of a bandpass filter (or lack
of it?) of what ends upon the CCD chip?
Click to expand...
Those cameras don't see mid-IR but near-IR (wavelengths shorter than
1500 nm, almost entirely shorter than 1300 nm).

As for seeing through clothes - that's something that there is a little
truth to, and only a little. What actually happens:

1. Most dyes are transparent to near-IR past maybe 750-800 nm, so a lot
of clothing apears much lighter in color to those wavelengths than in
visible light.

2. A few fibers have surface roughness of size comparable to wavelength
(or half wavelength or whatever it is that counts) of the shortest near-IR
wavelengths. When the wavelength gets longer, the fibers get more
transparent. However, the difference that makes is normally only
comparable to that made by getting the clothing wet.

3. At longer wavelengths the index of refraction of the fiber material is
less, and the fibers scatter light less - but I think the index of
refraction is not much different at 950 nm than in the middle of the
visible spectrum.

- Don Klipstein (don@misty.com)
 
On Fri, 05 Oct 2007 13:17:20 +0000, Don Klipstein wrote:
....
transparent. However, the difference that makes is normally only
comparable to that made by getting the clothing wet.
Click to expand...
Which can also be fun! ;-)

Cheers!
Rich
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top