Maker Pro
Maker Pro

Light therapy device

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
Hello. I am hoping to explore the light therapy that NASA developed.

This video features a homemade one.

He uses 80 LEDs in what looks to be a half breadboard. I'd like to build something similar using infrared LEDs of a wavelength from 780 to 950nm.

I've heard from a researcher in this field that the upper limit of power density is 100mW/sq cm for these devices. So I'm wondering, were I using a half breadboard (300 holes), what infared LEDs and resistors would I need to fit as many as possible without exceeding the power density limit?

Please explain how to calculate the power density and what specification of the LED would make sense to look at. Also, when looking into LEDs, I saw a variety of viewing angles. When comparing two LEDs of the same light output in terms of wavelength and lumens, would it be reasonable to say that the concentration of light pointed in one direction would be higher for the LED with the smaller viewing angle? I'm equating this to the breadth of a flashlight beam - the tighter the breadth, the further you can see in one direction - but I'm not sure if this is the right way to think about it. The hope is to get as much light into the skin per LED as possible.

Thanks
 

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
I found this sort of LED at Ledsupply.com:

850nm-LED-specs.png

And plugged the values into the led wiz of led.linear1.org. I wasn't sure what forward voltage to use considering the specs on the LED above, so I went with the default value. What should I actually input - 1.2 or 1.6?

940nm-80-LED-input-values.png


This is the diagram I got:

80-NIR-LED.png

Does this look right for my purposes?

These are the notes the website provided:

80-NIR-LED-wizard-notes.png
 

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
I'm guessing that a square centimeter of LEDs would consist of 4 LEDs all jammed next to each other if they are each the 5mm size. Then that square centimeter would be producing 20mW/LED * 4 LED = 80 mW/sq cm. Is my logic correct?

Also, when I input 1.2V for forward voltage, the value of the resistor becomes 100Ω.

Any idea whether this will overheat? If so, what could I do?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
There is a significant difference between the power consumed by a LED and the energy emitted from it.

The energy conversion efficiency of LEDs ranges from 4% to 50%. If your LEDs are 25% efficient in converting electrical energy to light, then if they're operating at 1.5V and 20mA (30mW) they are producing 7.5mW of light.
 

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
There is a significant difference between the power consumed by a LED and the energy emitted from it.

The energy conversion efficiency of LEDs ranges from 4% to 50%. If your LEDs are 25% efficient in converting electrical energy to light, then if they're operating at 1.5V and 20mA (30mW) they are producing 7.5mW of light.

The spec for the LED I posted says that the "luminous intensity" is 20mW. On its face, it sounds like this is already considering the energy conversion efficiency. Not so? If not, how can I determine that percentage? It doesn't seem to be listed in the specs
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Ok, what is their forward voltage and what current is this (20mW output) quoted for?
 
Last edited:

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
I think your 80 IR LEDs will blind you because you cannot see IR heat energy. Your retinas will get cooked well done. I made a night light with 36 red LEDs crowded together and boy oh boy did they get hot.
 

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
I think your 80 IR LEDs will blind you because you cannot see IR heat energy. Your retinas will get cooked well done. I made a night light with 36 red LEDs crowded together and boy oh boy did they get hot.

Hence my question about cooling. What would you recommend?
Also, I'll be sure not to shine it in my eyes. Maybe just medium rare.
 

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
Ok, what is their forward voltage and what current is this (20mW output) quoted for?

The forward voltage is also in the pic above, though I'm not certain I'm reading it properly. I would guess 1.2Vf.

Likewise, I think the current is 20mA based on the specs above.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Likewise, I think the current is 20mA based on the specs above.

I'm not sure I trust those specs. They suggest a conversion efficiency greater than 80%
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
They say that the Vf is typically 1.2V and that at 20mA (when the total power available is 1.2 x 0.02 = 0.024W) and under these circumstances the IR output power is 0.02W.

That means 0.02/0.024 = 83.333% of the input power is converted to IR.

I'm not convinced. It's higher than the theoretical max efficiency for a white LED (53%). Whilst these and white LEDs are very different, the difference in claimed efficiency is huge.
 

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
One of the two pins on an LED carries heat to the printed wiring. At 20mA then a 5mm IR LED will not overheat because its heat is 20mA x 1.2V= only 24mW.

My 4-pins LEDs were 2.6V at 60mA each which is 156mW.
 

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
They say that the Vf is typically 1.2V and that at 20mA (when the total power available is 1.2 x 0.02 = 0.024W) and under these circumstances the IR output power is 0.02W.

That means 0.02/0.024 = 83.333% of the input power is converted to IR.

I'm not convinced. It's higher than the theoretical max efficiency for a white LED (53%). Whilst these and white LEDs are very different, the difference in claimed efficiency is huge.

Thank you for explaining that. No doubt a useful thing to understand. I came across this when I searched for "LED efficiency" on google. It refers to larger household lightbulbs, but maybe the tech has changed at the level of tiny LEDs too? Further, I've been reading that luminous intensity is typically not measured in mW, but in mcd. Any idea why they'd write it this way? I'd assume just for easy calculations of efficiency, as you showed me. EDIT: Ok, candela is based on the human eye, and IR is not visible.

In any case, can you recommend a good supplier for infrared LEDs? Just head to Mouser or Digikey?

Lastly, do you think that a smaller viewing angle makes sense for this application? I figure a smaller viewing angle would concentrate the light into a smaller area, perhaps increasing penetration into tissue.
 
Last edited:

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
One of the two pins on an LED carries heat to the printed wiring. At 20mA then a 5mm IR LED will not overheat because its heat is 20mA x 1.2V= only 24mW.

My 4-pins LEDs were 2.6V at 60mA each which is 156mW.

Wow. I didn't even know about 4-pin LEDs. I'll do some research.
 

Audioguru

Sep 24, 2016
3,656
Joined
Sep 24, 2016
Messages
3,656
Philips Luxeon LEDs in a 4-pins Superflux package are made in all colors but not IR.
 

Attachments

  • Luxeon SuperFlux LED.jpg
    Luxeon SuperFlux LED.jpg
    4.3 KB · Views: 61

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
. I came across this when I searched for "LED efficiency" on google. It refers to larger household lightbulbs, but maybe the tech has changed at the level of tiny LEDs too

It refers to lumens per watt, not the percentage of power converted to light.

Look at the specs for the cheapest IR LEDs on mouser it digikey or some other site that provides datasheets. Check out their efficiency. Your LEDs are probably no better than them.

I really don't know if a wide or narrow bean would be better. It depends on how close you put it to yourself. I do know that a narrow beam width will put your eyes at a higher risk
 
Last edited:

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
It refers to lumens per watt, not the percentage of power converted to light.

Look at the specs for the cheapest IR LEDs on mouser it digikey or some other site that provides datasheets. Check out their efficiency. Your LEDs are probably no better than them.

I really don't know if a wide or narrow bean would be better. It depends on how close you put it to yourself. I do know that a narrow beam width will put your eyes at a higher risk

I see what you mean. So how would you recommend I search for better ones, which would get me close to the 100mW/sq cm limit than these cheap ones? Is there a special type of LED, similar to the one audioguru suggested?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
It's really just a matter of looking at the devices available and checking out their datasheets.

LEDs are available in many packages. Some packages are designed to be mounted on heatsinks and are typically capable of much higher power.

Remember that there power rating you'll see in the headline specs will most likely be there input power. You'll need to dig a little deeper to find the radiant power.
 

mattfara

Mar 16, 2017
62
Joined
Mar 16, 2017
Messages
62
Philips Luxeon LEDs in a 4-pins Superflux package are made in all colors but not IR.

Too bad. Are they not made in IR for some physical reason, or is it possible that IR 4-pin might be produced by another company?
 
Top