How Dangerous are Infrared LEDs?
ฝัง
- เผยแพร่เมื่อ 9 ก.พ. 2025
- This video provides an objective analysis of infrared LEDs in comparison with other common light sources.
More specificly, I use an infrared light meter to detect the irradiance of various light sources to determine whether they meet standard thresholds, which can give you a better idea of what type of LEDs might be safe for use in applications like eye tracking. The video also goes over the various factors that influence the power and output of an LED to help you build better infrared light systems.
0:15 What is irradiance?
0:39 IEC threshold for infrared exposure
0:50 Low irradiance examples
1:16 High irradiance examples
2:00 LED comparisons
2:40 Current (milliamps) demonstration [**correction: note that the values were 4mA and 100mA for the spinel and generic cameras when measured in series with the current meter]
3:28 IR Torch
3:50 List of all measure irradiance values
All of the gadgets/parts in the video can be found on Amazon:
Irradiance meter: amzn.to/3EmtvZD
Spinel OV9281 camera: amzn.to/4jLQHAP
940nm 3mm LED diodes: amzn.to/4hq1TS8
IR Torch: amzn.to/3Cs4Wdr
Multimeter: amzn.to/3WOu0SO
Thanks to a few folks for pointing out that the current should have been measured in series, not parallel! I've posted the updated values in the video description after re-measuring. Also, a few people mentioned that the irradiance of the sun is measured to be around 1k watts per square meter, but the NIR meter I used was registering over 3k. I think it might be the meter itself, but I'm further investigating to see what might be the cause!
Measuring current in parallel?!? You are shorting out the LEDs! Never do this! If you did this in a high energy circuit your meter will explode!
Yeah, good thing these were very low current applications. Somehow this completely slipped my mind.
Thank you for your interesting and informative videos. At 2:42 in your video, you use a multimeter in current mode and attempt to measure the current through the IR LED. However, the measurement probes are placed in parallel with the LED rather than in series. Hence, I am afraid that in this configuration, the measured value will not correctly indicate the current passing through the LED. Am I missing something here? Once again, thank you for your videos.
Uh oh, let me check that!
I re-measured, and you are correct. The values when connecting the LEDs in series with the current meter were somewhat similar though: 4mA for the Spinel, and 100mA for the generic webcam. Thanks for catching that!! I've been doing CS for too long, lol.
Great news is likely the diodes are driven current limited; placing a very low resistance path (ammeter) in parallel with a significantly relative higher resistance load (the diode) The majority of current will flow through the ammeter giving a fairly accurate short current readout.
Cheaper devices or low current diodes may use a simple resistive load in series with the diode, which would give you a much different readout in a short.
Epic fail!
@@andrewn7365 No it's not. Expecting perfection from everyone are all the time without the learning experience is a fools dream. It's only an epic fail if it's repeated after the lesson.
Can you verify your sunlight measurements? You report 3000-4000 W/m2. However, the *total* solar irradiance at the Earth's surface is about 1000 W/m2 (entire spectrum) and the irradiance at the top of the atmosphere is 1360 W/m2 (en.wikipedia.org/wiki/Solar_irradiance#On_Earth's_surface and en.wikipedia.org/wiki/Solar_irradiance#At_the_top_of_Earth's_atmosphere). IR irradiance is about 50% of these values.
Yeah, I'm guessing the meter itself is off. I will probably conduct some other tests with it to see if I can calibrate it or how it compares to other devices. Hopefully it still gives everyone a good relative comparison of the different light sources though.
The meter might be reading off, a lot of meters have a photometric linearity filter that linearizes the response from 980 to 400nm. If you shine a broad band mostly visible light source on on without this filter it will give a false high reading.
This meter has readouts for both visible and IR spectrum, but I wonder if some sort of filter isn't present. I will likely get another device for testing at some point. Any recommendations?
Really cool experiment and a great informative content to explore. However, there’s a significant breakdown somewhere in the measurement process or in the conversion of correlating data points.
For reference, the maximum total power of sunlight, including visible and invisible spectra (summer solstice, on a clear day, at sea level, noon, near the equator) hovers around 1,000 W/m². Outside the atmosphere, the solar constant sits at 1,361 W/m² in direct exposure. Infrared typically makes up ~52% of that total, with slight fluctuations depending on atmospheric conditions.
The part that has a little bit to play with this but is often overlooked is wavelength-dependent damage. Wavelengths longer than ~1700 nm are mostly absorbed by the cornea, where they can lead to cataracts and surface-level thermal damage. Wavelengths between ~1700 nm and ~400 nm pass through the cornea and lens, focusing directly onto the retina. This sub-ionizing spectra induces damage almost entirely through thermal effects. much like using a magnifying glass to concentrate sunlight into a burn spot. This type of damage is commonly referred to as retinal burn.
The given power density of W/m² is useful as a guideline metric, but it's not one I would personally depend on for safety. The inverse square law has a lot to do with this and collimation, and focus drastically alter realworld exposure risks. A diffuse IR source at 1 W/m² is nothing compared to a tightly focused beam of the same power or divergent beam spread at extreme close range. Like you touched on; while exposure limits are often cited at around 100 mW/m² for eye safety, this is a generalized guideline, it doesn’t account for beam divergence, exposure duration, or pupil dilation. Yes the measurement of watts per square meter really makes sense as an instantaneous slice of the risk, but only tells part of the exposure factor when assessing biological risk.
All that said, this is an awesome topic, and I'd love to see this was some even more in depth exploration. It might be worth borrowing a calibrated irradiance spectrometer from a university lab and redoing the experiment with a bit more precision, and the collaboration could add some really solid content.
That's great insight, thanks! Hopefully this video will encourage others to test things on their end as well. Are you in aeronautical science by chance?
@jeoresearch general science and engineering.
Wait a minute, how can you measure the current in parallel?
See my note above. Had to remeasure it in series.
That is not how you measure amperage and the current flowing through the LED.
Yeah, someone else noted that as well. I already re-measured with the LEDs in series and put the new values (which were somewhat similar) in the description.
So what is safe, and what is not safe?
It's highly dependent on conditions, but you should mainly ensure that the irradiance on the eye is less than 100 watts per square meter per the IEC standard.
A very underestimated and rare topic. I love that someone has looked into it and is pointing it out to others. Unfortunately, the video isn't quite as well researched, and the readings seem a bit off. It would be new to me that sunlight has the shown IR power.
Yes, based on some other comments, it seems the NIR meter I used isn't 100 percent accurate. I will probably conduct some other tests with it to see if I can calibrate it or how it compares to other devices. Hopefully it still gives everyone a good relative comparison of the different light sources. Will improve next time!
now , how did our ancestors survived ??
With a much lower average longevity :)
@@jeoresearch eh buddy, I get the sentiment that you're trying to convey here but It's not even correct.
You can easily find multiple records of people in ancient Greece living to 70-80, and they didn't have "modern medicine".
Of course you were still more likely to die in general (lowering the average), as an infant or due to violence / disease. But the hard pill to swallow is that if you were lucky enough to make it, your life expectancy wouldn't be much different than what we experience today.
TLDR: They are not dangerous.
In many cases, yes, but running them at too high a current can be bad!
Stop, The Sunlight is 1000% safe in normal conditions. this is a BS video
🤣
So iec ratings are bs
I think they definitely err on the side of caution.
Why is on thumbnail picture of ultraviolet led if you talk about infrared led ?
Great question!! Infrared actually shows up as a red or purple hue on most cameras. I believe UV would show up as blue, but that's something interesting to test!
@jeoresearch this is really strange. I have ultraviolet lamp, and my phone camera see's it as light blue colour with violet lens flare.
I didn't expected that.
Also ir light on camera looks like dark purple.. totally unexpected
Sorry for thinking you were liar and for thinking you itentionally put uv light picture
oh god i used that exact ir flashlight with a full spectrum camera to look into my eyeballs 🫣
I hope not for a significant length of time!
@ less than 5 minutes i think