Surface Mounted IR LED's In Surveillance Cameras

Wondering if anyone here has performed a comparative analysis on the performance of these little Surface Mounted IR LED's vs. the traditional pronged diodes.

SMT IR LED vs Traditional LED

1. Do these SMT IR's output more light than the traditional ones thus requiring less of them to see the same distance?

2. Do SMT IR's have any other distinct benefits? (I've heard they do not heat up as much).

Any help would be much appreciated.

Thanks IPVM Community!

Candace, good question.

1. Yes, typically SMT LEDs can cover further range / distance than PTH (plated through hole) LEDs. Related, that's why you typically see cameras with many more PTH LEDs than cameras with SMTs. An IR camera with SMTs may have only 2, 4 or 8 while one with PTHs routinely have 24 or more.

2. The other claimed benefit is, as you say, better heat management. My understanding is that since the heat of the SMTs do not radiate out as much it is easier to manage heat.

In terms of a study, we have tested many cameras with both types. In general, the cameras with SMTs have better image quality / more even FoVs but that might also be because they are from 'better' cameras overall.

The other issue would be life time / reliability and that we do not have definitive stats. That being said, most of the mid to high end IR cameras have moved to SMT and you generally see PTH on real low end cameras.

Thanks for the informative response, John!

The surface mount LED chips do better for a few reasons. They come in 1-3 watt flavors but remember, wattage means little, it's luminance at 850nm that counts. Also, the lenses used are simply better than the plastic surround used in PTH versions. Remember to consider the focus of the lens, 15, 30, 45, 60, 90 degrees?

Finally, we demand our IR suppliers use the Epistar brand of IR LED chips. We have found them to have the highest luminance at 850nm. They cost a bit more but they're worth it.

John, excellent first comment, thanks for sharing!

We have found them to have the highest luminance at 850nm.

Welcome John!

Curious what are you using to measure "luminance" at 850nm?

I'm in Atlanta with lots of Georgia Tech connections. I asked a friend to use one of their light spectrum analyzers. Epistar emitted 30-35% more photons at 850nm. From an OEM (which we are) standpoint they are also about 30% more expensive at the chip level but that's 40 cents to 52 cents so... doesn't matter.

Nice to meet you. You all do a great job here. Best in the industry in my opinion.

Give me a few more months and I'll show you what we're doing. Very different, very focused, very neat.


Also, as for "luminance", as you likely know it's really radiance in the IR band. (W/m2sr)

Sorry, I forgot who I was speaking with..


Nice save! :)

Anyway, the concept of luminance in describing surveillance lighting leaves a bit to be desired.

I mean I understand the concept of normalizing the intensities by the typical sensitivity of human vision, but a human is not capturing the raw light, only a sensor. And in the case of B+W sensors, for instance, the captures are more in line with the radiance.

Even in color cameras the extra green tile in the mosaic is but a crude approximation of the complexity of our vision system.

There are several Bayer-less designs possibly making their way to surveillance cameras, we'll have to see if they might be better served by flux vs lux.