UV-B/C LED's

JOO©E said:

Looks like a decent chip. All UVB led are piss poor efficiency which correlates to higher operating temps (per watt input), and reduced longevity compared to visible spectrum LEDs.

Click to expand...

I agree on this. Since the uvb/c leds are so inefficient the actual amount of heat the chip has to deal with would be almost double of what a standard white diode would have to deal with at same wattage. Another approach would be going for more and cheaper uv chips and run them soft. Now with covid there are many new options since there is a new demand.
Another thing to think about is hanging height: if 15mW output is ok for 12", assuming inverse square root reduction in intensity, then 4mW per diode should be ok for 6" which is where our lights hang. 4mW outout was around 40mA on the last chips i checked. Means about 10x less watts and heat per diode.

Rocket Soul said:

I agree on this. Since the uvb/c leds are so inefficient the actual amount of heat the chip has to deal with would be almost double of what a standard white diode would have to deal with at same wattage. Another approach would be going for more and cheaper uv chips and run them soft. Now with covid there are many new options since there is a new demand.
Another thing to think about is hanging height: if 15mW output is ok for 12", assuming inverse square root reduction in intensity, then 4mW per diode should be ok for 6" which is where our lights hang. 4mW outout was around 40mA on the last chips i checked. Means about 10x less watts and heat per diode.

Click to expand...

Some good points here, and also in your previous comments too! Tbh it's probably closer to 25× - 50× more heat per watt input. Visible spectrum are around 50 - 65% efficient while UVB are like 1-2%.

Because the LED has only 1 way of shining, ie it doesn't shine in a whole sphere, but only a half sphere (or an even narrower beam), you have to adjust your intensity vs height calculation a little bit. You essentially have to double (plus a bit more) your typical figures. This means surface mounted LEDs won't reduce in intensity as much as a candle or a lightbulb (or any spherical emission) will when there's an increase in distance from the source. You can utilize beam angle and even various software to get exact. Most chips emit in a 120° cone and the % of volume that the emission cone occupies of the total sphere will be the determining factor in how the intensity maintains as the distance from the source increases, but in a very generalized sense you'll be much closer to real figures if you multiply traditional figures by at least 2.

I too am of the opinion that multiple chips at low intensity are superior to a single high intensity source. Uniformity is important to consistent results. Longevity increases at low current, and operating temps are reduced. Efficiency!!!!!

Cant believe we're having a thread on UVB and Solacure hasn't been mentioned. UVB definitely works but they have to be legit bulbs. Solacure is the only company i know thats doing it right.

Ima just throw this into the fray for fun....

I certainly have considered simply incorporating the F40 (or F20) Flower Power in my design - I wish thy didnt charge as much for the 2ft as the 4ft, unless thats a pricing error on the site. But either way, this was more targeted toward the led audience to see if a similar chip had already been tested. I certainly appreciate all the feedback Ive received, certainly some new information for me, as Im just starting to learn and understand the UV stuff, esp when it comes to LEDs. More specifically, as someone else pointed out, the heat dissipation - based on what I can find to make rated optical power (30-35mw), it takes 2.25w electrically - thats really not difficult for the aluminum substrate and frame to deal with, so Im not too concerned there, as we're only talking a couple of these, and their cooling will be discrete from the rest of the light. At this point, Im just curious what the coverage ratio/power would look like as I run my cobs around 12-18" to avg 800-900 ppfd. I have a quantum sensor, but afaik the SQ-420 is unable to measure below 400nm. Ill have to look into something for uv if the price isnt outrageous, might be worth it for proper experimentation.