Let's talk DIY lights


Yup! It’s all about sizing the load efficiently to the cooling system.

In my case it’s so far overkill that it’s kinda funny.


Hey, cheap is good free is better right?


150 F -> 35 F. You are confusing temperature with energy. Doesn’t work that way. Besides it’s all a non-linear system because heat transfer takes place at different rates for different temperatures of the heat sink and the air moving over the fins. If they are the same temperature, zero heat transfer takes place. The faster the air moves, the less temperature difference between the two because more heat is being transferred.

Believe me, it works. My heat sinks get too hot to touch with zero air movement and run about 90 F with a very weak fan. About 24 cubic feet per minute through each heat sink. They are about 4 square inches per watt, which is 1/4 or so the recommended 17 square inches per watt for passive heat sinks.


I know they work, but I don’t see where they are 4x better with a fan?

I don’t doubt for a minute that you spent less on heatsinks. But there’s not a formula, statistic, or document anywhere that will convince me to design any of my lights with active cooling. You simply don’t need to. Outside of trying to make a heatsink that you already have or is recycled work, I don’t really see the point. Depending on the design, passive sinks aren’t really that expensive when you consider the alternatives. And require zero additional fans or fab work that won’t already be in a grow space.


From Application Note AN30
Thermal Management for Bridgelux Vero Series LED Arrays:


I’m not sure what significance that chart is suppose to have? Needing more surface area on a passive heatsink is probably mentioned 20+ times In this thread alone.


If you read that document you will see from that chart and the next that fan cooling makes heat sinks about 6 times more effective, not 4 times more effective. My apologies.


You cited the same application note when you suggested that the thermal pads wouldn’t be sufficient for thermal transfer. And we can make the same assertions about heatsink cooling as we did that. Any difference will vary based on the exact application. The chart clearly shows that a properly sized passive heatsink can outperform an undersized active heatsink just as easily. In those applications, adding a fan won’t provide any advantage over a passive heatsink.

Maybe I misunderstood your comment about 4x better performance. It’s certainly not anything like that in led performance. If you meant solely an increase in thermal performance of the heatsink, then sure it can happen. But adding a fan doesn’t make it automatically better by any amount, let alone 4 or 6x.


would a street sign work as a heat sink? @dbrn32


They’re steel aren’t they?


alum/mag alloy i think


I guess I should sharpen up my sign game lol. I’m not sure how well. I suppose it would pull some heat, but probably not as much as the more pure aluminum alloys. I can do some digging around.


there are street signs on every corner around here for the taking!


Lmao! Somehow I feel it’s probably not the best idea.


they will all ready have a name too!


Looks like they are 5052, which would be fine. But I’m not sure about the coating they have them? @BIGE


the backs are uncoated on these…they are old outdated signs taken down…lol


I would think you can clean them up and give them a shot then.


It’s all about surface area. Passive is fine, it just takes 6 times as much surface area as a fan-assisted heat sink. You need about 16 square inches per watt passive and you can figure latest generation COBs send about half their watts into the heat sink. So for a 20 watt COB you would have about 10 watts and need 160 square inches passive or 27 square inches actively cooled.

So my comments about adding a fan is just relative to M4ur’s use of that microprocessor heat sink. It’s too small to be used passively, but is fine with a fan.

As for using a street sign, bar stock, or whatever: As long as it’s aluminum, it’s just a matter of calculating the surface area. If it’s steel, forget it. Heat transfer of steel is very bad. Unlikely that anyone would use copper, but it’s almost twice as good as aluminum at moving the heat from the COB out to all the fins, But the fin-air interface is really the bottleneck here so a copper heat sink makes little sense in passive mode.


Here’s the first run under the diy lights. They got moved over today after a bleach scrub and disinfect of the grow area. Lights are at 50% and approximately 12" from the tops.