COB Talk

[ttystikk]

Exactly what I'd discovered by spending hours playing the Product Characterization Tool (PCT).

I had built my bars for passive, but upon testing one over 6 hours I found my heat sink temp. getting up to 47c and slightly above with an ambient temp. of 24c.

I too would have loved to run these bars passive but given my entire reasoning for building my own was to get the best efficiency possible and so I built some add on fans that run @1300 RPM and are virtually silent. (Have to put my ear up next to them to hear the fan at all). The end result being that my heat sink temps. @ the same ambient temps are now a mere 28.5c directly behind the cob after 6 hours running. From my calculations over the 4 bars I've built that works out to be 16-20K lumens output.

I was afraid it would destroy my good looking bars but find them pleasing to look at with the fans.


PS. Need to tidy up the fan wiring and will before I hang them.
And they are truly heavy the heat sinks weigh in @10KG each.

View attachment 3663017 View attachment 3663018 View attachment 3663019

And as an aside I also cool my drivers with fans as they also suffer losses from high temps.

Brilliant! Chunky heavy heatsinks are a good thing, especially with a lil active cooling.

 

[ttystikk]

I built both my veg and flower lights actively cooled with Vero/CXA running @ 1.4 A. Not only were the fans a pain in the ass to wire up, but it added to the parts list and build expense. And I've since had to dim my lights since they are so bright, basically eliminating the need for a 100w heat sink.

Going forward I will be running passive sinks that are pre-drilled for my application, and probably be running the chips at 1.05A while still dimmable. This will make the build come together sooo much quicker and cheaper too.
I'll sacrifice some lumens for convenience. The amount seems neglible to me.

If you run your passive setup extremely softly, there may not be a big reason to actively cool it- on the other hand, we're paying large sums specifically for efficiency, a few extra bux for fans to maximize it seems like low hanging fruit next to the job and cost of building the whole unit.

The temperature droop curve says there's 10% efficiency to be gained between 85C and 25C. That's a lot. If fans alone can add a few points, my contention is that the incremental performance improvement gained is generally worth it.

I plan to take this idea to its logical conclusion in my own build.

 

[J Bleezy]

My active cooling is an oscillating fan on the wall. In all seriousness though, my diy passive light stays much cooler than my spydr1200, I can barely feel any heat when I touch the heatsinks.

 

[cdgmoney250]

@ttystikk This is true... But who is running their Tj close to 85C? Most of the temps I have seen being passively cooled have been 50C or less. The difference between 25C and 50C, is probably less than 5% in total light output (guessing).
Starting to split hairs here. I wonder if anybody has tested the ouput difference between high and low Tj? Other than the manufacture of course.

 

[PurpleBuz]

Because I like to be technologically disruptive, I thought I'd pop in here and ask why people think passive setups are so great? Is it because you're saving a few watts on a fan?

I have found efficiency wise, a small amount of fan power makes a huge diff to the performance of a well designed light system over a fully passive system. the amount of heat sinking (regardless of type of heat sink) required to go full passive just doesn't justify eliminating a couple of watts of fans. In some of supras testing with fans\heatsinks he found the same thing.

I just built a 120 watt light bar that uses about 3 watts of fans. Veros 18s running at 520 ma mounted on a 44" heatsinkusa 2.079" profile, which is mounted on a 48" 1/8 inch thick channel

Temps right behind the cobs (vero18s) on the 2.079" profile heatsink with ambient temp @ ~25C
with fans ~32C, without fans ~46C.

From my perspective the only reason to go fully passive is for an IP67 rating, and even then IP67 and IP65 fans are available at a reasonable price.

Furthermore consider the BML spyders. Biggest complaint that I know of besides price is that the units run hot AND the work around is to add a fan that blows on the unit. While a lot of grow rooms have enough air flow to manage this, that's really (in my opinion) BML ducking responsibility of keeping the unit running optimally.

Now just to be an ass.... can someone tell me how much more heatsink you will need to run at 33C without fans ?

 

[JorgeGonzales]

@PurpleBuz

In fairness neither 46C or 32C is hot to an led, and the heat stays in the room either way. I'm pretty sure these low current COB designs with passive radial or pin heatsinks are running much cooler than the BMLs with their long extruded designs, and single leds.

If the vero 18@500ma is 50% efficient, it's dissipating 7 watts or so, and Tj is roughly watts dissipated x total thermal resistance in C/W.

This ends up being easy math, since 32C-25C = 7C, so you would need a total thermal resistance of 1.0 C/W. This includes heatsink, led, and thermal interface material added up, so whatever that is. I think my COBs are .17C/W plus whatever the paste is. 32C is certainly doable, with something like this.

Your heatsinkusa profile is 7.02 C/W per 3 inches of length. You can get the effective thermal resistance by dividing your total length by number of cobs and using this chart. Your actual junction temp will be higher than the heatsink measurements you took, but those are good for relative differences.

 

[ttystikk]

I have found efficiency wise, a small amount of fan power makes a huge diff to the performance of a well designed light system over a fully passive system. the amount of heat sinking (regardless of type of heat sink) required to go full passive just doesn't justify eliminating a couple of watts of fans. In some of supras testing with fans\heatsinks he found the same thing.

I just built a 120 watt light bar that uses about 3 watts of fans. Veros 18s running at 520 ma mounted on a 44" heatsinkusa 2.079" profile, which is mounted on a 48" 1/8 inch thick channel

Temps right behind the cobs (vero18s) on the 2.079" profile heatsink with ambient temp @ ~25C
with fans ~32C, without fans ~46C.

From my perspective the only reason to go fully passive is for an IP67 rating, and even then IP67 and IP65 fans are available at a reasonable price.

Furthermore consider the BML spyders. Biggest complaint that I know of besides price is that the units run hot AND the work around is to add a fan that blows on the unit. While a lot of grow rooms have enough air flow to manage this, that's really (in my opinion) BML ducking responsibility of keeping the unit running optimally.

Now just to be an ass.... can someone tell me how much more heatsink you will need to run at 33C without fans ?

Thank you. Excellent points, all.

 

[PurpleBuz]

@PurpleBuz

In fairness neither 46C or 32C is hot to an led, and the heat stays in the room either

View attachment 3663306

that's true but whats your point ? how about compaing the cost of a passive heatsink to achieve 32 versus 42 ?

 

[Greengenes707]

I hear a lot of thought. Lets see some testing or proof anyone?
I used to be pretty anti passive. But things have changed and passive doesn't mean 85c and heavy as hell. As well as active does not mean 25-30c.

Everyone thinks their active units run a few degrees over ambient??? Do please show? I don't want to see any sink temps. I want case temps? Specially for an active setup. If you're a DIY'er of more than one light and don't have a thermocouple...come on. And if we are talking about an inclosed case unit(retrofitted lights, using project boxes, fin shrouds,...), active units lose most of their advantages over passive. Of course that is all in a controlled and stagnant testing environment that I have tested in. Once passives are put into a grow room...the slightest air movement is as good as an active fan. "Active raw/open frame" will gain some from a grow room air movement too, but "active enclosed" not so much.
I used to hate BML for recommending to use a circulation fan...and still do. You should not have to recommend that, the lights should be perfect with bad air movement and amazing with great air movement. But the fact of the effect, is true. In a grow room, fan drop Tc of a passive units significantly.

I know many people who have built either simple to pretty complicated "in case of fan failure" systems for their lights. They aren't needed passively. Fans and all components related...not needed.
140mm Pin sink will be under 20$ in a few week from PLC. A better cost value than HSUSA and you can space as needed and save a good amount of weight. For high efficiency build(50%+, basically every cob build from here out), no one is going to need more than a 140mm. And higher wattage only gets better for the pins vs extruded bars. And smaller profiles will be cost less.
All that said...the cost can be cheaper than an active unit to achieve the same Tc.


Then of course the very basics of heat in the room...active or passive...they both contribute the same amount of heat to the environment...technically the active actually contributing more.

 

[JorgeGonzales]

that's true but whats your point ? how about compaing the cost of a passive heatsink to achieve 32 versus 42 ?

Huh? My point is if the led is operating within it's specs, at an efficiency you are comfortable with, who cares what the temperature is. If the BMLs need fans they are poorly designed.

You asked how much heatsink you would need to get to 32C, I did the math, and linked to a $12 heatsink. So I guess my answer is $12. I have no idea what your heatsinkusa total was, how many COBs you run, or how many fans, so I guess you can tell me instead how much cheaper your setup is.

But as I said earlier active is usually lighter, cheaper, and more efficient. But also more work, and more things to break down, potentially catastrophically if you are running heatsinks that need active cooling to maintain a safe temperature.

 

[PurpleBuz]

All that said...the cost can be cheaper than an active unit to achieve the same Tc.
View attachment 3663344
.

absolutely good points.

but who says a pin fin can't be used with active cooling or should I say semi-passive ?

and btw I just showed you some numbers.

 

[Greengenes707]

and btw I just showed you some numbers.

Did you though?...

Your numbers...

Temps right behind the cobs (vero18s) on the 2.079" profile heatsink with ambient temp @ ~25C
with fans ~32C, without fans ~46C.

My request if we are going to talk accurately and homogeneously. There are many variables sink to sink...not Tc point of a 3590 to Tc point of a 3590 or whatever chip and is referenceable to a Tj and repeatable somehow by anyone...

Everyone thinks their active units run a few degrees over ambient??? Do please show? I don't want to see any sink temps. I want case temps? Specially for an active setup. If you're a DIY'er of more than one light and don't have a thermocouple...come on.

 

[PurpleBuz]

Huh? My point is if the led is operating within it's specs, at an efficiency you are comfortable with, who cares what the temperature is. .

look at the charts perf diffs between 32 an 45 C. i'm not looking to be in spec, i'm looking for more dollars/lumen/watt.

and btw you didn't answer my question how much heatsink (COST) will it take to bring temps to 32C to match the fans ?

 

[JorgeGonzales]

@Greengenes707 I just want to say I love that chart. I used it to decide to go passive, and double check all my math in the process. It was invaluable.

I ended up with a Mechatronix 99x80mm radial at about 1.02 C/W pair up with a 52V Citizen COB running a relatively soft 700ma, and knew it would be fine. It really, really helps being able to cross reference datasheets with hard testing, and shoot for the middle.

Anyway, thanks.

 

[PurpleBuz]

Did you though?...

Your numbers...


My request if we are going to talk accurately and homogeneously. There are many variables sink to sink...not Tc point of a 3590 to Tc point of a 3590 or whatever chip and is referenceable and repeatable somehow by anyone...

good enough to tell me that the cost of the fan and the watts that drive the fan are worthwhile for the unit I built.

that was really my point.

 

[PurpleBuz]

Huh? My point is if the led is operating within it's specs, at an efficiency you are comfortable with, who cares what the temperature is. .

REALLY ? wow
the lower the temp the higher the efficiency ..........................................

 

[VegasWinner]

Pin sinks biggest advantage over a radial heat sink is that they remain at peak efficiency when used at an angle, but star/radial heatsinks offer the same weight and cooling advantages, and don't cost as much. Having just shopped for way too long on a build, radial heat sinks are less expensive to buy for similar thermal resistances, probably because they are easier to manufacture, but thats just a guess.

Here's two I was going between before ordering for my build. The first will look familiar if you have seen the Cutter/Growmau5 kits, $17 at CDI: https://www.cdiweb.com/ProductDetail/LPF11180ZHEB--mechatronix/573584/

The second is a radial comes in under $10: https://led.cdiweb.com/ProductDetail/MODULEDXTRA9980C-MechaTronix/574062/

The radial is 573g vs 567g, and cools a tiny bit better at 1.02 vs 1.07 C/W. Definitely a little heavier, but nothing like an extruded heatsinkusa, and a lot cheaper to buy.

Pin heat sink would be great for side lighting or angles like Growmau5 is working on.

Edited to fix weights after re-reading datasheets.

I used the Moduled version and they are cool enough to hold with your hand. peace

 

[Greengenes707]

good enough to tell me that the cost of the fan and the watts that drive the fan are worthwhile for the unit I built.

that was really my point.

My point is that it is not as big a gap as you think when you measure the actual Tc.
And if you were to take the fan cost(fan+to power it) and put it towards sinks...you would lower your passive Tc more for the same price as active. In your situation you are only removing the fan...and not adjusting sink size. That is not a test of passive vs active cost and practice, it's jts your active(or semi passive, call it what you want) setup without a fan...doing exactly what it is supposed to...cool activity and cool not as good passive.

If you want to actually be real about it and compare the cost of 2 build each designed for its own benefits...then we can come back to this.

No one ever said active builds are a waste or worthless. Or that your build was off in some way. You just used it as an example of passive vs active and it's not.

 

[Evil-Mobo]

You guys are making me want to return my heat sinks to HSUSA and order the round pin models for my light build lol.........

 

[VegasWinner]

I have both active and passive DIY leds. I wired a fan for the passive light for the plants, not the heat sinks. great place to put a fan I thought, right over the plants, until lights out, at least, then another fan has to come on. For the active I have two fans on he heats sinks, and another fan for circulating fresh air. Never too many fans I guess. I find the additional colors help more than active/pasive approach.

I have been testing colors, and I have found royal blue seems to stimulate vegetative growth. peace.