Heatsinks for DIY LED lamps

churchhaze

Well-Known Member
I've seen a lot of temperature readout modules, but still not the obvious.

Instead of a display for outputting temperature and a knob for controlling fan speed, why not directly tie fan speed to temperature?

Taken a step further, what's with all the fancy displays and interfaces? Can't you just touch the sink to see if it's hot? The price of a potentiometer and an aluminum knob alone is more than the price it would take to setup the entire thermostat system to outright regulate temperature.

Adding unnecessary parts to a device that serves essentially one function is actually against my principles as a designer. It's cool, but only a novelty. Simplicity is real beauty.
 
Last edited:

churchhaze

Well-Known Member
It depends what you consider low power. Some people tend to drive their cobs harder and call it underdriving. :P

Also, I think the real hard limit to passive cooling is the ceiling space you can fit heatsink. Has anyone here actually got the largest profile with the tallest fins and used it to cover their entire ceiling, no gaps? That's what I think the real limit is.

If I remember correctly, @SupraSPL is running his 10.080" profile sinks totally passive and underdriving, and still not covering his whole ceiling. (If he did, I'd imagine he'd be able to cool enough leds passively to burn the floor with radiant power)

Passive cooled with zero fan intervention would be limited to lower powered setups I'd think.
 
Last edited:

Positivity

Well-Known Member
I've seen a lot of temperature readout modules, but still not the obvious.

Instead of a display for outputting temperature and a knob for controlling fan speed, why not directly tie fan speed to temperature?

Taken a step further, what's with all the fancy displays and interfaces? Can't you just touch the sink to see if it's hot? The price of a potentiometer and an aluminum knob alone is more than the price it would take to setup the entire thermostat system to outright regulate temperature.

Adding unnecessary parts to a device that serves essentially one function is actually against my principles as a designer. It's cool, but only a novelty. Simplicity is real beauty.
Having a arduino control temperature and fans would require programming skills I don't have right now. Not really the time to learn it now either.

A $25 fan speed controller and temp reader coupled with a $10 ammeter isn't novelty at all. Sure I can have R2D2 do it for me....but given the information I'm getting I can dial the fans where they need to be on my own. Instead of guessing what my led junction temperature is I know what temperature it is....instead of guessing what the fans are contributing and how fast they need to spin...I know how fast they need to spin.

The features these gimmicky products bring takes off the blinders and allow you to see what's really going on...:razz:

Best $35 I've spent yet..

I think my fan controller has a auto setting too...haven't got there yet..;-)
 
Last edited:

Positivity

Well-Known Member
It depends what you consider low power. Some people tend to drive their cobs harder and call it underdriving. :P

Also, I think the real hard limit to passive cooling is the ceiling space you can fit heatsink. Has anyone here actually got the largest profile with the tallest fins and used it to cover their entire ceiling, no gaps? That's what I think the real limit is.

If I remember correctly, @SupraSPL is running his 10.080" profile sinks totally passive and underdriving, and still not covering his whole ceiling. (If he did, I'd imagine he'd burn the floor with radiant power)

That's more heat sink and weight than I ever want to deal with. I want them as light and minimal as possible. Pretty sure supra was using the supply of sinks he already had. A artic 11 would weigh a fraction and cost a fraction and run cooler...
 

bicit

Well-Known Member
There are pros and cons to both methods. Everyone is going to have a different grow setup. It's evident that the industry and most DIY'ers, not all of course, incorporate fans into their fixtures for a reason and it appears to be the way to go. Although according to Supra's claim, they may have overlooked something in the process...



Running a fan at ~2W every day for ~15 hours over a 90-period would amount to ~$0.30 @$0.12 kWh for a single grow. A single year of use, with 4x grows, would be around $1.20 in electrical costs (keeping in mind that the cost of electricity rises over time). Multiply that by additional fans and the cost in electricity increases even more so.

The question follows: Is better to pay for the larger upfront cost of a bulkier HeatSink, while running fewer COBS and at lower currents or to purchase small slabs of aluminum and run more COBs per area at higher currents despite the annual charge in paying for those fans to operate?

Of course this question requires intensive research and knowledge of any of the given variables in order to accurately answer it but it ultimately comes down to the numbers clicking.

I don't doubt your response Bicit, as I totally agree with your points and that there isn't just one answer. We should all be open to the different ideas and configurations of COB fixtures, whether it be active or passive cooling, including myself.

It's not a matter of cost, it's a matter of ultimate efficiency and complexity. Even assuming no power losses the arctic alpine cpu cooler draws 2.6 watts of power. This means that the amount of power draw by the system overall has been increased by 2.6watts. If we use this to cool a 3000K, 80cri vero 29 running at 38.2 watts we need to recalculate the efficiency (lm/watt) value of the LED. For simplicity let's say that with the fan it cools the case temp to 25*C. For the emitter alone at 38.2 watts bridgelux rates it at 5300lm or 138 lm/w. However if it is actively cooled the system would now draw 40.8 watts of power, 129.9 lm/w, or 41.2% efficient.

By contrast let's look at a poorly sized passively cooled system that runs with a case temp of 85*c. Bridgelux lists the output as 4666 lm, 122.1lm/w or 38.8% efficient. A meager 2.4% gain in efficiency for the actively cooled system. These numbers were run in a vacuum assuming no driver losses. Most 12v power supplies have poor ratings on efficiency compared to LED drivers so in the real world these differences would be almost insignificant.

More failure points, more moving parts, more wires, just more of everything. Passive is much simpler.

Active cooling works best when running higher currents or when the power draw of one fan can be displaced over multiple emitters. Passive works best for simple systems. Cost would be almost negligible except on a large scale.

Also, just a note. Supra is a chaser of ultimate system efficiency regardless of cost or complexity. He's good at math.
 

AquariusPanta

Well-Known Member
Having a arduino control temperature and fans would require programming skills I don't have right now. Not really the time to learn it now either.

A $25 fan speed controller and temp reader coupled with a $10 ammeter isn't novelty at all. Sure I can have R2D2 do it for me....but given the information I'm getting I can dial the fans where they need to be on my own. Instead of guessing what my led junction temperature is I know what temperature it is....instead of guessing what the fans are contributing and how fast they need to spin...I know how fast they need to spin.

The features these gimmicky products bring take off the blinders and allow you to see what's really going on...:razz:

Best $35 I've spent yet..

I think my fan controller has a auto setting too...haven't got there yet..;-)
I'm digging my $17 fan speed controller (eBay). My ears would be bleeding without ;). The temperature reader was a must - without it, I wouldn't know where to set my speed controller (medium).

lol "R2D2"
 

AquariusPanta

Well-Known Member
It's not a matter of cost, it's a matter of ultimate efficiency and complexity. Even assuming no power losses the arctic alpine cpu cooler draws 2.6 watts of power. This means that the amount of power draw by the system overall has been increased by 2.6watts. If we use this to cool a 3000K, 80cri vero 29 running at 38.2 watts we need to recalculate the efficiency (lm/watt) value of the LED. For simplicity let's say that with the fan it cools the case temp to 25*C. For the emitter alone at 38.2 watts bridgelux rates it at 5300lm or 138 lm/w. However if it is actively cooled the system would now draw 40.8 watts of power, 129.9 lm/w, or 41.2% efficient.

By contrast let's look at a poorly sized passively cooled system that runs with a case temp of 85*c. Bridgelux lists the output as 4666 lm, 122.1lm/w or 38.8% efficient. A meager 2.4% gain in efficiency for the actively cooled system. These numbers were run in a vacuum assuming no driver losses. Most 12v power supplies have poor ratings on efficiency compared to LED drivers so in the real world these differences would be almost insignificant.

More failure points, more moving parts, more wires, just more of everything. Passive is much simpler.

Active cooling works best when running higher currents or when the power draw of one fan can be displaced over multiple emitters. Passive works best for simple systems. Cost would be almost negligible except on a large scale.

Also, just a note. Supra is a chaser of ultimate system efficiency regardless of cost or complexity. He's good at math.
I, with many others, would argue against cost not being a factor. I mean, this is a thread for DIYer's yes? (smiles)

I would tend to agree that Supra is a freak about efficiency, to the point that it frightens me (smiles again).

Aldous Huxley: "The worst enemy of life, freedom and the common decencies is total anarchy; their second worst enemy is total efficiency"
 
Last edited:

bicit

Well-Known Member
I, with many others, would argue against cost not being a factor. I mean, this is a thread for DIYer's yes? (smiles)

I would tend to agree that Supra is a freak about efficiency, to the point that it frightens me (smiles again).
The point being that one has to consider the skill set of the builder and the desired application. Since the efficiency between an active and passively cooled system is negligible assuming both are properly designed. Some people like playing with big heavy heatsinks and keeping the wiring simple. Others like messing around with extra wires and parts to keep the size and weight of the heatsink down.

No 'best' way to do it since what constitutes as 'best' is relative to the situation.
 

AquariusPanta

Well-Known Member
The point being that one has to consider the skill set of the builder and the desired application. Since the efficiency between an active and passively cooled system is negligible assuming both are properly designed. Some people like playing with big heavy heatsinks and keeping the wiring simple. Others like messing around with extra wires and parts to keep the size and weight of the heatsink down.

No 'best' way to do it since what constitutes as 'best' is relative to the situation.
I've been claiming that which you type, specifically that it's all dependable on the setup and environment. Glad we could agree on something ;)

Is anyone else noticing the high-levels of entropy found on this thread? jfc... lol
 

churchhaze

Well-Known Member
The thing is that he was specifically referring to "a poorly sized passively cooled system that runs with a case temp of 85*c".

What about a properly sized passively cooled system? What temperature would that run at?

A "poorly sized passively cooled system" is what I'd call semi passive, like my 80W UFO killer design at 700mA. A slight breeze brings it below skin temperature, but no breeze and it gets to about 70-80C. The same design at 500mA is cool to the touch even without the fan. That makes it a properly sized passive sink at 500mA. I turned the fan off because it wasn't making an appreciable difference in the heat sink's temperature. What's the point of using 2W to lower temperature by 1-2C?


Okay....

So you run passive with LEDs running at 80*c.

Better turn your fan up to evacuate that heat..bye bye efficiency. Pretty sure a inline uses more watts than a dc case fan..
 

bicit

Well-Known Member
The thing is that he was specifically referring to "a poorly sized passively cooled system that runs with a case temp of 85*c".

What about a properly sized passively cooled system? What temperature would that run at?

A "poorly sized passively cooled system" is what I'd call semi passive, like my 80W UFO killer design at 700mA. A slight breeze brings it below skin temperature, but no breeze and it gets to about 70-80C. The same design at 500mA is cool to the touch even without the fan. That makes it a properly sized passive sink at 500mA. I turned the fan off because it wasn't making an appreciable difference in the heat sink's temperature. What's the point of using 2W to lower temperature by 1-2C?
IIRC supra recommended 110cm^2 of surface area per watt of power in order to achievement a case temp of 40-50*C. However I could be mistaken on what the final temps would be. That's why I estimated a 'worse case scenario', rather than speculating.
 

churchhaze

Well-Known Member
Not so unrelated to this thread, I realized a few weeks ago that my 120mm CPU heatsink fan (on an actual CPU) had been stopped for some time. I got it unstuck and it started turning, and making tons of noise, and there was tons of lint as usual (I clean inside there regularly).

I took the fan out and all the plastic trim meant to guide the flow of air out. I figure it had been working all that time without a fan, why bother getting a new one. The CPU has been totally stable, and the sink doesn't get very hot.

I still wish I could find one of those lint free active cooling setups that don't have an inch of lint every time I go to check it out. The thinner and denser the fins, the more lint you will collect. I wish that was a myth.
 

churchhaze

Well-Known Member
The space between a purely active heat sink and it's fan is like a cloths dryer lint collector. It collects soooo much dust. I've never had an active design where that wasn't the case.

Also, after the years, the fans blades will get coated in lint and gunk. It's not the worst thing in the world, but it's far from dust free.
 

guod

Well-Known Member
...
I took the fan out and all the plastic trim meant to guide the flow of air out. I figure it had been working all that time without a fan, why bother getting a new one.
The CPU has been totally stable, and the sink doesn't get very hot.

.
Dynamic frequency scaling (also known as CPU throttling) is a technique in computer architecture whereby the frequency of a microprocessor can be automatically adjusted "on the fly," either to conserve power or to reduce the amount of heat generated by the chip. Dynamic frequency scaling is commonly used in laptops and other mobile devices, where energy comes from a battery and thus is limited. It is also used in quiet computing settings and to decrease energy and cooling costs for lightly loaded machines. Less heat output, in turn, allows the system cooling fans to be throttled down or turned off, reducing noise levels and further decreasing power consumption. It is also used for reducing heat in insufficiently cooled systems when the temperature reaches a certain threshold, such as in poorly cooled overclocked systems.

 

indianajones

Well-Known Member
when i get these built and running with 6x vero 18 i will take measurements
and report back.

20141128_222052.jpg

you can see the glass broke on the unit to the left during shipping. huge fucking
pain in the ass to get it out of the plastic wrap without getting glass shards all
over the place. these cost me about $150 a piece after shipping and will cool
~180-190w, they are rated for 200w.

20141130_195321.jpg
 
Top