(2) Fully controllable 100w Leds on the Cheap.
- Thread starter eznet2u
- Start date
Sorry, but I don't understand your 50% statement. A 100W chip needs a cooler capable of cooling min. 100W. Out side of the chips viewing angle the efficiency is cut by 50% but at the chip its self, contact point to the cooler should be figured at 100W. On this 100W Epistar chip http://www.ebay.com/itm/EPISTAR-100W-Super-Actinic-Blue-Hybrid-Led-Panel-Aquarium-/220926012078#ht_1605wt_700 it can actually be pushed up to a max of 252W depending on the current you apply to it. The cooler would need to be figured for a cooling capacity to the current x voltage = W you want to run your chip at. These coolers actually tell their wattage cooling capacity http://www.arctic.ac/en/p/cooling/vga.html. All that being said, there is a ton of multichip DYI build info on this thread http://www.reefcentral.com/forums/showthread.php?t=2128756 I hope this is helpful, and I am looking forward to seeing eznet2u's build pics. Good luck to all.---Rick
I'm no scientist but of the 100w some of that energy is dissipated as heat, some as light. If the chip was 100% efficient than all the energy would be dissipated as light and therefore the chip would run cool by itself. Since we know it puts out light, and we know it is not 100% efficient, then we know the amount of heat we need to dissipate is < 100w. This applies independently to the amount the power you supply the light with.
Thanks for the thread link! More reading woohoo!
Thanks for the thread link! More reading woohoo!
BeanAnimal
Premium Member
No... The LED array converts electricity to light. What is NOT directly converted to light is converted to HEAT at the back of the die. The light leaves the LED array as energy and is NOT converted to HEAT in the LED array (though it turns to heat somewhere in the tank, room, etc)
That is the 100W LED puts 100W of "heat" into the environment, as it does not create, destroy or store energy, it only converts it from one form to another.
If we assume 50% efficiency, that means 50% light (radiation) and 50% direct (convection, conduction) heat. So the heatsink needs to be able to handle 50W or so.
This has nothing to do with viewing angle or the efficiency of the lens system, it is just basic physics (the law of conservation of energy). A watt is a watt, no matter what form of energy it is converted to, it is still the SAME amount of energy.
I guess I don't understand the 50% efficiency. Are you saying you are actually only ending up with 50W of light emitted from a 100w led? and then 50w in heat generated? Thats completely contradictory to the 1181 posts in the Multichip build thread that I have read. In that thread the consensus is to size your cooling device capability, to the wattage of the chip you are going to use. In fact the recommendation is for over sizing. If your only getting 50% efficiency from a chip I will not be using leds as a light source.---Rick
BeanAnimal
Premium Member
I guess I should have been more clear, instead of just repeating the example numbers given and not discerning between efficiency and efficacy.
In the above conversation it was assumed that there was a 50% efficiency.... Reality, those 100W chips are likely far less efficient than that. I would guess 85W of heat to 15W of light, seeing that a cutting edge Cree is only 70/30 (maybe 65/35) heat to light.
Please note (more below) that we need to distinguish between Efficacy and Efficiency, as the two are often confused! Those 100W chips may have an efficacy of 50%...
Respectfully (and not singling any person out) most of the folks building anything electrical are fairly ignorant of the laws of physics and simply follow the heard. That is not a bad thing, it is just reality and dictates that many may stumble upon the right answer via the wrong understanding
In other words, most folks can follow instructions and end up with a working project, but very few of them understand the details of how and why it works.
That may not be a bad thing based on the fact that there are so many variables with regard to the actual performance of the heatsink itself, including thermal coupling ot the driver, airflow, orientation, surface area, etc and the fact that these chips are not likely as efficient as we think. In the end the [our] goal is not to use physics to optimize the heatsink to maximize cooling per surface area and materials cost, it is to keep our LEDs from overheating at "any cost".
So, the "rule of thumb" being followed is sound, even though the logic used to derive it is questionable
Ignoring losses in the reflectors, lenses, etc, the amount of light produced by a particular technology is easy to measure and a matter of fact and record.
Please study the chart below to understand how different technologies convert energy: This is the overall EFFICIENCY of the light source (often confused, with EFFICACY). This is a 2010 table representing the current state of technology. It may have improved a bit in recent months, but not much.
http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/thermal_mgt_white_leds.pdf
You will notice that THIS table has been dropped from CURRENT DOE publications because it does not convey the message that the govt wants pushed, even though it is factual.
Instead the recent publications have switched to using EFFICACY.
http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/energy_efficiency_white_leds.pdf
The difference? The tables are for the same light sources, but analyze the light output in a different way.
luminous efficacy - is the ratio between the total luminous flux emitted by a device and the total amount of input power it consumes adjusted by the luminosity function. In other words, it denotes the ratio of light to the theoretical maximum for the spectrum in question, not the overall radiant output. In our case "white light".
luminous efficiency - is the total amount of radiant energy compared to that lost as heat (sound whatever, is not radiant energy).
So why list efficacy instead of efficiency? First of all, it addresses the light we are concerned about (the spectral output in question) and secondly (and more importantly) the numbers LOOK far better to the casual observer ignorant of the physics
Do YOU want to use LEDs? Lets fully ignore the debate regarding spectral output with regard to aesthetics or the ability to grow coral. The efficiency/efficacy of a particular LED can be from very good to VERY poor. Some of the cutting edge products from Cree, Phillips, Osram, etc are certainly very efficient and getting better each year. There is another whole world of products that are marginal to crappy at best. Many (most) of the China knock-off LEDs have fairly dismal efficiency numbers. They also cost far less. I suppose it is up to the end user to determine what the "bang for the buck" is and if the cheapy LEDs will outperform the existing MH or T5 and/or offer value over the cutting edge offerings.
Have fun
I see said the blind man. Thank you for that input.---- reality and dictates that many may stumble upon the right answer via the wrong understanding
In other words, most folks can follow instructions and end up with a working project, but very few of them understand the details of how and why it works.
I must say I fall into this group. I read to learn, but if I need surgery I leave it up to the surgeon. Soooo how do you figure size in the amount of cooling needed for given LED's? If A 100W led can run on a 50W cpu cost wise, is less than buying a 100W capable cooler. Is there some formula in lay mans terms? Thanks---Rick
In other words, most folks can follow instructions and end up with a working project, but very few of them understand the details of how and why it works.
I must say I fall into this group. I read to learn, but if I need surgery I leave it up to the surgeon. Soooo how do you figure size in the amount of cooling needed for given LED's? If A 100W led can run on a 50W cpu cost wise, is less than buying a 100W capable cooler. Is there some formula in lay mans terms? Thanks---Rick
BeanAnimal
Premium Member
Most of us find ourselves in the same bout at one time or another... we are all ignorant of something, most of us many things
There are certainly formulas for sizing heatsinks, but most are well beyond easy explanation and are very heavy on math and thermodynamics. If you were designing a product for sale, you would want to do the hard work in order to minimize the heatsink footprint and/or cost, etc. For the rest of us, we use trial and error to get where we need to be. You will find, that ther trial and error method produces "rules of thumb" optimizes or not, that the heard can follow
There are certainly formulas for sizing heatsinks, but most are well beyond easy explanation and are very heavy on math and thermodynamics. If you were designing a product for sale, you would want to do the hard work in order to minimize the heatsink footprint and/or cost, etc. For the rest of us, we use trial and error to get where we need to be. You will find, that ther trial and error method produces "rules of thumb" optimizes or not, that the heard can follow
mussel and hate
New member
One thing to remember is efficacy for coral growth differs from human vision. MH has a bit of an advantage in that department vs "white" LEDs from near UV output in ~380-420nm range.
I'm glad the DOE changed the charts. The first one seems to violate the law of conservation of matter and energy.
I'm glad the DOE changed the charts. The first one seems to violate the law of conservation of matter and energy.
BeanAnimal
Premium Member
Only if you are not reading it correctly. The chart shows conduction, convection and radiation. It breaks down the total radiation into visible, IR, and UV. 1 unit (watt) in, 1 unit (watt) out. The law of conservation is in tact
mussel and hate
New member
It does seem to total up to 100% at the bottom. I must have been reading that wrong.
updates?
updates?
What kind of mounting solution have you come up with for that heatsink? That is to say, how will it be suspended over the tank? I see that on the Gemini II, there are some handy extra holes that could be used to mount it to an aluminum C channel or L. :thumbsup:
Any more updates on this build?
updates?
What kind of mounting solution have you come up with for that heatsink? That is to say, how will it be suspended over the tank? I see that on the Gemini II, there are some handy extra holes that could be used to mount it to an aluminum C channel or L. :thumbsup:
Any more updates on this build?
Wish the OP would update! I had no idea it took a month for stuff to arrive from China and I'm jones'ng to see how his comes together.
The TDP of a Geminii is 150W so I think it should be good to go BTW. I tried using a rapidled hanging kit and the screw holes aren't big enough so they have to be drilled out. You could also just thread the wire through and crimp it too.
I also see that the heatsink contact plate is smaller than the LED's base so I'm anxious to see what the op does about that- not sure you have to do anything I suppose.
The TDP of a Geminii is 150W so I think it should be good to go BTW. I tried using a rapidled hanging kit and the screw holes aren't big enough so they have to be drilled out. You could also just thread the wire through and crimp it too.
I also see that the heatsink contact plate is smaller than the LED's base so I'm anxious to see what the op does about that- not sure you have to do anything I suppose.
Stu Pickles
New member
what was the outcome?
DetroitReefer
New member
I live in Detroit. I placed an order on the 4th of September, and my package just landed in New York on the 16th or September.
I found a FANTASTIC company in Taiwan though. Ordered a lot of LED multi-chips last Wednesday, and they arrived Monday. I paid for DHL shipping $20), but not waiting a month made it very worth while.
Similar threads
