Question About LED's Efficiency

[Friday]

Can someone correct me, because I'm pretty sure I'm wrong about this part of LED technology.

If you were designing an LED grow light for plants, you want to get maximum lumens from the minimum watts.

Now, isn't it true that many LED's produce light more efficiently if they're underpowered? (Both colored and white)

If that's true, then would the most efficient grow light (the best lumen-per-watt ratio) be one that was composed of a collection of under-driven LED's . . . such as 100 of the 1-watt LED's being driven by only 50 watts?

I guess "efficiency" would include the cost of the components -- but wouldn't that objection be over-shadowed by long-term cost efficiency??? I mean, it'll cost much more to use 100 one-watters and drive them at 50%, than it would be to buy 50 one-watters and drive them at 100%. But after a year of use, wouldn't the cost efficiency make up for it?

Thanks for helping me with this one. Sorry if it's a newbie question. We all have to start somewhere!

~jessie

 

[AnAppleSnail]

Can someone correct me, because I'm pretty sure I'm wrong about this part of LED technology.

If you were designing an LED grow light for plants, you want to get maximum lumens from the minimum watts.

Now, isn't it true that many LED's produce light more efficiently if they're underpowered? (Both colored and white)

If that's true, then would the most efficient grow light (the best lumen-per-watt ratio) be one that was composed of a collection of under-driven LED's . . . such as 100 of the 1-watt LED's being driven by only 50 watts?

I guess "efficiency" would include the cost of the components -- but wouldn't that objection be over-shadowed by long-term cost efficiency??? I mean, it'll cost much more to use 100 one-watters and drive them at 50%, than it would be to buy 50 one-watters and drive them at 100%. But after a year of use, wouldn't the cost efficiency make up for it?

Thanks for helping me with this one. Sorry if it's a newbie question. We all have to start somewhere!

~jessie

You don't just want lumens - you want lumens that can be photosynthetically used. There are a few efficiencies involved:

Electricity into light
light into photosynthetic light
photosynthetic light into photosynthesis

Plant research is currently being done to isolate the peaks, but you want high output in the wavelengths plants benefit from most - the rest is less useful.

Any spend-more-now-to-save-later question is an economics one. A Cree XP-G R4 costs about $8, so I can spend $400 or $800 in your example. Driven at max (1.5 amp), each gets about 325% * 130 lumens = 420 lumens. Driven at half max (750 mA), you get about 260 lumens, 61.9% of the output per LED. What's the power savings?

750 mA. Vf = 3.2. W = 0.75*3.2 = 2.4W
1500 mA. Vf = 3.4 W = 5.1W

100 * 2.4W = 240W
50 * 5.1W = 255W

I could calculate the 50% photon power level but it wouldn't make much difference, the 100-LED set would be using somewhere around 2.35W per LED.

That's only a 15W difference in power use, assuming passive cooling. How long will it take to earn $400 by saving 15W per hour (with interest, because money spent NOW is worth more than money saved later)? That's a LONG time - every 2.7 days you'll save about a kilowatt - those cost a few cents?

Also, you'd need twice the mounting surface, although cooling under-driven LEDs is easier.

 

[Friday]

Hi AnAppleSnail,

Thanks for the post! I know about the photosynthetically available radiation (PAR), and that's exactly why LED's are interesting as grow lamps. You can target specific wavelengths, and just hit those -- and not waste light in the spectrum the plant cant's use.

Anyway, you answered my question so expertly that I'd love it if you could answer a quick follow-up question!

Let's say I wanted to build a lamp composed completely of just Red 630nm LED's. I'd like it to draw somewhere in the vicinity of 150 watts, and I'd like to keep the budget under $300 (if possible). What would be an efficient arrangement of LED's and drivers?

From my reading, it appears that the 1-watt LED's are more efficient than the 2w-5w varieties, so I'll assume that I'd want about 175 x 1w LED's. But that's as far as my knowledge will take me.

Once I find what arrangement of LED's and drivers would work well for this, I can take it from there.

Thanks so much (in advance) for help here! It's hard finding reliable sources of up-to-date information on a technology that's so rapidly changing.

You don't just want lumens - you want lumens that can be photosynthetically used. There are a few efficiencies involved:

Electricity into light
light into photosynthetic light
photosynthetic light into photosynthesis

Plant research is currently being done to isolate the peaks, but you want high output in the wavelengths plants benefit from most - the rest is less useful.

Any spend-more-now-to-save-later question is an economics one. A Cree XP-G R4 costs about $8, so I can spend $400 or $800 in your example. Driven at max (1.5 amp), each gets about 325% * 130 lumens = 420 lumens. Driven at half max (750 mA), you get about 260 lumens, 61.9% of the output per LED. What's the power savings?

750 mA. Vf = 3.2. W = 0.75*3.2 = 2.4W
1500 mA. Vf = 3.4 W = 5.1W

100 * 2.4W = 240W
50 * 5.1W = 255W

I could calculate the 50% photon power level but it wouldn't make much difference, the 100-LED set would be using somewhere around 2.35W per LED.

That's only a 15W difference in power use, assuming passive cooling. How long will it take to earn $400 by saving 15W per hour (with interest, because money spent NOW is worth more than money saved later)? That's a LONG time - every 2.7 days you'll save about a kilowatt - those cost a few cents?

Also, you'd need twice the mounting surface, although cooling under-driven LEDs is easier.

 

[blasterman]

You haven't said what type of LEDs you're able to work with. Are you talking about stars, or just the LEDs themselves which will require re-flow gear, etc.

Also, efficiency and wattage don't go hand in hand. I use the Satistronics 3watt reds for big lights because they are cheap (less than $2.00 per 3watt) but they aren't as efficient as others below...maybe 25lumens per watt. Still, you can build a pretty big light for the cheapest investment this way.

Star mounted Red Crees and Rebels come around 35-40 lumens per watt in common flux ratings, and can be driven up to 3watts with a bit of efficieny loss. Typically they cost around $6.00 if you shop around.

We had a thread on the most efficient readily available red a few months ago. As I recall Osram Golden Dragons lead in that category (some here at Mouser), but not sure if these can be mounted without reflow gear.

 

[AnAppleSnail]

Hi AnAppleSnail,

Thanks for the post! I know about the photosynthetically available radiation (PAR), and that's exactly why LED's are interesting as grow lamps. You can target specific wavelengths, and just hit those -- and not waste light in the spectrum the plant cant's use.

Anyway, you answered my question so expertly that I'd love it if you could answer a quick follow-up question!

Let's say I wanted to build a lamp composed completely of just Red 630nm LED's. I'd like it to draw somewhere in the vicinity of 150 watts, and I'd like to keep the budget under $300 (if possible). What would be an efficient arrangement of LED's and drivers?

From my reading, it appears that the 1-watt LED's are more efficient than the 2w-5w varieties, so I'll assume that I'd want about 175 x 1w LED's. But that's as far as my knowledge will take me.

Once I find what arrangement of LED's and drivers would work well for this, I can take it from there.

Thanks so much (in advance) for help here! It's hard finding reliable sources of up-to-date information on a technology that's so rapidly changing.

Broadly speaking, a particular LED will be less efficient at higher currents (This is true down to about 20 mA). But between different LED structures and brands that may not be the case. And you're not just looking at lumens/watt, but also lumens/$.

Planning a fixture requires picking the LED pretty early on. A 300-watt LED light will be really bright, but also require the rejection of around 200 watts of heat. That'll take a lot of heatsinking, which is going to represent a lot of your investment (unless you thermal epoxy or solder LEDs to adequate scrap metal you've got lying around).

 

[qwertyydude]

All the hydroponics guys use high pressure sodium lights because of their efficiency and cost benefit. Getting 16,000 lumens out of 150 power led's driven at 1 watt each would be expensive and lumen maintenance has to be considered there also. Whereas 1x150w HPS grow light has similar efficiency, 16,000 lumes at 150w. That's 109 lumens/watt at exactly the right wavelengths plants absorb so little wasted light. You'd need a much more significant lumen efficiency improvement to justify led's in this matter.

 

[blasterman]

Actually HPS does not emit at ideal wavelengths for plants, but it's in the curve. HPS is also a very,VERY efficient light source for yellow/orange, and cheap. There's no question common LED based lamps have a hard time keeping up with the brute fire-power of HPS.

Then again, one advantage with LEDs is you spread them out on rails and cover a much wider area with little extra cost. Most high wattage LED growth lamps are very square and try to compete with a HID fixtures in terms of density, which I find a bit silly. By spreading the LEDs on rails or C-channel you can virtually eliminate heat, get the lights much closer to the plants, get more side coverage, and not have to deal with fixture loss.