UPDATET: How many Lumen does a High Power LED have?

[Mad Scientist]

You obviously didn't understand what I wrote, let me try explaining again:

Watts is still a measure of electricity.

Calling it radiant watts only tells you how much juice that LED uses to make that much light in that specific wave length.

NO. Watt is the SI unit for POWER. Energy per time, simple as that. And that's not exclusive to electricity:

When dealing with electromagnetic radiation (let's say: light) - physicists call it radiant power or radiant flux - which is measured in watt. You could also use "radiant energy", doesn't really make a difference if you're using joule or watt here, as long as you keep time constant in your calculation.

Lumen on the other hand is a measure of how bright something looks to the human eye. It's calculated from radiant energy. And since the luminosity-function i linked above is the base for said calculation, a watt of 680nm light has less lumen than a watt of 650 nm light. And that was the point I was trying to explain above: Lumen is not a suitable unit for comparing light output from two different-wavelength sources (especially for wavelengths at the edge of the luminosity curve). Same goes for lux, since it's just lumen/m[sup]2[/sup].

You have no idea what percentage your LEDs are dominant by.

Your 650nm might only be 51% dominant while the 680nm might be 80%
....
No LED actually only puts out 1 wave length.

LEDs don't just have 1 wavelength, that's right. The way lumen are calculated is by splitting the spectrum into small wavelength portions, calculating the area for these portions and multiplying that value with the value from the luminosity function. Then you add up all those values and you get your lumen-value - it's called numerical integration. That's exactly what LED manufacturers do to calculate the lumen values in the datasheets.
And you can do this too, if you have the spectrum from the datasheet and the lumen value, you can calculate radiant power.

No LED is the same.
You can probably get different results from 3 different 650nm LEDs and 3 different 680nm LEDs.
They vary a LOT even made by the same company ,They vary a lot from Lot to Lot .

Of course there's variation, but that applies to all LEDs and the variation is specified in the datasheet. If you want to buy LEDs of a certain binning, with some manufacturers you can, but it will usually cost you extra. Other manufacturers don't give you the option but you know the variation in advance, it's all in the datasheets. There are manufacturers who aren't quite honest in their datasheets and use all sorts of tricks to make their LEDs look better on paper, all you can do is steer clear of those.

No matter what any grow light vendor says it's only going to be what they think will sell their product,Period.
99% of it is out of context or pure BS.

They can print all kinds of links/articles,whatever, like above ,but the fact is,I'd bet none actually know anywhere near the actual LED stats on the stuff they actually buy.
They order whatever LED in wave length,watt and Lumen they want.
They have no idea of the stats and probably could care less.50% or 80% dominant all look the same to the customers eye.
They use whatever LED is cost effective for them to turn a profit.

The more tech looking crap they can put out of context in their adds,the better...

I'm sure there's legitimate LED-Panel manufacturers out there. But yes: I agree with you, many of them will just put together the cheapest possible panel and then try to sell it with lots of pseudo-scientific marketing BS. And advertising with lumen-values is an example of that BS.

There's lots of people out there trying to build their panels based on all the marketing-crap they're reading. That's why I'm explaining the science here.

 

[smokemaster]

The watts on the data sheets I've ever gotten from suppliers means electricity...

The data sheets tell soldering and running temps.forward and reverse current,max. and min. volts , ma. ,wave length and lumens...

What the LED is supposedly rated at by the factory.

As far as voltage at whatever MA and what voltages are supposed to put out the wave lengths that LED tested out at.
And especially for Blue LEDS the data on a lot of data sheets is totally off.

Most data sheets either over drive the LED or run it at a lot higher power than they should be run at=very short life...

Some ,I think are actually giving data for pulsing the LED at levels you can't run that led at without pulsing them or you let the smoke out.
It makes the LED look better for sales purposes.
Some data sheets do say that you can pulse a 350ma LED at the same voltage but at 500ma to get more lumens.

From what I've read,plants don't see pulsed LEDs as light.

In plants light is basically a chemical reaction to specific wave lengths.

From what I've read,pulsed over driven LEDS can put out a lot more lumens than they are designed for by over driving them/pulsing them at higher ma.Say 500ma instead of 350ma.
Pulsing keeps LEDS from getting too hot from being overdriven-so I've read.
Though I think it would eventually catch up with you and still shorten the LEDs life.

Though sometimes they say to use voltages and ma. that don't make sense to me either...pulsed or not.
You can pulse/over drive them to a certain degree but not by what some of the data sheets say you can.

Pulsing LEDs is too fast and plants can't respond to it.


Never seen anyone give data on what you are talking about so far.
I've never seen any data on what percentage any LED is dominant by or what other wave lengths on either side of dominant the percentage that they are present in.
So you can only know that at whatever MA and voltage at least 50% of that LED is whatever wave length it's rated at.

And IF the sheet isn't bunk you might get an idea of maybe 2 of the other wave lengths also put out on either side but not in what percentage each is at 2 other voltage rates.
How many of the LEDS lumen output is other than the dominant color?

It seems since you are looking for specifics as far as I can see,why doesn't the portions of lumens of other wave lengths that are part of the whole amount of lumens said LED puts out matter?
It seems not knowing the amounts of each wave length of light an LED puts out would make it hard to even know what to divide the lumens up by to get your watts thingy figured out correctly.
You can only guess - say dominant is 50% and other wave lengths equal the other 50% combined.
Though depending on the LED you could get any % over 50% as dominant.


I still think the stuff you are talking about is pretty much useless to the DIYer because I don't see the data you need to actually know what any LED really puts out and even if you did,you would have to know what percentages of each wave length of plant usable light you need to put in your grow light to imitate the sun or what that specific plant needs in general.

I see the same data sheets for any specific LED on their sites for years.
I'm sure that there is a lot of variance between different lots of LEDs that you end up still guessing when you try and figure out your radiant watt thing.

One thing I see with my CD spectrometer is a lot of different colors are present in any LED i've looked at.

I see red,yellow,green,blue,purple orange etc. in different sized bands no matter what color the LED is dominant in.

Some have a lot of very thin bands with the dominant color being the thickest.Some have 1 thick band with only a couple thin bands of other colors.
I'm sure it depends on the quality of the LED as to how thick or thin any of the bands are.

But in general the data sheets I see are like a recipe for making meat loaf or something.
In general they get you a dinner but you may need to add or subtract ingredients(leaner meat,fresher spices) to get what works best.
Most times the LED data sheet only lets you know maybe what not to use or go over as far as running the LEDs.

Most times I have to put strings together of different LEDS and burn them a few days while putting a meter on them to see what they are actually using or whatever...how hot they get, brightness etc.

Usually I find out I have to run the LEDS at a different/lower ma. and or voltage or the smoke comes out after about 2 yrs of 16-18hr days,365days a year.
Data sheets aren't ,in general,what they are cracked up to be a lot/most of the time.

Just because an LED lights up it doesn't mean you aren't over driving it,or under driving it.
You can only go by your meter and hope the data sheet wasn't crap.

Figuring out making a panel can suck,I'd think figuring out what a ready made panel puts out would be near impossible outside of a lab.Unless of coarse you have that $4000-$5000. to spend on that gizmo...

 

[Mad Scientist]

As far as voltage at whatever MA and what voltages are supposed to put out the wave lengths that LED tested out at.
And especially for Blue LEDS the data on a lot of data sheets is totally off.

Don't buy LEDs from those manufacturers. There's alternatives.

Most data sheets either over drive the LED or run it at a lot higher power than they should be run at=very short life...

Some ,I think are actually giving data for pulsing the LED at levels you can't run that led at without pulsing them or you let the smoke out.

For the advertised values, they measure everything at the setting that produces the best numbers. The efficiency (lumen/watt) values are typically measured at 350 ma, since LEDs are more efficient at lower power. They typically advertise max. brightness values from higher power (typically 700 mA) - But all the numbers are in the datasheet, it's just a matter of looking at the actual data instead of the advertising, and use the numbers that are relevant to you. If you intend to drive your LEDs at 700ma, looking at the 350MA values obviously doesn't make sense.

Some data sheets do say that you can pulse a 350ma LED at the same voltage but at 500ma to get more lumens.

Not interesting for us hobbyists, since we simply use constant-current drivers to drive our LEDs. And that's also how the measurements for datasheets are done, at least for the datasheets I've seen.
There are companies out there, that try to sugarcoat their data, but typically by obtaining their numbers at lower temperatures (which you can tell by the fact that they don't specify heatsink-temps but chip-temps in their datasheets)

But Osram and Philips don't, their datasheets are legitimate and incidentally they also currently produce the most efficient LEDs. (And by currently I mean half a year ago, I haven't done any comparisons lately...)

I've never seen any data on what percentage any LED is dominant by or what other wave lengths on either side of dominant the percentage that they are present in.
So you can only know that at whatever MA and voltage at least 50% of that LED is whatever wave length it's rated at.

And IF the sheet isn't bunk you might get an idea of maybe 2 of the other wave lengths also put out on either side but not in what percentage each is at 2 other voltage rates.
How many of the LEDS lumen output is other than the dominant color?
It seems since you are looking for specifics as far as I can see,why doesn't the portions of lumens of other wave lengths that are part of the whole amount of lumens said LED puts out matter?

You don't just compare one wavelength. I just used a theoretical single-wavelenght LED in my above explanation because it's easier to explain the underlying principle that way. As you already realized, to calculate the actual numbers, you can't just look at the peak wavelength, you have to look at the spectral data (the curve in the datasheet) and do a numerical integration over all the wavelengths.

One thing I see with my CD spectrometer is a lot of different colors are present in any LED i've looked at.

I see red,yellow,green,blue,purple orange etc. in different sized bands no matter what color the LED is dominant in.

Yes, and the spectral curve in the datasheets plots just that: the relative intensity for every single wavelenght.

Most times I have to put strings together of different LEDS and burn them a few days while putting a meter on them to see what they are actually using or whatever...how hot they get, brightness etc.

Usually I find out I have to run the LEDS at a different/lower ma. and or voltage or the smoke comes out after about 2 yrs of 16-18hr days,365days a year.
Data sheets aren't ,in general,what they are cracked up to be a lot/most of the time.

Just because an LED lights up it doesn't mean you aren't over driving it,or under driving it.
You can only go by your meter and hope the data sheet wasn't crap.

Figuring out making a panel can suck,I'd think figuring out what a ready made panel puts out would be near impossible outside of a lab.Unless of coarse you have that $4000-$5000. to spend on that gizmo...

First things first: God, you seem to have seen lots of crap datasheets!

As for the "need a lab" part: To be honest, I don't know the going rate of spectral photometry equipment, and frankly I don't care since I'm lucky enough to have a friend who does the measurements for me.

 

[smokemaster]

Would you call this a good data sheet?

http://www.alldatasheet.com/datasheet-pdf/pdf/227038/OPTEK/OVSPBCCR8.html

I think it's pretty much useless.
All it tells you is the LEDs are 1 watt and run on up to 300ma.

You have to decide for yourself if you want to run them at 300ma and lesser voltages or a combination of lower voltages and lower ma.
Gotta keep the LED from cooking itself.
In general I run my 1 watt LEDS at around 280ma and 1.9v - 3volts.
Red lower than blues in general.

If you just run your LEDS by the data sheet you'll cook a lot of LEDs.

Since everything is based on running a 1 watt LED on 300ma. at a minimum of 3.2volts up to 4.4volts (.96watts up to 1.32watts) using their test voltages, it is running the LED in the tests at almost 1 watt to over 1 watt.

The data might be factual but it is of no use to some person who wants a long life for that LED.

Changing MA and Volts changes the output of the LED, thus making the rest of the stuff based on un realistic data as far as use in real life.
99% of the data sheets I see for 1 watt LEDs is just like this one.

The few I've looked at for the 3 watt LEDs I have are very similar.
About the same voltages but higher MA.
Actually I think a lot of the sheets are for the same LED group just run at different MA and volts so they test different.

 

[Mad Scientist]

OK, but could be better:

The std-Bins diagram at the bottom of page 2 is quite confusing. Also it doesn't tell you what distribution you could expect in a standard-order. (As I mentioned before, there's retailers who buy large batches, sort by BIN and sell them separately - but you have to pay premium for that)

The other table is OK, tells you the measurements were obtained at 25°C PCB-Temp (not junction temperature which is important)

Page 3 is pretty standard: first diagram tells you what voltage drop to expect for currents between 100 and 500 mA, third one tells you the spectrum of the 470nm-peak one (which is specified as typical)- you'd use this to calculate radiant flux. - This one could be better - there's a lot of unused space on both sides, which makes digitizing less precise. But I've seen worse.

On the last page there's just the dimensions, usually there's also temperature curves for reflow-soldering.

So - could be better, but all in all OK. - Hook it up to a constant current driver at 300mA, that provides at least 4.4 V and use a proper heat-sink and you should be OK, is how i read it. Those datasheets are targeted at engineers, so I won't blame anyone but myself if I'm unable to deduce if those LEDs are intended for long-period use. When in doubt: Ask the pros at an electronics or LED-forum if what you intend to do makes sense.

 

[smokemaster]

Kinda like-
The operation was a success , but the patient died.
Stupid patient was sick...shouldn't have had the operation....

 

[Mad Scientist]

Good analogy, let me use it in a different way:
Patient dead: Probably should've gone to a hospital instead of performing an appendectomy on himself. - Can't expect from a datasheet to turn you into an engineer.

 

[smokemaster]

Maybe the doctors shouldn't have given the guy the instruction book for doing the operation while saying it was an easy operation to perform...just follow the instructions and you'll be fine,we do it all the time...

My only problem is a lot of the engineers who are posting LED info on the net in places like here(not specific LED/electronics sites or whatever), all fail to tell average people that they can't run their LEDs according to the data sheets for very long or that when they change the voltage and or the current it also changes the lumens and wave length to a certain degree.

Posting about all the gizmos and do dads to people who aren't electrical engineers here just muddies up the waters.

It's fine on sites that are geared for electrical engineers,but the average grower just wants to make a grow light and can't afford the commercial stuff or have to spend $ on junk.
I think you can easily make a grow light with LEDs a lot better and cheaper than commercial ones if you want to.
You don't need to be an electrical engineer to do it.

People just need to know the basics for the most part.

It always seems to me that most engineers or wanta be experts keep muddling up LED info as bad as the vendors do.

People who post questions on pepper sites aren't usually engineers.
They want to simply know what to get/buy to make a panel and grow their plants.

Telling them the data on data sheets is what they should run their LEDs at is making sure they will not get satisfactory life spans from their LEDs in general.
LEDs are expensive and if they only last 2-3 seasons they aren't worth buying cost wise.
Other light sources work that are cheaper.

Since most people don't have access to all the do dads and gizmos you might have access to ,a lot of the watts per whatever stuff is pretty much useless and just confuses the average do it yourselfer.

A lot of times if they are using a driver if they only added an extra LED or 2 to a string they would be OK.
But since they overdrove their LEDS they shortened their LEDs life or just plain cooked them.

I think that Chinese LEDs have a bad name because people ran their LEDs at the data sheet values and cook the LEDS.
Also the cheep grow lights are purposely run hot so they can sell them using those lumens or whatever to sell them.

I've used both U.S. and (mostly) chinese LEDS and really don't see any difference in them so far except in price.
A few have bad soldering pads or minor flaws that I can deal with easily.
I don't get many that are screwed up as far as I can tell.
But I don't have all the gizmos to check them out with,only my multi meter...

I don't have any special gizmos to check wave length but as I posted several times.

I use as many different lots and manufacterers LEDs as I can get that are rated anywhere between 400nm and 480nm and red from 615nm and 680nm.
If they aren't putting out the rated nm .So what?They are in the ballpark.
I look at it as a positive thing.

As long as they are in the Plant usable wave lengths I'm happy,my plants seem to be happy too.

As I said before I like 1 watt LEDS because I can cover more wave lengths per panel and can put more per panel to get equal or more lumens per panel for usually cheaper than higher watt leds.

I like the overlap of LEDs beams on my panels,the closer they are, the more overlap I get and the more wave lengths I get evenly distributed on my plants the better.

1 Watt LEDS don't need as much cooling/spacing on my panel.
I can get 2X the stars on my panel in more different nm than with 5 watt LEDS.Close to the same Lux.

My goal is to get my panels to have as many of the plant usable wave lengths as I can per panel.
I don't care that I put 50 , $1.00 leds or 15 ,$3.00 LEDs per panel.
I want to get about the same Lumens and more wave lengths on each panel.

If all my 1 watt LEDS run on the same MA I can put one from each batch of different wave lengths in a string and probably cover more wave lengths than if I used several different LEDS from the same lots in higher watt LEDs.


Since it's somewhat unclear ,in general,that different plants might use some wave lengths in small amounts but don't have to have it all the time,or have to have another at different times,I go for trying to get as many wave lengths of anything red or blue that the sun puts out.

I want my panels to grow anything I want them to grow,no matter what they need or like in general.

In general,all the average grower needs to know to put together a panel is Volts X Amps = Watts.

If your LED is 1 watt you shouldn't run it at more than about 85% +/- of its watt rating.
If the ma is 300 then you shouldn't use more than about 3volts per LED.(3X.3=.9watts)

Using less amps/mA you decrease Lumens and to a certain degree you change wave lengths by reducing the volts.

If your power supply/driver is 12 volts X 20 watts you can't draw more than 20 watts off of it total.

If you are running a 2 volt,300ma LED off of that driver.

You can have your strings of LEDs at 6 LEDS per string(6X2=12).

That string is 12 volts and draws 300ma.

You can run 66 strings (396 LEDs)off that driver.(66X.3= 19.8 watts).

You can run as many 12 volt strings as your Watts will support.
I don't know why or need to know why volts are never ending and watts aren't.It's just the way it is...

If you aren't using a driver then you have to mess with other ways to regulate your Current.

If you are using a regulated power supply like a computer power supply you can probably get away with using just resistors.

Plug in the wall power supplies vary too much for anything other than 5mm or smaller LED use.
They still fry LEDS and usually overdrive them so stay away from them in general.

Go to a calculator like this

http://led.linear1.org/led.wiz

to see what resistor you need to use etc.

I use LM317's and resistors with computer power supplies because I can run more LEDS on them than most of the drivers will run.
My panels usually have too many LEDs for 1 driver to run.
I like 1 power cord for each panel...

With the LM317's I have to subtract 1.25-1,5 volts from the power supply or add that voltage to the total volts needed per string because that's what they use,however you want to look at it.
But making your own driver should be a new post...

One more point,I don't see that LEDs have a penetration problem.
I see my plants growing SUPER thick canopies because the light they get makes them so happy and NO light source is going to penetrate that thick of a canopy.
Once a plant gets big enough,past 12in -18in tall with hand sized leaves I put vertical panels to light up my branches.

I read that vertical lighting isn't that uncommon for other light sources too.

It's probably why so many people like grow boxes...

Since LEDs are spot lights,they don't have as much light to reflect that isn't already on the plant-depends on the angle of the LED and how far it is away from your plant.

Another thing I see ,is a lot of is people using LEDs too close to the plants.
Some panels are good at 18in away.others have a sweet spot that can be closer or farther apart.
Depends on what LEDs you used in your panel...

Yet another post subject.

More Lumens put out by a wide angled LED isn't better all the time than a tighter angled,less lumen LEDS.

I was lucky,a guy on another garden site has helped me any time I've needed help.
It's been going on a couple years now.
He taught and is still teaching me a lot of stuff about LEDs in a way that I can easily understand.

I have to admit.at first he must have been laughing his butt off at my questions but he patiently put up with my sometimes stupid questions and mistakes so I can make my panels for what my end goal will be eventually.
I in turn turned him into a chili head with seeds he was interested in growing and a care package here and there.