The Truth About LED Lights

[Hosebomber]

I know this is going to come across as conceded, but I would but any of my test lights against a 3 or 4 band red and blue only panel. Shit I would even give those red and blue only panels a 10-20% power advantage and still feel 100% confident that using a wider spectra will crush the results of a red and blue only panel. The truth is that Chlorophyll A-F totaled only register around 40% of the actual growth hormone light receptors in most plants. Cannibalism is one of the more complex plants known and very little research is done on it for legal reasons.

Do not under estimate accessory pigments. I will help your grow and final product.

 

[WillyP]

Hosebomber what wavelengths (bands) are you using in your test lights that makes them so good? Iwas going to do a comparison of my lights against my 180W Hyroponics Hut light, but I decided not to waste a good plant. As for the accessory pigments, the contribution to the total sbsorption cueve is quite small and In my design, a warm white led provides enough energy cover the spectrum of the accessory pigments.
What would be nice is if someone with an HPLC could do an analysis of leaves to determine their pigment content. I would have done this already if I had access to the instrumentation

 

[s1ingblade]

That's very informative thank you very much. I know with this site and it's great members I will figure out those LED's

 

[Buckshot]

Cannibalism is one of the more complex plants known and very little research is done on it for legal reasons.

cannibalism ...

 

[Hosebomber]

Hosebomber what wavelengths (bands) are you using in your test lights that makes them so good? Iwas going to do a comparison of my lights against my 180W Hyroponics Hut light, but I decided not to waste a good plant. As for the accessory pigments, the contribution to the total sbsorption cueve is quite small and In my design, a warm white led provides enough energy cover the spectrum of the accessory pigments.
What would be nice is if someone with an HPLC could do an analysis of leaves to determine their pigment content. I would have done this already if I had access to the instrumentation

At any given time I am running 3-4 test with different lights. I'm not welling to give out all of the information on my systems just in case I ever do decide to start producing them for profit (or simply to repay the tens of thousands I've burnt up in time and money). Most commercial products are very close and the differences have gotten much more narrow over the past few years. The hardest part of all the testing is that each plant species and strain tends to vary on what works best. So after reading nearly every peer reviewed article on light effects, it still doesn't give you a 100% clear picture on what each plant needs (because nearly all studies use a different species or plant type all together).

Testing your light against a mass produced model should be on the top of your list of things to do... you may be way off on your thoughts and they might outperform you. Then you have a baseline of where to move forward from. As for doing a HPLC test, contact a local university, they may have the resources to do the test for you cheaply or free even. University of Kentucky did quite a few test for me when I was living there a bit ago. Unfortunately, now I live over 100 miles for anything even remotely close to a agricultural lab.

Buck, sorry about that auto correct, been at the hospital with my wife for a week and my phone likes to type on it's own sometimes.

 

[Buckshot]

Buck, sorry about that auto correct, been at the hospital with my wife for a week and my phone likes to type on it's own sometimes.

I am sorry to hear your wife is in the hospital. There seems to be a lot of that going on at the moment ... she just made my list. I thought the mistake was a riot... worth a smile in tough times.

Buck

 

[WillyP]

I don't think comparing grams/watt (weight/power) is a good measure of performance. A better comparison would be weight per energy consumption or grams/watt-hr. For instance a 100 watt light run for 2000 hrs produces 300 grams whereas a 200 watt light run for a total of 1000 hours also produces 300 grams. The first light has a higher g/watt, but both plants had the same grams per watt-hr.

 

[savolainen]

Again I disagree. As we spoke about before, the naming of LED 1 3 5 etc watt is based off the drive current used (350 750 1000mA etc.). This is done strickly for simplicity. The voltage across each diode should not change more than 5% unless the system is very poorly designed. Even a 5% voltage change over 2 or 3 volts will ot make much of a change in the diode output. Radiant power is the only way to corrctly discribe >440 and <630 nm leds.

You are incorrect and stubborn. You should go to check and read those principles of constant current drivers little by little...

So that those drivers can provide constant current, they should alter something, and that something IS voltage.
They start from the low part of the given range, like if the driver output is for example 30-48V and current 1000mA, it increases voltage from 30V until current is what is needed to LED circuit to work.

So one who is going to build LED light should just know how to pick right LEDs for the right driver.
It might need little calculating.

But most obvious circuit is easy: if LED's Vf value is given 10-12V those should be put in series, 3pcs. of them... in case that forward current is same 1000mA.

If driver gives 2000mA and 30-48V, those same LED's can be used if connected three in series and two in parallel.

Parallel connections are not recommended unless series that are connected in parallel, have exactly same combined fV value, otherwise, if they differ, other strain draws more current than the other -> that causes the strain to get more warm than the other which causes that it takes even more current, until one of the LEDs gets broken.

Q. So how could that strain which gets more hot, have more current if they use constant current driver?
A. Another strain takes less current at the same time as that hotter one takes more. If Vf values are identical, current gets equally shared. So plain series connections are recommended.

 

[Hosebomber]

You are incorrect and stubborn. You should go to check and read those principles of constant current drivers little by little...

So that those drivers can provide constant current, they should alter something, and that something IS voltage.
They start from the low part of the given range, like if the driver output is for example 30-48V and current 1000mA, it increases voltage from 30V until current is what is needed to LED circuit to work.

So one who is going to build LED light should just know how to pick right LEDs for the right driver.
It might need little calculating.

Maybe you should go re-read those few post prior to what you quoted. I stated that the voltage should not change more than 5% or it was a poorly designed system. Likewise, I stated the the CURRENT would not change. It's a constant current driver... that's what it does, keeps the current constant. The voltage varies slightly to regulate the system. If you have more than a 5% change in voltage you have chosen the wrong driver and CAN lead to thermal runaway. You do not chose LEDs to fit the driver. You make (or buy) the driver to work within your design parameters. As for running them in parallel or series you are correct. I have yet to see a LED panel wired in a parallel though. It's just bad electrical engineering in ALMOST all cases (there are always exceptions).

Please read and educate yourself before calling people names. I'm all for debate and disagreement, but failure to simply read the context of the conversation and then calling people names is poor etiquette and makes you look dumb.

 

[lenngray]

How 'bout someone explains why, if an HPS wastes 70% of its energy making HEAT, and LEDS, which necessarily generate soooo much less HEAT, and can put all that available energy into making something OTHER than heat (hmmm, what might that be), that with all the genius of targeting all the energy to Photosynthetically Active wavelengths, that these LED guys can't get better than 1/3 better yield. Even ignoring the spectrum magic, just retargeting the HEAT energy should give better than 200% improvement. Is all the rest of the energy being retargeted to FRAUD?

 

[Hosebomber]

The best answer to that Lenn is the fact that everyone in the LED manufacturing community forgets about all of the accessory pigments. That and the fact that the 70% figure is just plan wrong. It may be close to 70% if you include heat and green/yellow wavelengths like some people try to claim. Those other wavelengths are used. To what extent? I'm not 100% sure and neither is anyone else in the world. There simply isn't enough funding for the artificial grow light industry to test and document everything scientifically.

If you go back 5 or 6 years ago when LED grow lights where first coming out, you had 2 band 460 blue and 630 red. then it went to 3 band then 5 then 9 now you have 11 and some claiming as many as 15 bands. We know for a fact that the 9 and 11 band LED panels produce far better than the 2 3 or 5 band LED lights. I'm sure penetration power of higher efficiency LEDs play a part but the addition of wavelengths play a much larger roll.

 

[Buckshot]

How 'bout someone explains why, if an HPS wastes 70% of its energy making HEAT, and LEDS, which necessarily generate soooo much less HEAT, and can put all that available energy into making something OTHER than heat (hmmm, what might that be), that with all the genius of targeting all the energy to Photosynthetically Active wavelengths, that these LED guys can't get better than 1/3 better yield. Even ignoring the spectrum magic, just retargeting the HEAT energy should give better than 200% improvement. Is all the rest of the energy being retargeted to FRAUD?

This is the second run of this strain for me grown under LED only... no fraud that is an 18" ruler hanging there

want a closer look? yeah no fooling Lenn

here are two more going right now ... check my journal for updates

 

[savolainen]

You do not chose LEDs to fit the driver.

Yes i do.

You make (or buy) the driver to work within your design parameters.

My design parameters are:
1. choose what bandwidth i want
2. choose cheap LED's and
3. find driver which gives correct current and voltage from EBAY.

Which means that it's voltage starts near that value which comes from LED's after connecting those into series.
I don't see any problem there if driver lifts it's voltage more than 5% until it reaches LED's combined fV. If it lowers drivers efficiency little bit it doesn't bother me much.

If suitable driver can not be easily found, i change my design parameters so that they will, normally by choosing another LED's + driver combination. LED sellers often have suitable drivers available as well.

I don't care if driver need's to lift it's voltage 1%,4% or 6% as long as it gives correct current (or little less) and LED's work. I guess you were talking about tolerance which is often given +-5%.
This output voltage that constant current drivers can give is different thing.
How else this kind of products could even exist? They must be really REALLY purely designed as voltage changes so much... Mean Well MW 12 48V 350mA 16W AC DC LED Driver APC 16 350 TUV Brand New | eBay
-there is voltage output range from 12-48V. (Tolerance is +-5V and current 350mA).


Of course i choose most suitable parts that there are -though i have used parallel connection in one light. It is 11 x 1W in series and two rows in parallel.

LED's were left overs from a bigger project and as i also had one unused driver i put those pieces together.
BTW. that 680mA driver i bought from EBAY was actually 600mA. And as LEDs had 350mA forward current i put 2 x in parallel with that one. It would otherwise been quite good if that driver would had only 20mA short, but when it came out that it was only 600mA, i was kinda disappointed. LED's could work much brighter than they are now, at least i believe so.
Good thing is that under driving keeps temperature down. All are mounted to big aluminum bar.

So be aware about those 680mA constant current drivers in EBAY.
Light has been buzy about 6 weeks with 16/8 schedule and works well. Both rows are equally bright. All LED's were bought at the same time so it is possible that they have identical fV values.
If this fails i'll buy those kind of drivers next time i have 350mA


Besides ain't there parallel+series connections in those biggest LED components out there? Like 10W, 20W, 50W.

Other LED's i currently have: 20W Blue (one component), 3x10W Royal blue, 10x3W 660nm RED, 4pcs. 7x1W Red+blue LED lamps with E27 base (6xRED,1xBue) and lots of 9W White spotlights.

White LED's are good for patching up missing wavelengths.

At the moment there is too much blue compared to reds but i'm working on it. My "under construction" list have: 60x3W RED (660nm), 77x1W RED (660nm), 1x50W RED (one component), 1x50W white (one component).

So far all lights have passive coolers (big ones) but i guess that those 50W LED components are gonna need fan just because there is pretty much power (and heat production) in small area.

I am not gonna make any journal from my plants. At least anytime soon... there is so much else to do, and i haven't took pictures from the beginning.

 

[s1ingblade]

can't wait to see some pics savolainen, when the time's right of course


That cola is squashing that ruler BuckShot, bout time to get a yard stick

 

[Buckshot]

can't wait to see some pics savolainen, when the time's right of course


That cola is squashing that ruler BuckShot, bout time to get a yard stick

... I actually was looking around for and old fashioned wooden yardstick to dispell some of the bunk about lack of penetration

Thanks AM

 

[hightimes4evr]

Cool! Good to learn more LED tech

I will follow this thread

 

[Hosebomber]

So let me get this straight, you are calling me incorrect and stubborn based off of your rudimentary buy it off Ebay and toss it together model?

First, that MeanWell has a voltage tolerance of +- 5% not 5V. The spec sheet is at the bottom of your link.

Secondly, no the large 10, 20 ,50 watt LEDs are a series of diodes. If any one of the small diodes in the series burns out the whole light stops working. You can see this in the mechanical portion of the datasheet for the given LED.

You may have a very basic understanding of creating a circuit that will drive LEDs but you are on dangerous ground in your thought process. If you continue to make panels with parallel circuits and do not match your drivers efficiently you have a very good chance at creating a fire. Over pulling or under loading those drivers will make them excessively hot and can cause them to explode. When using these off the shelf drivers you do not have any amperage regulation. This can cause an additional fire risk from drawing more power than the wiring or circuitry can handle. This includes the wiring in your walls. Voltage * current = Amperage (watts). I should have clarified the "You do not chose LEDs to fit the driver." comment. When you are "properly" designing an led light panel you design the desired lighting output and ratios,then power the LEDs to the desired voltage and current, then develop the heat management system, and lastly aesthetics.

Please Please Please, if nothing else, before you start making your next panel, place a circuit breaker or fuse (that is no more than 15% of your "designed" load) in the line BEFORE you wire into your drivers and diodes. 12.5A would be the max on a single home outlet (in the US). I suggest retro fitting your older panels as well. There is no need to die or burn your house down to try and prove a point.

 

[lenngray]

Actually, they could not be counting green and yellow of HPS toward that 70% heat "waste", as those represent an actual majority of the "lumens" that make up the 55,000 of my 400 Watt HPS. But, since you bring it up, I'll insist that the actual "not knowing" is in the arena of "where is all that energy going in LEDs if not into heat?!!" Think "conservation of energy", and what is being consciously chosen when such selection is available.

You guys act like you discovered PAR and bring up your absorption peaks in Blue and Red, but I can find photosynthesis graphs across many species of plants that show why the entire spectrum gives so much more than just those peaks, and the papers go back at least as far as 1969. This is part of what's now making me think in terms of fraud.

And, no, mr picture guy, I'm not saying one can't grow plants under LEDs. The fraud is about "3x better". Hosebomber has at least offered a quantification that it's 50% better (400WLED==600WHPS), but it took a while to pull that truth out of him.

 

[savolainen]

So let me get this straight, you are calling me incorrect and stubborn based off of your rudimentary buy it off Ebay and toss it together model?

First, that MeanWell has a voltage tolerance of +- 5% not 5V. The spec sheet is at the bottom of your link.

Now, gimme a break. Of course tolerance is % -if i wrote V it was typo because i was so upset because of your hostile attitude. You seem to be very happy that i made one mistake that you can grab on.

Just before where mistakely wrote V instead of % i also wrote: "I guess you were talking about tolerance which is often given +-5%."

Besides you seem to have habit to speak things that you assume as absolute truths -it normally tells about very big ego... which is easily hurt.

There is said "Figure 1 shows a 10W LED, in this case it already uses LEDs in series/parallel internally.".
Which is exactly what i speculated earlier and you denied here:

Secondly, no the large 10, 20 ,50 watt LEDs are a series of diodes.

Like that "SHARP Mega-Zenigata 50W class LEDs" which has 16 in series and 10 in parallel.

AND you have successfully avoided the very point why i registered to this forum and posted my first (it was meant to be my only post as i have lots of other things to do) post.

You claimed that constant current drivers don't have variable voltage output -which was simply incorrect. Period.

Thanks for wasting my time and pulling me to your level, meaning lack of good and constructive spirit.

Pls. do not answer anything. I don't have any reason to follow this conversation anymore. This does not educate me in anyway and i can ruin my days better ways.

I'm out of here.

 

[Hosebomber]

You claimed that constant current drivers don't have variable voltage output -which was simply incorrect. Period.

You have no real knowledge or experience in making LED panels and you attach a lie in my name to back up your "point".

Changing the voltage will not change the current. It will change the overall wattage or amperage draw, but not the current.

The voltage across each diode should not change more than 5% unless the system is very poorly designed. Even a 5% voltage change over 2 or 3 volts will ot make much of a change in the diode output.

Those are the exact quotes from my post. Nowhere have I ever made that statement. You still have not learned how to properly read in context. I didn't avoid your point, I ignored it because it was already answered and invalid.

As I stated, it is generally a bad idea to run LEDs in a parallel circuit in MOST CASES, there are always exceptions and the overall design goal is the requirement. It is not a great idea to do it in grow lights. In those high wattage/voltage leds that do use internal series parallel wiring they have built in fault detection to burn out the whole of the LED when there is a failure. Thus why I stated look at the datasheets for them.

lenngray, if you use the search function you will see that me and a few other LED guys on the forums had the 2/3rds discussion about 8 months ago. I'll state again, I'm not tied to any LED company and have not let anyone other than family test any of my lights other than myself. I have yet to get a "mass production" type panel to grow and equal harvest at 50% HPS power and that has been my goal for a few years. If I ever do hit that mark I may have my lights produced but until then I simply do my test and help share knowledge and links to peer reviewed articles (not random web pages made by sound people or advertisements of some led light company). I have made it a point to call out nearly every company about claims of USA only LEDs, power of a 1000 watt HPS, using brand name LEDS from companies that do not make those wavelengths, and all of the other ludicrous claims that have been made.