What's new

Building a Home Made LED

hempfield

Organic LED Grower
Veteran
He mentioned the XM-L models, I just underline that they will not be driven at their full potential, just that.
 

xmobotx

ecks moe baw teeks
ICMag Donor
Veteran
i have wondered whether under-driving might be a good solution for the real small micros like PC and such

it seems like running a lower rated diode might be a better solution

i would definitely try to move towards vertical/intra-canopy in that small of a space at least have some corner mounted directed angular into the canopy along w/ top lights

something i feel might be a nice fixture would be something you could mount at a slight angle on a wall these could be placed 8-10" down from the top then 8-10" down from that and so on as one might feel the need
 

DrFever

Active member
Veteran
Rrog, something to keep in mind with LEDs is that most of what you have read is bullshit. They are potentially able to be more efficient than HIDs, but the claims of 2x - 4x are wildly overstated. I think that most of us have found that you can pretty equivalent results at somewhere around 70% (+/-) of the HID wattage.

The other thing that sounds like it might be entering the equation right now is the idea that LEDs don't make any heat. They do. They are roughly similar to HIDs in that there isn't that much difference in the relative efficiencies of either light source - they both range from 20-30% efficient at converting electricity to light, and the rest of the power gets converted to heat. Where LEDs do have the advantage on heat is: a) the 30% savings in the overall power means that much less being converted to heat, and b) they don't create infra-red heat like HID lamps do. HID lamps "cast" IR and heat up the surrounding objects - this isn't a factor with LEDs. The downside is that it is sometimes more difficult to do the heat extraction on LEDs because they don't lend themselves well to devices like vented hoods.

Starting your own thread would probably be a good idea - I think that you are about to get caught up in the rip-tide of LED technology! It would also be good to document this for others.
I think the main problem with LED is there overall coverage which sucks period when you think as a comercial or under ground big yield grow op you really don't care about efficiency nor heat you worry about cops and pulling huge yields
i personally can't see LED jumping or taking HId's place in the comercial end for atleast 20 more years hahahaha
 

WeedIsGod

Member
xmobotx: I'm building three panels almost exactly as you describe, but not for a micro grow. I'm hoping to build a 100W spotlight for over the canopy, but I've already built three panels of 36 under driven 1W LED's as side lighting (~30W per panel, everything on a 330mA driver). They are very thin, completely passive (cooling isn't a huge deal at this wattage) with a Pink/White/Red sorta tinge to them. Everything in a 2'x2' space. You might see something in the near future from me... maybe not. :/

DrFever: But where are the forward thinkers?

They make 100W, 200W, 300W, 500W, etc. multichip LED's. CPU coolers can be made to efficiently cool these for relatively cheaply. I think it's only a matter of time before we start seeing stuff like this on the small scale DIY front. I've already seen commercially built panels utilizing 1, 4, and 6 x 100W chips with custom spectral distributions (Red+Blue, etc), and that same manufacturer specifically asked me if I knew a green house outlet in order to get their product line out there (just a foot in the door). Philips is also gaining ground in commercial green houses, but this is most likely to grow food, flowers, etc. I think there's interest on both the manufacturing and consumer sides of this, it's just a matter of how much, I guess.

20 years? I think it depends. I don't see it taking that long, but of course that's only if another technology doesn't eclipse LED within that time.
 

tenthirty

Member
I don't see one cash cropper here......at least on any kind of scale.
As far as I'm concerned, it's about growing a reasonable quantity of excellent quality on the least resources possible. Thus the HML.

This is a hobby for most of us, ether through desire or necessity.
So you can assume that we at least like what we are doing.
At least for me, it's all about the journey and learning.

I don't think that you can say the same thing for a lot of cash croppers.
 

rives

Inveterate Tinkerer
Mentor
ICMag Donor
Veteran
i personally can't see LED jumping or taking HId's place in the comercial end for atleast 20 more years hahahaha

I think that your crystal ball is myopic. If you consider that 20 years ago, LEDs were largely relegated to pilot light usage on electronic equipment, they have made remarkable progress. Cree just developed an LED that puts out nearly 1-1/2 times the lumens/watt of an HPS. Electronic product development is non-linear - at the current rate of development, you will have as hard a time buying a 1000w HID in 20 years as you currently do trying to find a 100w incandescent lamp.
 

xmobotx

ecks moe baw teeks
ICMag Donor
Veteran
xmobotx: I'm building three panels almost exactly as you describe, but not for a micro grow. I'm hoping to build a 100W spotlight for over the canopy, but I've already built three panels of 36 under driven 1W LED's as side lighting (~30W per panel, everything on a 330mA driver). They are very thin, completely passive (cooling isn't a huge deal at this wattage) with a Pink/White/Red sorta tinge to them. Everything in a 2'x2' space.

yeah i had kinda 'gone on' in my thought process and i did mean for larger grows ~lol; in a micro grow you could put the leds in the floor & be intra-canopy
 

Kukok

New member
At just a fifth of their maximum output power. Sound like a waste to me. With proper cooling he could easily achieve more PAR lumens per square foot if driven at 3/5 or even 4/5 of the max power.

I dont get that. How can I drive them at 3/5 or 4/5 of max power? Use mor LEDs in the line and get the voltage limited by the driver?

Maybe more XP-Gs (total of around 30W) are better because of the lower lumens.

@xmobotx:
Vertical was the way I wanted to go with my cfls. However I could imagine a smal panel with 4 XP-Gs on the left side wall (see pic in original post).
 

xmobotx

ecks moe baw teeks
ICMag Donor
Veteran
yes, the LEDs will operate on the power they get within their parameters ~note they are typically rated w/ a range and will operate throughout that range

a less intense lamp may well be better suited to the tiny micro grows ~idk but makes sense

LEDs lend themselves quite handily to vertical because their light is directional and because of their low profile besides leaves and such can get quite close to them w/o harm
 

jcmjrt

Member
I dont get that. How can I drive them at 3/5 or 4/5 of max power? Use mor LEDs in the line and get the voltage limited by the driver?

You drive them at a higher current than 700mA. I just finished a build last night on a fixture with Cree XMLs and I'm driving them with an Inventronics at 2100mA. That's still coasting pretty well and they could be driven harder. The temp right at the LED on the heatsink didn't pass 100 F so I'm pretty happy with that.

I'm not as height limited on my canopy as you are. You might want to consider some lower wattage LEDs and drive them at 700 to a 1000mA so that you can get the plants very close to your fixture. Take a look at Steves LEDs; he sells luxeon chips which are quite excellent at a reasonable cost and his drivers are inexpensive too (separate power supply required). I'm going to be making a couple of fixtures later with his LEDs and drivers so I'll be able to report back in a couple of weeks. With the inventronics (power supply included in unit) and Steve's drivers too, it's very easy and cheap to add dimming....just a small pot. Now the pot does not dim to off just low light. I'm figuring it allows me some leeway for using the fixture with seedlings and young plants...and was just too cheap and easy not to do.
 

hempfield

Organic LED Grower
Veteran
I dont get that. How can I drive them at 3/5 or 4/5 of max power? Use mor LEDs in the line and get the voltage limited by the driver?

Maybe more XP-Gs (total of around 30W) are better because of the lower lumens.

@xmobotx:
Vertical was the way I wanted to go with my cfls. However I could imagine a smal panel with 4 XP-Gs on the left side wall (see pic in original post).

The current is the factor which decide the light intensity, not the voltage. LEDs maintain a relative constant voltage (forward voltage) but they tend to over burn if the current is not limited by the driver.

The maxim current for a XM-L is 3000mA, but with active cooling can be raised a little bit (decreasing the light efficiency and the life of the LED itself). As far as I know, the 700mA drivers actually limit the current to a lower value ~650mA, which mean a little more than a fifth of the maximum current.

Must keep in mind that the forward voltage also vary with the current, but not to much (from 2.8 at 700mA to 3.3 at 2800mA).

So you don't need aditional LEDs, but a more powerful driver, preferable dimmable, able to deliver a maximum of 2800mA.

I am waiting some XM-L U3 driver to arrive in the mail, together with this constant current step-up boost driver http://www.ebay.com/itm/181031410566 and this gadget http://www.ebay.com/itm/290700266076 to display the value of the current. In addition I will also have this monitoring the output voltage of the driver with this digital mini voltmeter :http://www.ebay.com/itm/170846281102
 

rrog

Active member
Veteran
I'm planning to run the XM-Ls at 2000mA. I'm just not sure about the U3, etc nomenclature (yet)
 

jcmjrt

Member
I'm just not sure about the U3, etc nomenclature (yet)

There is a great deal of variation in quality of product within the same batch of LEDs. Of course, the variation from batch to batch is even larger. The manufacturer's test the leds for performance and then bin them by level of performance - one of those factors is how close is the frequency of the light produced to the desired frequency.

The U3, T5 thing is just the way they label their bins.
 

rrog

Active member
Veteran
Oh, really? Thanks for that! Had no idea. So does a U3 perform better than a T5 or something like that?
 

hempfield

Organic LED Grower
Veteran
The T6 series (cold white) have the maximum light output. The XM-L T6 U3 is the latest generation of the XM-L, with a slight improvement of the light output and some other minor improvements.
 

rrog

Active member
Veteran
Would it be wise for me to mix bins? In other words, all my WW are one bin, the CW are T6 U3, and the NW are yet another bin? I gather from your post that NW and WW are not available in T6 U3 like the CW are.
 

Kukok

New member
Thank you all for the very patient mood as most time on icmag.
The current is the factor which decide the light intensity, not the voltage. LEDs maintain a relative constant voltage (forward voltage) but they tend to over burn if the current is not limited by the driver.

The maxim current for a XM-L is 3000mA, but with active cooling can be raised a little bit (decreasing the light efficiency and the life of the LED itself). As far as I know, the 700mA drivers actually limit the current to a lower value ~650mA, which mean a little more than a fifth of the maximum current.

Must keep in mind that the forward voltage also vary with the current, but not to much (from 2.8 at 700mA to 3.3 at 2800mA).

So you don't need aditional LEDs, but a more powerful driver, preferable dimmable, able to deliver a maximum of 2800mA.

I am waiting some XM-L U3 driver to arrive in the mail, together with this constant current step-up boost driver http://www.ebay.com/itm/181031410566 and this gadget http://www.ebay.com/itm/290700266076 to display the value of the current. In addition I will also have this monitoring the output voltage of the driver with this digital mini voltmeter :http://www.ebay.com/itm/170846281102

I think,...I slowly get it.... thank you all for the clear answers. May I sumarize and see if I understand:

1) The max. light output of LEDs is at their maximum current (normally at 3000mA)...got it..
2) underdriving LEDs (eg. 30V driver with 31V LEDs) is better then to have unused power (eg. 30V driver with 20V LEDs) which results in additional heat(for dummies)

So I think I go this route:
11x 2W XM-Ls with dimmable driver = pot+driver 2800mA+power supply(could you suggest one from this site? because I am having problems with the dependencies between electrical power/current still, and was not able to find a 30V power supply)

xmobotx: Due to my cfls are used vertical I was thinking on a small Al-profile with dimable 4 WW XM-Ls on a seperate driver for flower, too.
My decision to use XM- over XP-LEDs is that they bring double power (W/lm wise, no need for a big array in my situation) for the same power consumtion (do I see things right?) and with the pot I can dim the driver to max (or lower) light out-put of the XPs anyway. And that at same costs.

jcmjrt: I am sitting in Europe. US shop means shipping fees explode, taxes and bureaucracy. EU ftw. in my case

My last OT-question: Has anyone tried intra-canopy or bottom lighting? I always thought that our plants only can work with the light which hits the top-side of the leaves. Anyone has some information or link for learning?

tenthirty and all: sorry for hijacking, will open a seperate thread if requested. Maybe I will start a report on my HML when I start but not sure I am a lazy stoner ;)
 

xmobotx

ecks moe baw teeks
ICMag Donor
Veteran
the bottom lighting thing was *kinda* a joke ~but i would use them so the beam was cast along the side of the grow if i applied it; still speculative for the reason you mention & probably more to just bring ambient levels up

however side-lighting {vertical} is a tested tried & true tech; combining vert side lighting w/ top-lighting doesnt have as exaggerated effect as strict side lighting
 

Kukok

New member
hehe, ah ok...no english native here, so didn't get the joke.

One thing I still wonder is which/how many power supplies(ps) are needed. Most of them are 12 V. Does that mean that I need one(one driver) per 3 LEDs/4-5V usage for full power. And is this the maximum quantity of such LEDs used on a 12V ps.

Could anyone post a link or some details of the power supply side of HMLs please?
 

hempfield

Organic LED Grower
Veteran
Thank you all for the very patient mood as most time on icmag.


I think,...I slowly get it.... thank you all for the clear answers. May I sumarize and see if I understand:

1) The max. light output of LEDs is at their maximum current (normally at 3000mA)...got it..
2) underdriving LEDs (eg. 30V driver with 31V LEDs) is better then to have unused power (eg. 30V driver with 20V LEDs) which results in additional heat(for dummies)

So I think I go this route:
11x 2W XM-Ls with dimmable driver = pot+driver 2800mA+power supply(could you suggest one from this site? because I am having problems with the dependencies between electrical power/current still, and was not able to find a 30V power supply)

xmobotx: Due to my cfls are used vertical I was thinking on a small Al-profile with dimable 4 WW XM-Ls on a seperate driver for flower, too.
My decision to use XM- over XP-LEDs is that they bring double power (W/lm wise, no need for a big array in my situation) for the same power consumtion (do I see things right?) and with the pot I can dim the driver to max (or lower) light out-put of the XPs anyway. And that at same costs.

jcmjrt: I am sitting in Europe. US shop means shipping fees explode, taxes and bureaucracy. EU ftw. in my case

My last OT-question: Has anyone tried intra-canopy or bottom lighting? I always thought that our plants only can work with the light which hits the top-side of the leaves. Anyone has some information or link for learning?

tenthirty and all: sorry for hijacking, will open a seperate thread if requested. Maybe I will start a report on my HML when I start but not sure I am a lazy stoner ;)


I think you forgot something or the things are not so clear, but let us try to put them in order:

To light a LED you need a power supply that give and output voltage just a little bit higher than the forward voltage of that LED (of chain of LEDs if you want to connect several in series - which is recommended for simplicity of the circuit).

A power supply, especially a switching power supply, will have a constant voltage on it's output but if you connect the LED(s) directly on the power supply , any difference between the output voltage of the PSU and forward voltage of the LED(s) will be converted on extra current flowing through the LED(s). The higher the difference, the higher the current, and this means LED(s) failure.

To avoid this situation, you need to add a intermediate device which does not allow the current to exceed a specified value, wherever you connect just one LED or two, or three or N LEDs (with the condition that the sum of the forwarding voltage does not exceed the maximum output voltage of the driver). This is the constant current driver that you mentioned earlier (http://www.led-tech.de/en/LED-Contr...wer-Supply--2800mA,-30V--LT-1290_118_119.html). But keep in mind that the input voltage of this driver is 30V, so you will need an additional PSU that give 30V on output.

To simplify even more :

220V AC --> PSU --> 30V --> Constant Current Driver --> LEDs

So to keep your project as simple as possible, try to choose a constant current driver that take 220V AC on input and give you 2800mA on the output , at a output voltage equal to the forward voltage of an XM-L , multiplied by the number of LEDs.

I'm talking nonsense ? :biggrin:
 
Top