[m_t_r_86]

So, a little bit ago, I decided to build my own LED pendent lights. So I found some 50 watt LED chips, bought some electric water tight boxes (roughly 8"x3.5"x2"), bought 2 100watt heatsinks, some 22awg 2 strand wire with connectors. 2 9volt dc walwart (9.3volts at 500mA) and had 2 meanwells eln-60-48d already at the house.

If anyone needs a write up on how to do it, pm me. It's self explanatory.

It was so bright that I only ended up using one.

Here are pictures

Please excuse the cloudiness of the tank. I just transfered everything from my 20long over.

 

[Jamo33]

Looks the goods! What was the total cost of this build if you don't mind me asking?

 

[m_t_r_86]

Minus the meanwells, it came out to about 35 per light. I found the meanwells a yr back for 25 a piece.

 

[Jamo33]

Minus the meanwells, it came out to about 35 per light. I found the meanwells a yr back for 25 a piece.

That's awesome! Thanks mate, still tossing up my lighting ideas for the future and I'm always looking for others ways, LED being one of them obviously

 

[jeffkrol]

Think you could run the 2 chips in parallel off one driver..That would allow 1)under driving 1.3A/2) and saving the chip and 2) better coverage..

running 2 chips at a lower wattage usually produces more light per watt..so you would actually have a slight increase in output..

 

[m_t_r_86]

Each chip needs already 1 amp and if you run in parallel you'll cut the 1.3 amp in half. I tried and it wasn't bright enough.

Bump: If you run in parallel, the voltage doesn't change but the amps will. So each chip needs 36 volts at 1 amp. In parallel, the meanwells will put out 36 volts and 650 mA. If you run in series the voltage changes and the amperage stays the same. So 2 chips in series would require 72 volts at 1 amp.

 

[jeffkrol]

Each chip needs already 1 amp and if you run in parallel you'll cut the 1.3 amp in half. I tried and it wasn't bright enough.

Bump: If you run in parallel, the voltage doesn't change but the amps will. So each chip needs 36 volts at 1 amp..

chips don't "need" 1A.. They can run at almost any amperage as long as the V(f) is met..more or less..

surprised that you found 2 in parallel to be not as bright as one alone.. somewhat counter intuitive to LED..

Lumen efficiency increases as drive current decreases but of course gross output increases/ increasing current..
normally running 2 chips at 1/2 current will, overall, produce more lumens than 1 chip at 2x of the current..
The gain may be small but you are also not shortening the chips lifespan.

 

[m_t_r_86]

When you decrease amperage you increase led life. Agreed. When you decrease amperage you decrease lumens. So tell me how, when wired in parallel, two chips on one driver at half amperage provides more light than one chip at recommended amperage. I'm sorry but it doesn't work that way. Efficiency may be better. But that doesn't mean brighter per say. For example, at 500 mA you may get the total 100% lumens you were looking for. And at 1000mA you may get 75%. But no matter how you look at it, 1000mA current will be brighter than your 500mA current. I'm an electrician and this is what I do all day everyday.

 

[jeffkrol]

When you decrease amperage you increase led life. Agreed. When you decrease amperage you decrease lumens. So tell me how, when wired in parallel, two chips on one driver at half amperage provides more light than one chip at recommended amperage. I'm sorry but it doesn't work that way.

Actually w/ LED's it does..

total wattage at 2 at 1/2A is the same as 1 at 1X amps (roughly, the V(f) is only moderately different)..Higher lumen efficiency means you will sum more lumens per watt..

But no matter how you look at it, 1000mA current will be brighter than your 500mA current

1 @ 1000mA is the same total wattage as 2 at 500mA..

At 500mA you get 152 lumens/watt at 1000mA you get 132 lumens per watt..
IF you have 2 at 500mA vs 1 at 1000mA your gross lumen output increases

1 1000mA chip will be brighter than 1 500mA chip but duller than 2 500mA chips..

 

[m_t_r_86]

I'm sorry but your wrong. I've done tests with leds. And I can't take pictures showing you a multimeter hooked up in line showing you it is dimmer at lower amperage and brighter at higher.

 

[jeffkrol]

I'm sorry but your wrong. I've done tests with leds. And I can't take pictures showing you a multimeter hooked up in line showing you it is dimmer at lower amperage and brighter at higher.

as the die heats up it also loses efficiency..

Driving LED lamps ? some simple design guidelines - LEDs


The hotter they run the faster the output decreases

 

[m_t_r_86]

I do have to apologize. I miss understood what you were comparing. I thought you were comparing only one led @500mA to one @ 1000mA. I misread your post. You are correct in saying that one LED at 100% power compared to two LEDS at 50% power of equal wattage would be dimmer than the two running at half power. I really should thoroughly read posts before jumping off the deep end. Again, sorry. Also, on a side note, most electronics operate under the same principle.

Bump:

as the die heats up it also loses efficiency..

Driving LED lamps ? some simple design guidelines - LEDs

This chart is a general target chart. Without you knowing What LEDs I'm using, it is irrelevant. My LEDs will maintain 100% lumen efficiency based off of temperature alone until they pass 130 degrees F.

 

[jeffkrol]

I do have to apologize. I miss understood what you were comparing. I thought you were comparing only one led @500mA to one @ 1000mA. I misread your post. You are correct in saying that one LED at 100% power compared to two LEDS at 50% power of equal wattage would be dimmer than the two running at half power. I really should thoroughly read posts before jumping off the deep end. Again, sorry. Also, on a side note, most electronics operate under the same principle.

Bump:

This chart is a general target chart. Without you knowing What LEDs I'm using, it is irrelevant. My LEDs will maintain 100% lumen efficiency based off of temperature alone until they pass 130 degrees F.

Hey no problem. I suspected that you were not quite responding to what I was trying to say..


On a side note..hitting 130F (54C) at the junction is usually not that difficult.. heatsink be darned, w such high density chips..

I still stand that 2 in parallel have more upside than downside than running one at high current.. but it is a choice..

 

[m_t_r_86]

Hey no problem. I suspected that you were not quite responding to what I was trying to say..


On a side note..hitting 130F (54C) at the junction is usually not that difficult.. heatsink be darned, w such high density chips..

I still stand that 2 in parallel have more upside than downside than running one at high current.. but it is a choice..

I didn't want to run them both on the same driver for several reasons. In case one LED chip went to hell, I didn't want a surge to the other, even though I'm sure it would handle it. I also wanted to individually control brightness per side of tank. Left side is a more high light required side and right not so much. But in the end, the one alone was plenty bright for the whole tank meeting the high light required. If I ran two, I probably could end up running both on one driver with a fuse and maybe a resistor or two inside a junction box.

 

[PlantedRich]

I like the way you think! There are just a lot of things that work too well to go by the book of designers and engineers! If it works, KISS it?
I keep African cichlids and plants and all the books tell me you can't! Apparently my fish don't read the books either?