Diy led
- Thread starter flworkd
- Start date
brassmonkyballs
BMB
Yes its fine to under drive your LEDs. I didn't make my point which was that a person is using a particular driver is taking it at face value what the performance would be. In this case the specs of the FEDY is 220 lm at 700ma and so that's the expectation of output. When the driver is under driving them, not by design but by mis-labeling a driver, then the output is not as advertized, even if its acceptable.
Hope that clears up what I tried to convey. But yes, it is in fact ok to under drive them, so long as you know what you're driving them at. But in my opinion its silly to drive them less than their rating if they're designed to handle it.
Hope that clears up what I tried to convey. But yes, it is in fact ok to under drive them, so long as you know what you're driving them at. But in my opinion its silly to drive them less than their rating if they're designed to handle it.
I agree with you both BMB on the fact that you should get out of a product what it is advertized to do. As with any product you buy you check out the specs. I also agree josephv that under driving will most likely give us more life of the led's.I have finished wiring my fixture and man it is bright no doubt, but
I would like to know the par readings. And if they are going to work on my tank with these drivers.:spin1:
I would like to know the par readings. And if they are going to work on my tank with these drivers.:spin1:
brassmonkyballs
BMB
I'm not sure what 'near' means in your statement. I agree for most people this is good advice.
In this case though the rating is not the limit. Its the max continuous not point of failure. They are designed to be run at this current level. Take a Cree XPG.... it outputs 139lm @350ma yet at 1300ma it's output is exactly 300% of that...why run it far less than what it's rated for? In fact Cree changed the max to 1500ma on the XPG after further testing but most use the Meanwell driver which puts out 1300ma so thats where they are running them. The earlier XREs have been running at their rating of 700ma for a long time now in various fixtures.
I fully intend to run the FEDYs at the 700ma they are rated for as well....and likely set up a test strip and push them to 1000ma for test purposes. if the guys running these things at 700ma on C-channel are getting no heat, then they could be pushed more with a heatsink. It all comes down to junction temperature when determining max forward current.
Again, to be safe and extend life, under driving them is fine...but If I'm using Crees I'm going to minimize my cost for coverage, while driving the hell of them for output.
schudini
Active member
Yes, I was also interested in torture testing them.
Do you have an infrared pyrometer or thermocouple to measure junction temperature as a function of current?
brassmonkyballs
BMB
I do have a Raytek handheld. SHould be at least give me an idea of whats happening.
Well, I just finished up wiring 80 LEDs 10 in series 8 in parallel to the 200w FEDY driver, and after about 10min of running it, the aluminum is uncomfortably hot to handle. The LEDs are mounted ~1.5 inch on center diagonal side by side on two pieces of 72" 2X1" channel aluminum. I tested it in the open not enclosed in the hood. It doesn't seem like anyone else had heat issuing running similar configurations in the open air, but I guess I could put fans on them. Am I overdriving the LEDs? A mis-solder touching the aluminum would just cause a short, not a heat build up, right? I don't have a way to test them right now. If they are being overdriven, could I add a resistor to each series to compensate?
bigt0706
New member
from all the testing being done so far u should be underdriving them ,when u tested them were they laying on a table or floor ? if the back of the chanle is close to anything restricting air around them they will get hot
Agreed, I really thought I'd be under-driving them, but I was thinking if the drivers weren't as advertised, they could be outputting more than less too.
They were on top of my open top hood...top of channel in the open air, LEDs pointing down, 8" of air to top of water and only to 1" pieces of wood in contact with the bottom.
brassmonkyballs
BMB
How about taking some measurements so we can figure out what's going on? 1st, I would never use the channel every is using but hey, its being reported as working. 2nd I think you're using smaller channel and if so that's not good. 3rd, I agree with you that if they are running under then they have the potential to be over driving just as well. Therefore, you need to take some measurements to determine exactly what's going on.....and on each string.
You're right, I didn't realize until I went back and looked, everyone's been running 3x2" channel. The good news is, they're not overdriven. Honestly, I didn't test every series, but as everyone else indicated they are under-driven, ~678ma, so not by too much. I've been running them for the last couple of hours with a small external fan, and the heat is no problem now.
I found a bunch a 2" server fans headed to the recycler today that run at 650ma, so I think I'll run one in each series (or maybe every other series) for cooling and to under-drive the LEDs a little more...
I'm getting complaints that there's too much light coming from the tank, and right now I only have 80 running...another 80 are waiting to be soldered.
Two questions...
1. I think I remember reading...the hotter the LEDs run the more mA they'll pull and eventually they could burn out...is this true?
2. Could a POT be added for manually dimming?
brassmonkyballs
BMB
678ma is pretty good....about perfect. The thing about LEDs is the cooler you can keep them the more you can push them with current....they are current driven devices. They don't pull. Your driver is pushing current through them at a constant level and the temperature of the LED has no merit with the driver, no feedback loop. If its pushing 678ma, its going to keep pushing 678ma. Whether the LED likes it or not is of no concern to the driver. In fact, if an LED does not like it and decides to give up the ghost and fail open what happens? All that other current gets equally distributed to the other parallel strings. We discussed this in the risks of running parallel strings and the need for fuses. The driver is a constant current source and that current is going to flow regardless of the state of the LED. What happens if you cannot keep it cool is the possibility of continued heat rise but the current will remain the same.
Whats happening is a function of thermal resistance of the LEDs. You have to think of the driver as a heat source, the current is the flow of that heat and the voltage across the LED is the heat itself (which is basically the LED acting as a resistor). The ability of your entire setup to run cool is a function of thermal conductance. Everything electronic generates heat. The ability of your entire setup to dissipate this heat will determine whether it runs hot or cool....efficient or not. The fact it may run 'hot' is a matter of efficiency and how much 'power' you're wasting.... and then if it exceeds the max rated junction temperature of the device itself determines whether or not it may suffer a shortened life. Reality is its more complex than this and in fact when heat rises so does resistance and with the current constant you'll get more heat, and its a viscious cycle, etc. But cooling it will not lower the current....lowering the current though will lower the temperature it runs at thereby 'cooling' it.
So I suspect the use of the 3x2 channel is enough surface area to conduct away the heat (though marginally) while your 1x2 channel is not. Though its clearly dicey using channel and I highly recommend against it. With a proper heatsink there would be no issue of heat at all. Thats why there's an entire field around engineering and manufacturing of heatsinks.
As to your question of using a pot the answer is no. You need a different driver with dimming capability if you want to turn it down. Or add a parallel string to lower the current...within voltage constraints of course.
oh...and your fans run on voltage, not current...that current is in fact draw in this case...run them at the correct voltage which is probably 12V but check the labeling..... but don't run them off the driver.
Whats happening is a function of thermal resistance of the LEDs. You have to think of the driver as a heat source, the current is the flow of that heat and the voltage across the LED is the heat itself (which is basically the LED acting as a resistor). The ability of your entire setup to run cool is a function of thermal conductance. Everything electronic generates heat. The ability of your entire setup to dissipate this heat will determine whether it runs hot or cool....efficient or not. The fact it may run 'hot' is a matter of efficiency and how much 'power' you're wasting.... and then if it exceeds the max rated junction temperature of the device itself determines whether or not it may suffer a shortened life. Reality is its more complex than this and in fact when heat rises so does resistance and with the current constant you'll get more heat, and its a viscious cycle, etc. But cooling it will not lower the current....lowering the current though will lower the temperature it runs at thereby 'cooling' it.
So I suspect the use of the 3x2 channel is enough surface area to conduct away the heat (though marginally) while your 1x2 channel is not. Though its clearly dicey using channel and I highly recommend against it. With a proper heatsink there would be no issue of heat at all. Thats why there's an entire field around engineering and manufacturing of heatsinks.
As to your question of using a pot the answer is no. You need a different driver with dimming capability if you want to turn it down. Or add a parallel string to lower the current...within voltage constraints of course.
oh...and your fans run on voltage, not current...that current is in fact draw in this case...run them at the correct voltage which is probably 12V but check the labeling..... but don't run them off the driver.
Last edited:
brassmonkyballs
BMB
potentially bad things.....if you drew the circuit as 10 resistors in series and then 8 of these in parallel you'd have a somewhat balanced design. Now add 3 resistors of unknown values in parallel and resolve it. I'm too tired to think about it but it wouldn't be beyond me to try behind closed doors 
I can't recommend it though. Drive the LEDs with the purpose built device and go get a wall wart and a couple of jacks from radio shack for the fans and for 10 bucks you know all will be good.
I can't recommend it though. Drive the LEDs with the purpose built device and go get a wall wart and a couple of jacks from radio shack for the fans and for 10 bucks you know all will be good.Similar threads
