Design Does anyone know about LED wiring?

[JoeML]

Before my brother passed away, he bought 18 3W LED par cans from China really cheap. go figure, lol

Anyhow, after having them running for around 2 hours full on, the R, B, or G set didn't display. I took the can apart, to notice that the resistor must of gotten so hot, it melted the solder and disconnected. So I re soldered it and the same thing happened. I looked at the resistor and it says YAHE 5W10ΩJ.
There is 12 LED's of each RBG, 1 resistor for each color. Even if it is wired in both series and parallel, wouldn't I still need a even number for the wattage unlike the odd number of 5??

I was wondering if i replaced the resistor with a higher wattage... maybe a 6 or higher or replacing it with a aluminum heatsink type resistor.

Any thoughts would be appreicated greatly! BTW, they are bright, so I would like to keep them, instead of selling them.

 

[n1ist]

In general, you use a resistor with a higher wattage rating than what you need, since it is a max rating with proper heatsinking. In this case, it's marginal. You could replace it with a higher wattage part, but it wouldn't run much cooler unless it is much bigger (more surface area to dump heat from) or heatsunk.

/mike

 

[ajb]

If the solder is melting, then a higher power resistor won't help. The problem is that the resistor is dissipating more power (heat) than can be adequately handled by the overall assembly (resistor, PCB, housing). You could replace it with a higher value (resistance) part which will dissipate less power, but you will also lose output--it dissipates less power because it allows less current to flow.

If you really want to put any money/effort into making these work somewhat reliably, see if you can add a small fan to the housing.

 

[kendal69]

Those lights are not designed to burn for two hours they are stage lighting LED and only designed to burn on short bursts.

Can they be rewired to burn longer...that I don't know.

 

[Les]

Those lights are not designed to burn for two hours they are stage lighting LED and only designed to burn on short bursts.

Can they be rewired to burn longer...that I don't know.

Might wanna change that to DJ Lighting. Real stage lighting should have no duty cycle.

 

[JoeML]

You could replace it with a higher wattage part, but it wouldn't run much cooler unless it is much bigger (more surface area to dump heat from) or heatsunk.

/mike

So I could buy a 50 watt 10Ω resistor and it should be better?

And adding more or less resistance would probably be a bad thing, resulting in LED failure?


I was also wondering about the fan too. I was thinking of getting a 3" fan and mounting it somehow to suck the heat off the resistor. Cause I don't want to blow on them, resulting to suck dust into the fixture. Or I could buy filters for the vent holes?

As for the stage lighting / DJ lighting, I understand what you mean. But my understanding when my brother ordered them, was for bar lighting gigs. They did a great job for front lighting, but that was a outdoor cool night. Maybe because they are *[FONT=&quot] air convection, they are made for outdoor shows or the manufacture didn't know how to make the lights work, lol.[/FONT]

Anyhow, I guess I'll try looking for a larger wattage resistor and giving it a try, while taking heat measurements on the resistor and solder. I was also thinking, if a higher temperature solder would fix the problem.

 

[n1ist]

A larger wattage part won't help by itself, since it will still get as hot. One with larger surface area or that can be heatsunk to the case may help, as would adding a fan.
/mike

 

[JD]

A larger wattage part won't help by itself, since it will still get as hot. One with larger surface area or that can be heatsunk to the case may help, as would adding a fan.
/mike

That would probably be the trick if you want to modify the unit. Epoxy the resistor to the case to dissipate heat and jump wire it to the board. Its a matter of moving the heat from the resistor so it does not heat up enough to de-solder itself. Poor design, and a bit wasteful of energy, but you got what you got, so I guess make the best of it.

 

[dcollins]

For clarity:
Resistors dissipate power. They also limit the current passing through them (the amount they do so is based on their resistance and the voltage applied to them) - they 'hold back' some of the power from the LEDs, it seems. The higher the resistance, the more they will decrease the current running through them to the LEDs. This will make the LEDs dimmer and increase the power dissipated and therefore increase the heat generated. A lower resistance will increase the current running to the LEDs, probably burning them out. You want to keep whatever resistance is provided.

The watt number is a rating, not an electrical property (well, it is, but...) Just like how a dimmer can't supply more than (say) 2400W safely, the resistor can't take more than 5W. Power = Voltage * Current and Voltage = Current * Resistance, so Power = Voltage^2 / Resistance. A higher 'rated' resistor (say 50W) will still dissipate the same amount of power (as determined by that equation) and generate the same heat. Your problem isn't that the resistor can't handle the power, but that the housing can't handle that much heat. As previous posters have said, the only ways to fix that are more ventilation, a fan, using it over shorter duty cycles, or using it at a lower level (dimmed down).

 

[gcpsoundlight]

It probably blew in the first place because of a short circuit, in which case no higher rated resistor is going to help!

 

[ajb]

For clarity:
The higher the resistance, the more they will decrease the current running through them to the LEDs. This will make the LEDs dimmer and increase the power dissipated and therefore increase the heat generated.

Nope, a higher resistance will decrease the current through the resistor as well. Since the resistor is in series with the LEDs, the current through the resistor will always equal the current through the the LEDs, and ecause the Vf of the LEDs is basically fixed, so is the voltage drop across the resistor (within the regulating tolerances of the power supply). Hence a higher resistance part is a potential remedy if the associated loss of output is acceptable.

 

[JoeML]

I've been busy since the last post. But I have taken a broken fixture all apart and then fixed it, to work.

I've taken measurements using a Infrared Thermometer and recorded the following temps from the back of the panel where the heatsinks for the LED's are located. Normal temp with power off is 75°F. 1st start-up is 83°F, then after 5 minutes, the temp was 303°F. Max recorded was 335.4°F. At the resistors from 1 min after start-up to max for Blue was 114°-144°F, Green was 163°-208°F, and Red (most problems with this color) 219°-260°F.

So getting another resistor, i believe wont cure the problem, as the resistors are sitting right on top of the heat sinks (loose wires). I do have enough room to mount a 3" 24v fan in there, but wondering what CFM number i would need?
I would be mounting it using 1" stand-offs to the inside body and having the airflow through the back of the fixture.

 

[n1ist]

Ouch. I would definitely put a fan in there. At that temperature, you are above the ratings for most LEDs; almost into the solder preheat temperature. Since elevated temperature degrades the performance and lifetime of the LEDs, that's a problem.

Unfortunately, I am not sure how to calculate the fan volume.
/mike

 

[ajb]

Wait, so are the resistors already mounted to heatsinks?

Any chance you could tell us the voltage across the resistor during operation? With that figure it'll be possible to determine the actual dissipation of the resistor ( P=V²/R ) and I, for one, am quite curious in train-wreck kinda way.

When in doubt, start with a higher flow rating. You can always slow the fan down if you need to. And you're setting this up so that you're moving outside air through the fixture, not just stirring the hot air around inside, right?

 

[JoeML]

The resistors are not mounted to the heatsink. If you could tell me how to put a picture in here, id do that.

As for the voltage across the resistors, I'll take it soon.

And I will be having the airflow facing the top of the par can, where the power wire is located. I don't want to be sucking in dust from the vent holes into the light. I guess, i figure, its easier to dust from the front LED side, then to open the can up and take everything apart.