So, being the newbie that I am, and not knowing any better, I want to solder some LED’s directly to copper wires, nothing else. I have some Cree XP-E2 emitters that are currently mounted on 20mm aluminum stars. I want to remove them from the stars and solder them to 18ga copper wires. I was then going to encase them in clear epoxy. I’m not looking to run these at max output, but I’d like to know how much (milli)amperage these could handle mounted this way. Also, I suppose adding a third wire down the middle to pick up the thermal pad would help, but I’m not sure how I’d get the three wires soldered that close together on this little emitter base, without bridging. I’ve been reading everything I can here, and in data sheets, and at Engineering Toolbox. But, I still can’t figure out what level of heat is being generated and how well it can be dissipated through various materials.
I’ve seen that some emitters can be over-driven up to 2 or 3 times their amperage rating when there is a good thermal path. So, since I’m fine with under -driving these, I thought maybe I could get by with less than ideal thermal management. Can someone help me with some off-the-cuff guesstimations? Also, how to do the soldering, since there is no base, no pcb or anything to solder everything to. I have some solder paste, and it does seem to be easier to use than regular solder, but I’m still having trouble learning how to solder small parts.
If you’re not going to drive them very hard, why 18 gauge wires? 18AWG wires are going to be thicker than the pads you want to solder them to. An unnecessary headache. You definitely need a thermal path that leads somewhere it can dissipate.
What exactly are you building?
I might have to go with 20ga if the 18ga wires won’t fit, but the reasoning is that more copper = more heat can be conducted away, which in turn means more power can be applied, creating more heat! J)
I suppose there is a lack of experimentation in the direction of minimizing thermal management and yet still seeing how much amperage an emitter can take. Lots of people push emitters, but they usually do so by pushing thermal management to the extreme so that more amps can be pushed to an emitter. I want to see what an emitter can take with little heat management help. Not none, just little.
I’m looking to light up these little XP-E2 emitters with less than 1/4 of their rated amperage. I intentionally chose emitters that are over-rated for the amperage range I’m targeting, because I know I will have such a low amount of heat dissipation. As for what I’m building, it will be something like a LED rope light. LED rope lights that can be bought use cheapo 20mA LED’s and don’t give out much light. So, I thought I’d see what was actually possible and maybe push that boundary a little. I will have at least one disadvantage. They use clear PVC. I will be using clear flexible epoxy. The epoxy is slightly less clear and slightly less flexible than the PVC.
Some problems with those - they aren’t flexible enough, they aren’t bright enough. And they don’t have the right size and shape for what I want. Also, the light only comes out on one side. I want light coming out all around.
Wires to the pos./neg. pads aren’t going to conduct any heat away from the emitter. Only the center thermal pad conducts heat from the emitter. For what you plan to drive the emitters at, I’d go with really thin wires. Thick wires will do nothing but make it harder to solder.
If you ran 3 wires (awg 24 solid) with the center soldered to the center pad and the outer wires to the +/- pads it might be ok at 350mA, with nothing at all you might not be able to go over 100mA but that’s just me guessing.
I’m thinking of doing this when boosting my Trustfire X6. It already has a MT-G2, but the MT sits a little low in the reflector opening. At first I thought I was going to raise the emitter on three pieces of copper, but I realize that one should work just fine, as long as I leave enough room to solder the wires to the underside of the pad.
I put an XM-L T6 on a pair of 22AWG solid wires coming out of one side. It runs fine at 350ma without overheating just standing up in the air.
Interesting thought about how effective the power leads are in heat dissipation. Both the power pads a have metallic via up to the top of the chip. This should actually help draw heat form above to the wire. But the thermal pad in the middle is directly below the die and has a larger area. The substrate is likely an engineered material for maximizing heat conductivity (but still very low). So if you made the positive or negative wire part of the center pad as well, you should have the best of both worlds.
For such a small emitter, consider “rolling” a 16 AWG solid wire (flattening) right at the connection. Loop the wire so it comes out on both sides of the emitter. 20-22 AWG for the “single” connection will work fine. Just use a size you can manage easily.
Solder it with a very small tip. Pre-tin both the emitter and the wire. Use a small amount of flux and hold things together with a “3rd hand”. Then it is a simple matter of touching the wire with the iron and its done.
Thermal problems are about the difference in temperature of the die to the extreme where the heat can reach. Without the mass, you will not accomplish good heat distribution. There is a state of saturation where you reach equilibrium.
If you want to find your output limit with your setup, just ramp the power until one of the wires pops off (before you pot it f course). Both the emitter and the solder have about the same thermal limit. Note the power and cut it back by whatever margin you deem appropriate.
This will be a horribly painstaking job to do. The problem is not the center, but to solder the ever so tiny + and - on correctly. If you are still set on doing this, you can solder the center with the - and then only separate the + to save you some work. I suggest using XML because it is much bigger and easier to handle compared to the XP size. Good luck!!!
Thanks for all the comments. I guess I will try this:
Pre-tin the wires and the emitter’s pads.
Solder to small wires and use a small tip.
If I think I can make it, solder a third wire in the center.
If I can’t get a third wire in, use a larger wire for the - and cross it over to center.
Start powering it up with a low mA and slowly raise the amperage, watching the heat.
I see no reason not to use two wires with one bridging with the center pad, make it a nice and thick wire and it will run at 100+ mA, not sure about 350mA though..
Look at this chart from AWG concerning copper conductor properties. Especially note the maximum ampacities listed for various wire sizes! There is a note below which says “The current ratings shown in the table are for power transmission and have been determined using the rule of 1 amp per 700 circular mils, which is a very conservative rating.” Wow! What an understatement that is!
I look at the resistance ratings instead to decide which gauge to use and estimate voltage drop over the length of wire. This is more critical in flashlight use. If you drill some dome sized holes in a piece of wood spaced the way you want then overlay them with pieces of double sided tape you can use that to hold the LEDs while you solder the wires in place.
Wire ratings typically consider temperature rise so you don’t cause fires with a huge amount margin. Typically, resistance is not an issue for most applications in wire harnessing or magnetic coils. They are simply managing heat or safety issues.
As Scott says, we care about voltage drop across terminations. It’s a tough tradeoff when you start going over a few amps.
Use that wire flattening technique. Copper wire deforms easily and you can get really close contact for the negative-center pad termination.
I do use bare wire and use stripped Teflon insulation as a sleeve. Alternatively, you can use the silicon rubber insulation. Next order from MTN Electronics will have a variety of wire sized just for the insulation.
I wish I had a way to measure the temperature of 350ma XM-L mounted this way. There is nothing on the thermal pad and only 22awg wires in one direction on the +/-. It just puts out light and the solder is staying put.
edit: I just ran the XM-L for 1 hour at 350ma. The wire temperature rise (as heat sink) was barely detectable. I could pinch the emitter between my fingers without issue. With a little more wire mass and sinking the center pad, these could easily handle 700ma and maybe even 1A in this arrangement.
Thanks for all this great information you’ve given me. Since watts is watts and heat is heat, I’m going to assume I can do similarly with my XP-E2’s. I’m not going to crank them that high, but it’s good to know I’ve got a little head room.
