I’ve been looking for a new project and bought the parts from Richard at Mountain Electronics to build up a Nichia 219C triple in a Convoy S2+ host.
I’ve yet to drill the hole in the pill for the wires, but wondered if I should solder the copper spacer to the brass pill. I’ve got a good soldering station so enough heat won’t be an issue. I’ve got spools of Kester 63/37 no-clean rosin core solder as well.
So should I hit it with heat or just use thermal paste? I appreciate the assistance.
Of course. I would also solder MCPCB to spacer.
I would just use thermal paste. Soldering is a better thermal path, but the performance difference will not be significant. This is because the joints between the MCPCB/spacer/pill will be small even using thermal paste and are not the bottlenecks in the total thermal path. So if you don’t mind the extra steps of soldering and the increased difficulty of taking apart, go ahead, but thermal paste is fine too.
For example, a 0.08mm thick layer of 4W/mK thermal conductivity paste spread out over a 16mm MCPCB surface has a thermal resistance of 0.1 K/W. This value is already small compared to other thermal resistances in the path like the LED junction-to-solder point resistance which might be 2 K/W. Soldering might reduce the MCPCB-to-pill resistance by a factor of 20-30. So not a negligible improvement, but not significant.
Solder the MCPCB to the spacer, and the spacer to the pill.
You need a lot of heat or you will get a cold solder joint, I honestly don’t think you can do it with just one soldering iron - you have to get both components hot simultaneously.
I use a hot plate to get both spacer and pill up to solder-melting temps side by side, then carefully but quickly (and using tweezers - not pliers which are larger and will suck out the heat of the spacer when they grab it) transfer the spacer over onto the pill which is waiting with melted solder paste.
Once it cools down, try REALLY HARD to break the spacer off of the pill (by hand) to confirm the quality of the joint. If you can break it, it was a crappy joint and you need to re-do it. If you can’t break it apart by hand, it’s a good joint.
OK, so I’ll heat them up and solder them together. I like the idea of using a hot plate to heat the pill and spacer. I’ve got one for various science experiments with my students.
The rub is this: I’m using a Dr. Jones H17F driver and because it’s double sided, the retainer ring on the pill won’t work. I was planning on soldering that part to the pill. Should I solder all four parts (MCPCB, spacer, pill and driver) or leave one unsoldered in case I’ve got to fix it? Would it be wise to leave one unconnected? Which one?
Or, should I connect the driver to the pill in another way? It fits on the shelf with a little wiggle room. It’s not a press fit.
I did mine in a plain stainless frying pan.
Prep mcpcb with solder paste and emitters then coat the bottom and leave resting on a couple of toothpicks on the side.
Coat the spacer in paste and place onto pan.
Heat until solder melts and copper starts to discolour then give it a hefty tap to force out excess solder.
Place mcpcb on it and wait until solder melts then lift it out of pan and place onto a wooden board.
The solder will stay wet enough for to your adjust leds as needed for a short while.
Thom, yup that’s pretty much exactly what I do, just with a hot plate instead of a skillet.
PBW, you don’t want to have that driver in there when you get the whole deal hot enough to stick the two big parts (spacer and pill) together. Unless you know for super duper certain that the solder used on the driver has very high melting temp, there is a good chance components from the driver could come loose with the heat required to get the spacer/pill up to soldering temp.
A driver can be soldered into the pill (after the spacer/MCPCB has been completed) with a couple localized “spot welds” of solder if you have a good soldering iron and heavy tip that can quickly transfer a lot of heat to a localized area of the pill to make the solder flow and stick. I usually put on my heaviest tip and turn the heat up momentarily to 430C for this part (usually work around 370C).
What you don’t want is a low wattage or small tipped iron. With such a setup, in order to get enough heat for the solder to stick to the pill (now that you have a highly functional heatsink like a copper spacer), you will have to hold the soldering iron on the pill for a long time, and will just wind up slowly heating the entire assembly - and risk getting the solder between the pill/spacer liquid right about the same moment you get the solder to stick to the pill for the driver. Asumming you are doing the with the driver up (spacer down), the spacer will fall off and you’ll have to start over.
Ask me how I know, I’ve developed this method from trial and error. learn from my mistakes. lol.
Thanks again for the info. No, I’ll tack the driver on last with my soldering station up high. I’m pretty used to getting in and out, hot and fast with the guitar pedals I’ve made and delicate germanium transistors and diodes.
I soldered my PCB to the spacer and the spacer to the pill as well. I used a butane torch though. Applied soldering paste between all the components and started heating evenly until everything settled in place properly. I let it cool enough for the solder to harden and then set it in a cup of water. No problems, transfers heat well. DTP to the palm of my hand.