I decided to replace the LED in my silver Convoy S2+ since I didn’t like the tint. I’m going to replace it with a triple LED. I got a spacer from Mountain Electronics and took apart the S2. I was surprised by how little the thermal compound spread between the pill and MCPCB. I sanded the MCPCB before I assembled the S2 so it seemed the pill shelf was not flat. After I drilled a hole in the pill for the triple, I put a straight piece of paper on the shelf and I could see the issue.
Strange thing was that I’ve ran the S2+ in turbo mode lots of times with it getting very hot and stepping down due to heat, but never thought there was an issue. So I sanded down the pill to make it flat.
As you can see, the pill is still cupped. So I kept sanding.
Eventually I sanded enough to make the pill shelf flat.
Unfortunately I had to do the same thing with my last S2+ host (gray), so I’m concerned that others may also have this issue.
I love my Convoys and I’m not trying to be critical of them, but it seems strange that this happened to 2 of them. I haven’t checked my other S2+s, but certainly will if I modify them.
I recently made a triple S2 and didn’t notice an issue with the pill shelf, so it could be just with the S2+.
If you have an S2+ host, you may want to check to make sure the pill shelf is flat before putting everything together.
First convoys used thin alu pcb that is not rigid at all. Screwed down with reasonable force, pcb could flex to the shape of shelf so there were no issues.
If you are using copper pcb and/or copper spacer, soldering it to the pill is simple and effective way to forget about shelf imperfection.
Another way is to sand shelf to the shape of spacer surface. Use double-sided tape to stick a piece of sandpaper to the bottom of your spacer and rotate it pushed in the pill until perfect fit. Even without machine tools or chucks to hold parts, it can be made fast with bare hands.
Sanding flat all surface including rim is too much work IMO.
Personally I prefer an aluminum spacer to keep the weight down, so soldering it to the pill is not an option. When I checked the spacer, both sides were perfectly flat. It took me about 5 minutes to sand down the pill, so not too much effort.
Interesting idea about taping sandpaper to the spacer to match up the surfaces. Given the amount of cupping the pill had, I’m not sure how well the double sided tape would hold up, but I like your ingenuity.
While the aluminum spacer may be a bit lighter and much cheaper than the copper spacer from MTN heat transfer to the aluminum body is much better with copper.
I have 3 Convoy S2+ brass pills that were removed from hosts and replaced with Hoop's copper pills that eliminate the need for a spacer in triple LED builds.Hoop's copper pill also has a unique locking ring design and function.
I've been using cut circles of 0.5mm 17.0W/mk Fujipoly Ultra Extreme thermal pads between the pill and MCPCB to avoid the paste mess.
I just examined the 3 brass pills for flatness and all 3 are good.
You might be able to lay it pill top down on something flat and place a socket slightly bigger than the hole around the hole then hit it with a hammer checking for trueness each hammer blow. It might not come out perfect but much better than the way it comes.
Being I lathe and mill owner for a couple years, that cup shouldn’t happen unless the carriage isnt locked down when using the cross slide. If starting the cut in the center running the cross slide outward causes the carriage to be pushed out from the force of the cut. I have never seen anyone start from the center and run outwards though. You always start from the outside and run to the center. If the carriage moves slightly this cause the center to be high. Locking the carriage down when facing prevents this.
These are all made on a CNC lathe so I dont think the cup is happening during machining unless something is bad wrong.
More than likely I suspect the cup is formed from drilling the center hole to fast with to much pressure. The drilling causes the center to flex downward from the pressure.
If that is the cause, it coud be easly resolved by slowing the drill plunge speed. If brought to simons attention, he might find the problem.
Just my guess.
The shelf of the stock pill is only 1mm thick iirc and drilling requires a lot of pressure. I would say yes it could be the cause. OP believes it was unflat prior to the drilling.
The issue was first noticed when I took apart the S2+ and saw how little the thermal compound spread. I used a flat piece of paper on the pill shelf before I drilled the hole and that’s when I noticed the cupping. It wasn’t until after I drilled the hole that I decided to document the issue.
I also drilled the bottom side of the pill to make the hole, so if anything, the shelf would’ve been domed if it’s thought the drilling caused the shelf not to be flat.
Sounds like its not from drilling then. Something has to be wrong in the machining process, I just couldn’t tell you what. When facing you always run from the outside to the center. The only thing I could think of is some how the axis isn’t running true perpendicular. Either out of adjustment on the machine or its out in the code.
Drilling the hole didn’t cause much heat. I actually drilled in 2 steps, first with a smaller drill bit, then finished with a larger bit just big enough for 2 wires. All the time holding with my fingers.
If you have another factory pill. Try laying something flat across the entire top of the pill. Is the top outsides of pill where the pocket is cut flat, or do they also appear to be cupped. Does it appear the top face was cupped before the mcpcb pocket was cut, I guess is what I’m trying to say.
I’m pretty sure Simon would look into this cupped problem if brought to his attention.
Its possible it has already been fixed and a few bad ones still went through and were sold.
I have one laying around that’s a year or so old, I’ll check it when I get time.
