Today I was testing new clamp meter with 30Q (~4V) and triple 219c on kaidomain PCB with TA board in S41. I saw 17A and light turned off. The reason is that positive wire desoldered itself from the led star. How to prevent it from happening in the future?
At least in the first minute that should not happen if the board is in good contact with the copper head.
But the heat is generated very fast (60W) so you don’t have ages before the head reaches unsoldering temperatures. The S41’s high performance depends entirely on soaking up the heat for a while, but can not even shed a small fraction of that 60W, maybe just 10, so the power must step down fast!
Is it a copper mcpcb? I had the same thing happen on one of my D4s. The problem was just a bad solder joint. I fluxed it and turned up the heat on my iron and there was no problem after that.
Best is to pull the MCPCB and look if the heatpaste has a constant pattern with a very thin film
At such a high current the heat path from star to copper head needs to be very good
Is the Kaidomain MCPCB thinner than the stock one?
Either a bad thermal path or a bad solder joint. Mcpcb needs to sit flat on the shelf and be held down with sufficient pressure.
I try to fit a solid single conductor on the positive connection, hammered flat, and use high melting point tin-lead solder. With either a bit of Kapton tape, or silicone RTV on top. Behind that, transfer to usual stranded.
The consequences of the +V lead de-soldering and maybe flipping up and shorting against whatever don’t bear thinking about.
Yes, kaidomain MCPCB is thinner than the stock (1mm vs 1.5mm). The heatpaste under MCPCB looks good.
Mcpcb sit flat on the shelf, I had a problem to pull out optics.
I think that it was bad solder joint. Solder was “reused” (meaning no new solder was applied on led star) when I was changing the driver. Thank you all for help 