WLCSP-12 reflow

[Buriedcode]

Hi,

I saw a previous post about soldering these manually (hot air) here: https://www.eevblog.com/forum/projects/soldering-wlcsp-for-prototypes-by-hand-placement-and-reflow/

The general consensus was, using hot air on the underside of the board, because the chip is so light, even gentle air flow seems to blow it about.  As a test, I desoldered one from a scrap board, cleaned and fluxed the pads, as well as re-tinned the chips pads (not reballed).  As soon as the hot air went on it floated ontop of the flux, and seemed to just not have the weight to sit on the pads.  Of course, trying to steady it with tweezers was a task in itself.. so is there a reliable way of doing this?

 Perhaps using much thinner flux - I'm using Amtech 559 jelly right now.

The actual job is on a samsung phone board, but I daren't touch that until I can reliably - not as in an every day job but reasonably - solder one on a board.  I can't really use hot air on the underside of this board as there is a large connector there, which could drop off under its own weight.  I guess I could strap it up with kapton tape.

Please note, I'm not trying to "hand solder" with an iron.  I'm fairly sure I could dead bug it with 36AWG vero wire but thats a bit of a fudge, and will have to take 2A through it anyway.

[Dubbie]

I have never had any luck doing these by hand with hot air. The reflow oven plus stenciled paste does it perfect every time though.

[mark03]

I did a couple of 4-ball GaN FETs with tacky flux and hot air.  Those are even lighter (the width and height of the chip body are comparable!).  I don't understand why so many complain about the hot air blowing the chip around.  Just turn down the flow, and blow only from the top.

[thm_w]

WLCSP10, hot air on top side of board, set flow rate to ~4/10 on the 858 station dial. Flux is MG 8431, similar paste flux, softens quickly with heat. Out of ~5 boards I had to scrap one IC, lost the solder ball.

Don't think I can give any advice as its probably a hell of a lot harder on the samsung board due to a lot of internal layers sucking away heat? Preheat if you have it.

As a test, I desoldered one from a scrap board, cleaned and fluxed the pads, as well as re-tinned the chips pads (not reballed).

If you are saying you removed the balls, that is going to make it very very hard to re-use, if not impossible. You will have to re-ball or buy new IC for a good test.

[Buriedcode]

As a test, I desoldered one from a scrap board, cleaned and fluxed the pads, as well as re-tinned the chips pads (not reballed).

If you are saying you removed the balls, that is going to make it very very hard to re-use, if not impossible. You will have to re-ball or buy new IC for a good test.

Yeah thats was I was thinking.  I have some new ones on order, in a pack of 5, so these will have their balls so I can test from there.


As for "blowing the chip about" it didn't really blow it above, it was more floating on top of the flux, even after the flux flowed. I'll set up my hot air to be perfectly perpendicular to the board, that might actually force it down enough.  It's tough as there are plenty of passives surrounding it about 0.5mm away.  Thankfully these stay put under hot air, as expected (they're 0603 thankfully).