5 ms·
felder 62/36/2 [0] is a flux-cored solder. So you were using flux every time. It's also a leaded solder, which is much easier to get nice joints from than lead
by bitdivision 4y ago
felder 62/36/2 [0] is a flux-cored solder. So you were using flux every time.
It's also a leaded solder, which is much easier to get nice joints from than lead-free.
Flux is great, adding some before you solder is a good tip, and will likely help you make better joints if as you said, you haven't perfectly cleaned everything. The only downside is perhaps some more cleanup afterwards. There's no reason to disparage it's use.
0: https://www.felder.de/products/electronic-applications/manual-soldering-mechanical-soldering/solder-wire/loetdraht-iso-core-ra-18641030.html https://www.felder.de/products/electronic-applications/manua...
- gw99 4y agoYes it’s additional flux that you don’t need.
- randombits0 4y agoDepends. New work versus rework versus what-is-this-piece-of-crap?
- exmadscientist 4y agoDing ding ding. Production soldering isn't easy, but it's much easier to dial in than rework jobs. It's definitely a desirable goal to not need flux for production work. Rework... good bloody luck.
- lightedman 4y agoI most certainly do not use flux in my rework. I am in a zero-clean facility, the only flux allowed is that which comes with the solder paste for stencil and SMT placement - on initial production work. Fluxless rework isn't difficult at all. I do it dozens of times a day.
- exmadscientist 4y agoI'm having a little trouble believing you unless by "rework" you meant "reflow". Production reflow without extra added flux is normal. It's the goal for every assembly line anywhere and is almost always achieved. Rework of existing joints after assembly without flux is just about impossible. The oxidation of the joints guarantees it. Many times the flux in the paste or wire is sufficient, but this is more true for through-hole than for surface-mount, mainly due to joint geometry and available volumes. If you are regularly reworking old fine-pitch surface mount joints previously assembled with no-clean flux, without flux of your own... then I'd like to know what materials you're using, because that is not possible in general. The "no-clean" fluxes are notorious for causing rework trouble, and they are difficult to clean off once reflow temperatures have been achieved (that's why we call them "can't-clean"). They are never active twice, so they must be removed and replaced once heated the first time. We have to use special solvent flux removers around here (based on DuPont's hellish Vertrel XF) to get the crap off just to be able to access joints again. (This is on assembly work done out-of-house or on COTS devices, made with who-knows-what.)
- lightedman 4y ago"I'm having a little trouble believing you unless by "rework" you meant "reflow"." Nope, we don't use flux at all for rework. As the majority of our boards are solid metal, we just slap them on a heater and poke the components down or slide them into place. Flux isn't needed. Even with crappy RoHS tin solder. If you're getting significant-enough oxidation that interferes with rework with the product fresh out of the oven, or even a day or two later, you're doing something extremely wrong. Check and make sure your production paste isn't contaminated.
- OJFord 4y agoSo flux is 'a crutch for people who don't know how to solder', because that knowledge would lead you to use 'expensive' flux-cored solder instead of 'shitty solder' plus flux? I really don't think either's wrong, and anyway I assume 'production engineering' is not using how-to-solder tutorials.
- gw99 4y agoShitty solder + flux is wrong. Simple. That is part of production engineering which is mostly about making sure that your product doesn't come back again because it's broken.
- OJFord 4y agoYes, you're concerned with the yield on tens, hundreds of thousands of boards. It's a different world from a hobbyist's one-off project. This advice was never intended for you, and appropriate advice can differ between audiences.