Got old, unsoderable component from Mouser

[dmg]

Hi,

I just sent a board off for assembly along with some critical componets I had purchased and they complained about solderability issues with a QFN packaged component. Turns out the solder didn't even wet the pads, balling around them. They sent me pictures and I immediatelly saw the issue. It's an Analog Devices part packaged in one of their LFCSP packages, but in the old molded variant instead of the newer sawn one. I checked and in fact they stopped making the molded version in may 2014, and in fact the component has a datecode from early 2013, which is also confirmed in the traceability data from the Mouser order. So a 6 year old component, unsolderable after all that time. Can't find which materials where used for this component's lead terminations but whatever that was it did become unsoderable.

Anyone has had a similar issue?

[jbb]

I thankfully haven't been stung in production,  but I have had wetting problems with old stuff from the parts bin.

... sent a board off for assembly along with some critical componets ... the old molded variant instead ... the component has a datecode from early 2013...

That is quite old.  My guess is that it might not have been stored properly.  How was it packaged when you got it and how have you stored it?

[KL27x]

I'm curious if Mouser at least listed it as discontinued and "obsolete/ not recommended for current projects?"

[splin]

Are there any cost effective ways of cleaning oxidised/contaminated components like these? Dumping old resistors or the like is one thing, but many ICs are pretty expensive and it would be painful to have to scrap them.

[EDIT] I'm assuming that using more aggressive flux might not be an easy solution due to quality/reliability issues, compatibility with other components or residue cleaning difficulties etc.

[T3sl4co1l]

I can think of many chemical and mechanical processes that might work, but choosing one that's cheap and reliable will be a huge challenge.

For example: soaking in a mild acid, or complexing agent (e.g., EDTA?), may help to remove oxides, but there's little guarantee that the bare metal will remain pristine until soldering.  Solutions could be administered to whole reels (notwithstanding their labels), but washing and drying the reel would be a huge challenge.  (Needless to say, a baking cycle would be needed for MSL parts too?)  It might honestly be better to de-reel, process, and re-reel the parts.

Mechanical processes like wiping, burnishing, brushing, sanding or cutting, could very easily develop huge charges, resulting in destruction by ESD.  Or apply stresses to the part causing internal damage, if not cracking that's externally visible.  (I would have to guess, a vibratory polisher with gentle media like nylon + walnut shells, with a water / cleaner bath, might be okay.  Still have to deal with sorting, drying and taping the parts.)

Better yield may well be had, simply from heating it higher/longer, and using a slightly more active (but still not corrosive) flux.  This runs the risk of damaging parts, or other parts besides the chip in question, but maybe that's worthwhile?

Tim

[ConKbot]

Aren't water washable fluxes generally more aggressive than no-clean?  If your other components are ok with a wash, (I.e. no unsealed switches or pots) it could be an easy fix to switch pastes, at the cost of having to clean and dry the boards. If you can use leaded solder, a paste with RA flux will solder it unless its really really bad, but again, cleaning is a must.