Acceptable thru-hole solder joints, IPC

IPC-A-610 class 2,3 would require >= 75% vertical hole fill.

Doing some stuff with J-STD for satellites-

I don't have the IPC docs with me, but from a draft of IPC-A-610E

Space/Military is Class 3/A, which adds some stuff to Class 3, mostly to do with staking and gold issues (replacing some older NASA etc. specs).

Anyway, I would say if you can't inspect the fillet on the solder (secondar

for Class 1, and even if the primary side looks okay.

Class 1:

"This is the most lenient when it comes to allowing potential defects. Thin k of the electronics assembly in a very cheap toy, for example. The PCBA wi ll be hidden well away within some stuffing or plastic enclosure and, as a result, the quality of solder joints or component positions won?t b e a priority. It won?t be expected to last very long and often, has to be assembled to a specific price ? usually translating to "as c heap as possible", whilst ensuring it still functions as expected."

We do industrial products and AFAICR we use minimum 70% via fill

I know but they don't talk about 360 degree capillary effect on one side. I don't mean a valley with would be ok and still wetted but really at the pin. In all companies I've ever been at that was flagged as unacceptable.

Aha! Now we are getting somewhere. If you happen to be within reach of the IPC binder and have the time, could you look up which paragraph? I don't have IPC documentation because my stuff is usually governed by other standards and I don't have much to do with actual production.

This is more for commercial grade stuff.

Only says "We can?t find what you?re looking for" :-(

It's not about via fill or what is often called barrel fill in America (I always get thirsty when I hear that ...). It's about downward capillary around the pin, from the component side down into the solder. Like a fence post that looks loose. IME that is a recipe for transport failure and we had exactly that happen.

These links have a timeout.

At what kind of facility is this being soldered?

My boards are built at a smaller facility and they typically find issues on a few boards that get hand touched up to fix. I spoke with a Motorola guy back when they made a boatload of cell phones and he said they don't do anything other than test. So they optimize their process so they have very few failures and little problem with reliability. I'm wondering what they can do to improve their process enough that it's not worth doing either an automatic camera inspection.

Seems to me it shouldn't be hard to control the process well enough that defects are rare. Any idea what can go wrong that causes these issues?

A contract assembler but I can't reveal which one.

It's a hand-soldered part so probably dependent on who did the particular job. Their automated SMT soldering process is always nice and of superb quality.

I am looking for some ironclad written standard. What I have seen in IPC usually talks about one side only, how the solder joint should look. It doesn't state cases where the solder side looks great but the component side indicates a potential failure down the road, only about barrel fill. I know when it can fail and usually does in the field but there is a slight "convincement issue" :-)

Thanks, that's it!

I thought the link JM posted had the full spec including both sides.

Yes, it does. I just could not read it because of a time-out issue. Today I could.