I spent 25 years operating my own hybrid shop working with surface mount components, soldering materials and procedures. Dealing all the while with it's associated surprises and problems such as yours. Now-a-days I get paid to investigate and attempt to cure problems such as yours.
This one's a freebie.
This is the best I can do with the amount of information you supplied. My suggestion is a quick look at the problem that should only take about an hour or two of your time that will hopefully point to a solution.
You have not said what the terminating metalization on the capacitors is. I am assuming solder over nickel barrier.
On a component as small as 0603 gold embrittlement or other contaminant of the solder joint is a prime suspect.
Look closely at the capacitor you say you "popped off" with finger pressure, and the remaining solder on the board.
If solder remains on both, embrittlement is probable. Try a "quick and dirty " test (do this test with some solvent cleaned and some "out of the box" components, to determine if the supplied capacitors are contaminated in some way). Apply some solder to a few land patterns on a test board. Wick off the solder, thus removing the gold or other contaminant from the board and leaving the land tinned. Hand solder capacitors to these AND several other un-tinned lands. Try as best you can to simulate the production soldering conditions ie peak temp & dwell (if these tests yield no results, those factors will bear further looking into). Pop them off the way you did when first discovering the problem. If you see a dramatic difference, embrittlement or other contaminant is at work.
If the metalization has been removed from the end of the capacitor, or the solder remaining on the board shows the impression of the the capacitor terminations (indicating that the solder seperated cleanly from the capacitor)other than a "cold solder joint" (also determined by looking), the capacitor is the prime suspect. Investigate the metalization of that particular lot of capacitors on your own and with your vendor (Mfg.). Do some comparisons of other vendors products or implement component cleaning procedures prior to assembly.
This a "shooting from the hip" reaction to a line down problem. If it restores immediate production, take the time afterwards to refine the solution, improve economies, and QC monitoring (component qualification, incoming inspection, plating selection on the boards, solder selection, process controls, etc.)
I hope this has been some help! Good luck.
Jim Arnold
reply »