Component damage
Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.

Component damage

by Warrior84 on Thu Oct 08, 2020 11:04 pm

hello my name is Victor !! I would like to know if you can support me with a problem that I have, i work in the smt area, in the last weeks we have a problem with a part number which has no value after going through the reflow, 5 out of every thousand components are damaged no all.
is it possible that the component gets damaged in smt process? or is it a supplier defect.

Posts: 1
Joined: Thu Oct 08, 2020 10:41 pm

Re: Component damage

by adafruit_support_mike on Wed Oct 14, 2020 2:41 pm

Without knowing the specific details, it's impossible to say. A defect rate of 0.5% seems unusually high for anything made by the major manufacturers though. There's at least a chance the failures could be the result of damage that occurs during the reflow process.

The most likely suspects would be overheating, thermal shock, and steam damage.

Overheating won't damage the functional parts of most electronic components, but can mess with the package and connections. For an IC, the silicon die, leadframe, and plastic enclosure all have different rates of thermal expansion. The parts are designed to stay roughly the same size within a given temperature range, but outside that range the differences in expansion can be large enough to do damage.. breaking the connections between bond wires and the leadframe or the silicon, cracking the silicon die off the leadframe, or even cracking the silicon die itself.

Thermal shock is similar, but tends to be more of a problem for SMT capacitors. Most caps are made from a barium titanate ceramic because it has a high dielectric constant, but the ceramic is also fragile. The outside of any SMT component will cool faster than the inside, and capacitors that cool too quickly can crack like ice cubes in warm water.

Steam damage also creates pressure, but from a different source. The plastics used for most IC packages will absorb a small amount of water vapor from the surrounding air. That isn't a problem at room temperature, but 1cc of water expands to about 1300cc of steam when you heat it past 100C. If that steam is trapped inside the plastic of an IC package, it can generate enough pressure to break the plastic, damage the bond wires, damage the silicon, etc.

A failure rate of 0.5% suggests the process causing the damage is just at the edge of its safe operating range.. the maximum reflow temperature is just a little too high, the temperature ramps up or down a little too quickly, the soaking times are a little too short, etc. It's normal to push those parameters to make the reflow process as fast as possible, to maximize production, and increased failure rates tell you you've gone as far as possible.

If you're seeing an increased failure rate in a process that's been working without problems for a while, I'd take a serious look at steam damage. The good news is that you can solve the problem by baking the components before running them through the reflow process. The usual process is to heat the chips to about 80C and let them soak for a while, then ramp the temperature up to about 150C and let them soak again. The 80C soak gives most of the water vapor time to leave the plastic without causing unnecessary pressure, and the 150C soak gets rid of any water that's left. The manufcturer will be able to tell you the temperatures and soak times for a specific device.

Posts: 63074
Joined: Thu Feb 11, 2010 2:51 pm

Please be positive and constructive with your questions and comments.