What Causes Chip Blow Off During Reflow?

Phil
Welcome to Board Talk. This is Jim Hall and Phil Zarrow of ITM, also known as Pick and Place, the Assembly Brothers, here to attempt to answer your assembly and process questions regarding SMT.

Jim
Today's question comes from D.L.

What can cause a 0603 passive component to blow off the circuit board during reflow as if a firecracker was put underneath? That's a pretty graphic description.

Phil
That's radical. Well, the first thing that comes to mind is a firecracker under the component? Maybe you have some prankster in your assembly department.

A couple thoughts on this. The question we want to know is if it is accompanied by splattered solder paste?

Jim
I am curious why this person would choose to use the words "as if a firecracker blew off" because I'm thinking of how you're going to know this. The board goes into the reflow. It comes out. The part is missing. Why would you describe it as a firecracker unless there was some evidence of some sort of explosion, which would lead us to believe that it would be solder paste-related.

Phil
Generally, we think of solder paste-related violence - eruptions of a volcanic nature. Typically, we find this is similar to out-gassing and that's usually due to too rapid a preheat.

Jim
Preheat slope too high.

Phil
Now, what if there isn't any solder paste there? The component's is gone.

Jim
My experience as a former oven manufacturer, listening to complaining customers, is that many displaced components relate from oven problems. Conveyor vibration has always been the biggest.

Align your conveyors, lubricate your conveyors, because if they're vibrating during the reflow cycle, during the ramp-up and before you actually start wetting, those components can move.

Phil
The key word here being "blow off" is improperly set or maladjusted convection. Either the blower's set too high, somebody's tooled around with the buffers and some of the other stuff, or since a number of machines have adjustable convection levels, somebody cranked it up too high.

Jim
Or a more benign maintenance issue is you've got some flux or materials blocking up some of your hole pattern causing a high-velocity jet or an improperly directed jet in one part of the furnace. So, your 0603 goes under there, gets hit by this blast of air and, whoosh, it's gone.

Make sure you're really profiling the actual board you're running. Put a thermo-couple on the area; the type of chip that you see being blown off; measure what the real, actual preheat slope is.

If you've got conveyor vibration, you can usually feel it by just observing it. Air flow issues are a little more subtle to diagnose.

Phil
We hope that helps. This is Phil and Jim, the Assembly Brothers. So, again, in the meantime, I can highly recommend - don't solder like my brother.

Jim
And don't solder like my brother.

Phil
And don't let the kids fool around with the convection controls on your oven.

I saw this many years ago relating to 1206 100nF ceramic caps. The component had moisture ingress and out-gassed under reflow. The board would go into reflow with everything positioned correctly, but on exit from the oven the relevant caps were either elsewhere on the board or nowhere to be seen.

We had to sample test all our stock of that component by placing them on a hot plate and watching what happened to them. The component supplier actually accepted that the problem was his component, carried out the test at our premises and replaced the faulty stock.

We've seen this issue with high humidity. Greater than 70%RH (rainy day). Solder paste absorbs it and components are missing or severely shifted after reflow with craters around the pads. Pastes was dripping off the spatulas.

Chip blow off can also result from voids in the component terminations which are covered by plating layers. Upon heating trapped vapor or plating solutions can violently out gas and propel the chip off the board.

Moisture absorbed by the solder paste from the air can cause parts to be moved from their correct positions during reflow. They can move with enough force to be blown completely off the board; hence the "fire cracker" description is applicable. Try reducing the amount of time that the solder paste is exposed to the atmosphere before printing. It is important to go from the paste printer, straight to the pick and place, and straight to the oven, with minimal delays. Don't leave large amounts on paste on the stencil for long periods, either.

It could be caused by PCB warpage or via hole trapping air blowing out (if this is a via hole plugging design) during the IR reflow process.

Is it happening to all 0603 or just one part number? If it is just one part then it is not happening because of paste or profile. If it is a single part you have moisture in the part somewhere. Sometimes moisture (plating solution) can be trapped in voids in the plating of the terminals.

If there is a via under the center of the 0603 that it is covered with soldermask on both sides. Is it possible for impurities to get trapped in the via barrel during fabrication and heat up and blow out the hole causing the part to fly off... just wondering?

Have seen a similar problem. This was caused by a component placed nearby. It was a small relay, which had a small hole for releasing air during reflow. The air from hole was blowing the SMD component away from the PCB.

I have had exactly the same problem. It wasn't the chip or the paste or the profile that was causing the problem. It was a tantal component next to the chip which was "out gassing" and blowing the chip out of position. Try to remove the surrounding components and see if the issue disappears. Initially we pre-baked the tantal until the component supplier rectified the issue.

Couple of years ago I faced this issue in chips with sizes 0805, 0603, also in tantalum capacitors with cases types C/D, when such components should be placed on PCBs that had some vias closed by solder mask at positions just below bodies of the components. I also remember we should solder such components on PCBs with their date code 1 or 2 months expired, although all of them were well vacuum sealed. We did not try to bake the PCBs prior the assembly and soldering. Then, after reflow I noticed that some 0603/0805/Tantalum capacitors have been blew up from their locations.

During inspection of the PCBAs I observed that in all positions where components were blown up, vias existed and solder mask that was covering the vias have been exploded too, reminding me of a geyser cone. To complete my suspicions I took the components blew up from their positions and I inspected them on their bottom sides. They had circular marks on bottom sides as if they have been hit by a shot of gas (these marks were easy to see with a magnifying glass x8). I decided to run second PCBA batch with all PCBs properly baked prior the production. The issue was not observed anymore.

I just listened to your discussion on chip blow off in reflow. I can't believe one subject never came up. What about absorbed moisture in the solder paste prior to reflow? That used to be a really big issue for us about 20 years ago down here in Florida. Maybe its not an issue anymore because of better practices and improved technology in the solder paste but I just wanted to bring that up.

Moisture in the smaller components can cause them to blow off, specifically if the ramp rate is too high.