How to Prevent EOS failure?

Guest

Why EOS are considered latent failures? High Leakage Current or Low Vb due to oxide breakdown are most often manifest at customer's inspection site. These failure modes often escape during electrical test. Can you give me some advice on what preventive measures to implement to avoid reaching the problem to customers? Is the IR reflow help to screen the marginally weak products or products that are potential to EOS failure?

Guest

IR Reflow is a package rel test, so it will NOT help in screening out ESD/EOS failures. Burn-in is a possible screen, but it is also known to induce EOS and ESD itself. Have you checked if ALL your packaging and shipping materials are ESD-protective? We've encountered some reels before that are not dissipative and can cause ESD during shipment to customer.

Guest

Now I see the purpose of IR reflow and Burn-in. The reel used to the parts is not made of plastic but a cardboard. So it is less likely to induce electrostatic charges. I have just received a customer complaint involving short failure. FA revealed that a massive metal burn-out was observed on the die originating from the oxide. The failed units observed to have solder paste overflow to the oxide. Is the solder paste on the oxide induced EOS? Aside from ESD control, what other preventive measures should we implement to minimize the influx on EOS failures from the field?

Guest

Hi Noel,

Cardboard will also tend to accumulate charges over time, so you may want to switch to static dissipative reels if possible.

Where did the solder paste come from? Are these solder bumps on the bond pads? If so then it is really possible for EOS to melt these.

Are the rejects you're getting from your customers already used by them? Because if so, then it is more likely that they are using your part incorrectly. Just investigate how they are using your part so you can correct any application issues.

But if the EOS failures are coming fresh from the tubes (customers haven't touched them yet), then you may actually be seeing either gross ESD or reject mixing from your line.

Guest

Hi Gmailer,

Thanks. Actually we are using solder paste in connecting the die to the leadframe. It plays the same role as solder bump. The returned and failed parts are already used by the customer. In fact, the returned units exhibits crack and carbonized package which suggests that an intense heat was experienced on the package during application.

By the way, could you elaborate further or give me some hints about "incorrect use of parts" leading to EOS problem? I'll try to investigate how the customer use the parts to hopefully lead me to determine the cause of failure.

Rgds, Noel

Posted: Wed Dec 14, 2025 12:25 pm Post subject:

Hi Noel, any luck? what did your customer said.

btw, if you have a carbonized package try inspecting the sample on the X-ray. chances is the affected bondwire have melted and it is usually either the ground or supply pins. note, it is hard or you can't decapsulate a carbonized package.

Guest

Customer can induce EOS in many ways --- misorientation of device, application of reverse or very high voltage, connecting excessive load, etc. All of these are either due to human error or they do not understand how to use the device in actual application. Best solution is to work more closely with customer on how they're using your device.

MB

Posted: Fri Dec 16, 2025 12:56 pm Post subject:

I have experienced before. We manufacture cordless phone and during pcb assembly testing we have a high failure rate. We isolate it on one of the IC component and keep on feed backing it to the supplier. Only to find out that it is our testing process that induced the damage on the IC (eos).

The root cause is that we have 2 pogo pins on the test jigs that are so close together that during testing there was a tendency that one of the pogo pins get shorted to the other. Corrective was to relocate both of the 2 pogo pins.

Guest

i think it would be good to sample fresh units from unused tubes/reels. if no failures from these fresh parts will be seen then case is strong that customer induced the problem. if failures are found from untouched tubes/reels then case is strong that test-related or esd problems may be the issue.