Why BGA soldering ball always crack(2)? The composition of PCBA bonding-force

This article will talk about the composition of the PCBA bonding-force.

The PCBA bonding-force actually contains many items and factors. But working-bear think we can refer to the conclusion of red-dye test to discuss it. The BGA red dye and pry test summary the crack as 5 kinds of defect positions. These locations also can be applied to the PCBA all soldering issues.

1. The bonding-force between Cu Pad (Copper foil) on the component and component substrate (FR4)

The copper foil bonded to the BGA substrate is represented. If there is another type of electronic components, then different substrate material, plating, bonding solution on the soldering foot will present different bonding-force.

2. The bonding-force between BGA ball and Cu pad on component

The IMC layer on the component site is represented.

The BGA package always uses the Nickel (Ni) base of the substrate, so it will come out the IMC of Ni3Sn4. Be noticed that components with different plating material on the soldering foot will result in different IMC composition with different bonding-force.

3. The bonding-force between BGA sphere/ball and Cu pad (Copper foil) on the PCB

The IMC layer grows during SMT is represented.

Different PCB surface finished may present different types of IMC and result in different bonding-force. The Cu base of the PCB will make the IMC of Cu6Sn5 and Ni base of the PCB will make the IMC of Ni3Sn4. Basically the bonding-force of Cu6Sn5 is better than Ni3Sn4. Unfortunately, the Cu6Sn5 will gradually transfer to poor bonding-force of IMC of Cu3Sn as yeas pass.

4. The bonding-force between Cu pad (Copper foil) on PCB and PCB laminate (FR4)

The copper foil bonded on the PCB substrate (FR4) layer is represented.

Recommending NSMD (Non-Solder-Mask-Defined) pad design to improve the copper foil pad on FR4 but NSMD is not good for the solderability on BGA ball. We can have this discussion later article.

Sometimes the humidity attack on the PCB is the key point to result in PCB de-lamination issue due to the moisture will expand quickly during the high temperature reflow process.

5. The bonding-force within solder ball

The void inside the solder ball is represented.

For the other components then the solder paste quality and voids is the key.

Now you know already how many kinds of the bonding-force was included in the PCBA. The next step is to see where is the weakest point then the crack will start from that weakest point while stress applies.

The experience shows there are more than 60% of solder crack show on the IMC layer locate on the PCB side. (Are you still remember what IMC means?). The 2nd one break between the Copper foil pads on PCB and FR4.  The 3rd one break between IMC layer between BGA ball and Cu pad on component side. Only few of them break within solder ball or solder.

When people do the analysis for the solder crack for the BGA or component and found there is slightly black color show on broken section people smelled blood and launched a ferocious attack on the SMT factory. Most of the people think this black color must be oxidization happened on the components or PCB soldering surface and it must be the root cause of solder crack. Does this true? We found the color of IMC layer is generally fainter than the flesh-solder since it been exposed at atmosphere quite a long time and most of the solder crack just happen at IMC layer. So, checking the color on the crack section may not give you a correct information.

You still have a question? It is not that the IMC layer growth between the solder alloy and Cu substrate is a necessary condition to form a good solder. Why The solder crack always occurs at the IMC layer?

Related article:
Why BGA soldering ball always crack(1)? Stress > bonding-force

Why BGA soldering ball always crack(3)? IMC layer growth is a certain result to form the soldering joints

2 responses to “Why BGA soldering ball always crack(2)? The composition of PCBA bonding-force”

  1. Stephen Lim says:

    I have a question concerning BGA ball crack. A phenomenon was observed, after a product was deployed in the field and few months later reported functional failure. Analysis led to BGA and eventually cross-section show that the corner tip BGA ball cracked. But it was cracked at the component FR4 side, not the product PCBA side of it. The customer assumed it was related to strain/strain rate but a thorough checked in our processes had proven all the data was below the IPC-9704 defined (least square )straight line. By design, there is a need to use a tool to push the locking pin to secure the heatsink at 2 corners. But the strain/strain rate was good. So, I can’t help but to think the BGA IMC layer might have a problem. Customer internally had another opinion was on the nickel thickness. My question is, what is the role play of the nickel, and reading your article, you suggest IMC should be between 1 to 5 um but I understood this is a by-experience numbers. Is there other factors I need to consider what could cause the BGA ball crack at the component side? (totally cracked away from component FR4).

    • WorkingBear says:

      Stephen Lim,
      If you read my article about the “Waht is IMC?” then you will find “The IMC layer can be grown and distributed evenly at the interface location will be good enough.“. So you need to check IMC layer of BGA grows evenly and not focus on the thickness of the IMC layer.
      The IMC will be Ni3Sn if your PCB is ENIG finished. That means the Nickel will be one element of IMC instead of the copper in OSP finished.
      Back to your BGA ball crack issue, as I said “the crack is because stress > bonding force”. I do recommend you check the strain/stress by gage at each stage may give the stress to the PCBA include testing, assembly process and drop testing.

Leave a Reply