We Bake, But Still Have Delamination, Why?

We Bake, But Still Have Delamination, Why?
When our PCBs are wave soldered and then pass through reflow most had issues with delamination. Where should we look to solve this?
Board Talk
Board Talk is presented by Phil Zarrow and Jim Hall of ITM Consulting.
Process Troubleshooting, Failure Analysis, Process Audits, Process Set-up
CEM Selection/Qualification, SMT Training/Seminars, Legal Disputes
Phil Zarrow
Phil Zarrow
With over 50 years experience in PCB assembly, Phil is one of the leading experts in SMT process failure analysis. He has vast experience in SMT equipment, materials and processes.
Jim Hall
Jim Hall
A Lean Six-Sigma Master Blackbelt, Jim has a wealth of knowledge in soldering, thermal technology, equipment and process basics. He is a pioneer in the science of reflow.


Welcome to Board Talk. This is Jim Hall and Phil Zarrow, the Assembly Brothers. What do we have today, Jim?

This comes from KLR. We baked our bare PCBs at a 100 degree C for six hours. And then dipped in our HASL machine at 250 degree C for 10 seconds with no delamination problems.

The same PCBs were then assembled and passed through our wave soldering system at minimum conveyor speed with no delamination.

Finally, the PCBs had SMT components placed and the PCBs passed through our reflow system at 217 to 230 degrees C. Most of these PCBs had issues with delamination. Where should we look to solve this puzzling problem?

I guess the first question we ponder here is the temperatures that were being used. I believe the IPC has a spec.

I might start, asking about this process sequence because they're talking about sequential processing HASL exposure, followed by wave soldering, followed by SMT assembly.

What were the timing between these steps? Was it possible that between wave soldering and reflow soldering the boards were exposed long enough to an ambient moisture condition that brought enough moisture back in. Remember, your soldering cycles don't bake out moisture.

Assuming that they're conscious of this problem. I was just glancing at IPC 1601. I noticed that all their bake cycles call out for 105 to 120 degrees C. These boards were baked at 100 C.

I'm guessing that maybe the baking was just a little too cool. And that you got some residual moisture deep in the core of the board such that these shorter cycles and HASL and wave soldering did not pump enough heat into the board to cause the problem. But the longer 30 to 60 seconds in a typical reflow profile may cause the problem.

Deep penetrating moisture might be the source of the problem. An easy thing would be to bump your bake temperature up 10 or 15 degrees. I'd recommend looking at IPC 1601. It's the first document that really addresses in numbers and in hard facts or hard recommendations the idea of PCB baking. This new IPC spec is a first in the industry for a universal specification and looks like KLM may be a little outside of it.

And then also with any multiple processing, is there another source of moisture within the process that's getting in there. Did you clean after wave soldering? Typically, a water cleaning process is too fast to put a lot of moisture back in. It is puzzling to think that you didn't have delamination with the HASL immersion, 250 C for 10 seconds. 

But again, you can't see moisture. It's sort of a mystery, out of sight.

Yeah, out of mind.  You've been out of your mind for years, you know, all of those flux fumes. So hopefully that answers Mr. KLR's question.

I hope he doesn't have the fundamental problem like soldering like my brother.

And I hope he doesn't solder like my brother and whatever you guys do, don't solder like my brother.


Was this really a genuine question? How would it be possible to mount SMT to an already assembled and wave soldered PCB? Then the reflow would destroy the conventional parts anyway. I think not understanding moisture ingress and outgassing is the least of their problems.
Robin Moore, TTi
As above: Weigh,bake and reapeat until weight does not drop. You really need a mg scale. We use one up to 250 g. We typically bake at 140C. The steam pressure there is about 3 bar, to be compared with 30 bar at 235 when boards delaminate! Below 150 should be safe. Bare boards, no components!
Sten Bjorsell, Shipco Circuits Ltd
Also lets not forget the way he stack the boards in the oven. If to many boards are in a stack then the middle boards may not release the moisture as well as the outer boards would.
Liz Cramer, Global PCB Technologies, USA
Things to consider when dry baking to minimize delamination.

1. Did you get all of the moisture out? How do you know that you did? Answer: Weigh a sample on an accurate gram scale. Bake the sample for a set period of time at a set temperature. Weigh the sample. Bake at the same parameters for another hour. Weigh the sample again. Bake at the same parameters for another hour. If you see a reduction in weight, bake for another hour. Weigh and bake until you do not see a weight reduction. When you reach the point where there are no weight changes measured within the scale tolerance range add up your time in hours. This shall be the minimum time to dry bake at the temperature that you baked at.

2. What is a good temperature to bake at? One thing to consider is the product and this is where UL comes in. Materials and product that are UL certified are evaluated at a maximum continuous operating temperature, aka MOT. If many of the components are rated at a MOT of 130 deg C and one small smd cap is rated at 120 deg C then 120 deg C is your upper bake limit. Above the MOT rating and materials theoretically start to decompose. If your bake oven has a tolerance of +/- 5 deg C from the set point then your bake temperature would be 115 +/- 5 deg C. Keep in mind that when you open the oven door or load a preheated oven with cold product the oven temp shall drop. The clock starts ticking when the oven reaches the set point temperature.

3. Surface oxidation and the effect on solderability. This is where a solder sample can be of use. The IPC has guidelines in the baking spec. However, proving that your dry baking process (oven, product placement, air flow, ventilation, etc...) can meet or exceed the limits and remain solderable is something that needs to be qualified. This is where the solder sample can play a role. If you receive a solder sample from the PCB fabricator use it to qualify your dry bake operation and assembly profile. Keep in mind that there is a difference to an oven profile when baking one part as opposed to a fully loaded oven. Long term observation and testing is part of the overall qualification process.

4. Mechanical strength of the laminate. Dicey cured epoxy is mechanically stronger with regards to bonding. Phenolic cure epoxy has a weaker mechanical bond strength. Thermal shock at higher temperatures through an assembly line from zone to zone may result in added mechanical stress that can result in a mechanical delamination. Materials expand when exposed to heat. Different materials expand at different rates. Different areas of the product increase in temperature at different rates. These factors add stress to the product through the assembly process. Modifying the assembly process to allow the product to pre-heat for a slightly longer time shall help minimize the shock and stress. This shall help reduce the occurrence of a mechanical delamination.
David Duross, The Bare Board Group, USA
I am a sort of newcomer for SMT, however I have some experience with component moisture absorption. I would suggest that besides increasing the bake temperature to bring it in line with the IPC spec you quote, the time between baking and SMT reflow be as short as possible, to remove the possibility that moisture may be absorbed again. If you have the appropriate equipment, you could maybe try to measure the weight gain in the PCBs between the 1st baking and the moment of reflow, just to be sure about the presence of moisture.
Francesco Basta, Cisco, Netherlands
The effectiveness of the 100 deg.C bake may be completely negated if the bake oven is not properly vented. It is important to maintain the greatest difference possible in water partial pressure during bake. This means that a sealed oven may not work at all, a vented oven will be much better, and a vacuum oven will be the best.
Tom Salzer, Hermetric, Inc., USA
I am one of the writers of IPC-1601. Let me just state that 6 hours @ 100 may not be sufficient to remove enough moisture to get to the target, which is less than 20% of the saturated moisture content. In other words, if you can get 80% of the saturated moisture content out of the PWB, it generally will not delaminate during reflow.

The boiling temperature of water is 100 deg. C., so you need to get slightly above that temperature to change the moisture to steam and gradually drive it out. The key word is gradually. Doing this too rapidly (such as during reflow) causes the moisture in the PWB to be driven out violently, and thus the z-axis expansion damage (delamination or via cracking and shearing). Now, many times the PWB pre-preg itself is the problem. If the PWB laminate is of poor quality, and there are epoxy skips inside the laminate between the layers or in the layers themselves, no amount of baking can remove these air pockets and after baking the moisture content can be right back at the saturation level within an hour or two at between 40-60% RH. Thus the amount of time after baking and prior to reflow is critical. If the PWBs are not going to be printed, populated, and reflowed within a couple of hours, then after baking either drypack the PWB or keep the PWB in a drybox such as DR Storage or McDry until ready to process.

If the CCA will require any selective soldering or hot air rework, the bake will need to be repeated. The ramp-up during reflow may also be too rapid. The ramp rate should be no more than 1 degree C per second. A typical leaded reflow profile is generally 7 or 8 minutes long from beginning to end, possibly much longer for lead-free paste on larger, more massive CCAs.
Richard Stadem, General Dynamics AIS, USA

Submit A Comment

Comments are reviewed prior to posting. You must include your full name to have your comments posted. We will not post your email address.

Your Name

Your Company
Your E-mail

Your Country
Your Comments