We mistakenly assembled circuit boards using BGA components that were not properly stored. Can we now rebake the assemblies? Jim Hall and Phil Zarrow, The Assembly Brothers, discuss this question and share their own experiences and expertise. 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
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
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.
Transcript
Phil
And welcome to Board Talk with Jim Hall and Phil Zarrow, of ITM Consulting, although we're affectionately known as The Assembly Brothers in this venue.
Today we're coming to you from the ITM Elegante Ball Room high atop Mt. Rialto and we're here to talk about electronic assembly.
Materials, equipment, components, practices, procedures, and whatever else comes to mind.
I believe today Jim, we have a component procedure related question.
Jim
Yes this is from M.M. We mistakenly assembled circuit boards using BGA components that were not properly stored.
The BGA components had moisture tags that show indication of moisture exposure. All the assemblies passed testing.
Can we now rebake the assemblies to ensure that they will not have moisture related failures on these devices, or should we remove and replace the suspect BGA components?
Phil
I think this is a case of closing the barn door after the horse is out, or the BGA is out. Am I correct?
Jim
Absolutely Phil. This is a classic example of MSD potential damage.
A part was not stored properly. It was exposed to moisture. We didn't realize it. We reflowed it. The answer to your first question is no.
Rebaking at this point will do nothing to help that part. The damage occurs during reflow if there's too much moisture in there it can crack, delaminate and all the other bad things we know that can happen to the internal structure of moisture sensitive components. So no.
The question is, you say all the assemblies passed testing so there's no dead in the water failures. Was there any damage inside the parts? The best way to look for the most common types of moisture sensitive damage, which is cracking and delamination, is using scanning acoustic microscopy.
Better known as C-SAM. It's a non-destructive test. If you have it available, it's not real expensive or time consuming and again you can find it if you do the parts that are suspect you can see a lot of the typical kinds of damage.
You may want to x-ray them to look for potential cracks in wire bonds that may not have been caught during your quality testing for be it in-circuit or functional.
But, otherwise yeah, pull off all the parts and replace them. Good luck and you've learned, as my brother said, you've got to follow these procedures.
Phil
I just want to add that a lot of people feel that if they didn't get corner lifting on the corner bonds of the component that they're free and clear. We want to point out that the popcorn effect doesn't always occur.
That sometimes you don't get enough belly bulge during the reflow expansion to actually lift the corner leads. So just because you haven't lifted the corner leads doesn't mean those symptoms my brother Jim has mentioned aren't there.
Jim
Yes, the scariest thing about moisture sensitivity damage is that we have no idea how big the problem is.
Because if as Phil just illuminated, if you don't get this catastrophic pop-corning during reflow that lifts the corner, that cracks the packages, or we've seen actually parts being blown right off the surface of the circuit board, you can get situations where they look okay, they test okay but you've got some latent damage inside that has reduced the reliability of the product and you get failure in start-up, burn-in, early infant mortality, or worst of all intermittent.
Because you've got some delamination that's not a failure but the part overheats or you get some connections that aren't reliable and you get intermittence.
Typically if they've gotten out of the factory very few people die and fault analyze to the internal component level. So think about all the intermittent problems that are out in the field. Some percentage of those are probably due to moisture sensitivity damage during reflow in assembly.
But we have no idea how much or how big the problem is. So keep those moisture sensitive parts in their moisture barrier bags or in their dry cabinets. Keep track of them and don't put them on a board if you're not sure.
Bake the parts before you reflow them. Baking them after reflow does no good whatsoever.
Phil
You have been listening to Board Talk and whatever you do ...
Jim
Don't solder like my brother.
Phil
Don't solder like my brother either.
Comments
Would it be worth doing V/I testing on the BGA part to detect potential failures? ABI's BoardMaster does "3-D V/I Testing" which compares each pin with a "gold standard part" and many frequencies, detecting even subtle faults.
Alan Lowne, Saelig Co. Inc.
In addition to my previous comment, I would like to add two more things. The first is that MSD controls, like ESD controls, should ALL be present and working. This means parts are baked properly according to J-STD-033, boards are pre-baked per IPC-1601A, the humidity in the production area up to and including reflow and rework is kept between 35% to 55% RH, the parts are kept in proper sealed MSD packaging with desiccant right up to pick and place and reflow and not allowed to sit out for hours between side one and side two component placement/reflow, there are dryboxes present capable of meeting 5% RH just in case you cannot populate opened parts or process CCAs through reflow within 4 hours or less, etc.
Then, if any of the spokes of the MSD wheel are found to be broken, like in this case, but the functional test results are all fine, no evidence of popcorning (measure the thickness of one or two removed BGAs and compare with known-good parts), you have justification for using the product. Secondly, you can also take two or three of the affected assemblies (just as they are) and run them through whatever vibration and/or thermal cycling or ESS testing that makes sense, depending on the product. If they pass functional test and visual inspection after that, the rest of the suspect product is exonerated by implication.
Richard Stadem, General Dynamics Mission Systems
The advice to pull the BGA parts off the board, is in my opinion, not good. When you remove the parts you introduce various levels of degradation and damage to the PCB and then yet another level of said degradation and uncertainties when you solder the new part to the board. Unless the environment is extremely humid, >>80% RH, then it is likely the parts were not damaged.
The encapsulant used for the bodies today are quite superior to those used 25 years ago when moisture damage was a fair expectation and bake-out was the best step to take to minimize damage. On the other hand, if the assemblies are any portion of life support, scrape the assemblies and start over.
Jaye Waas, Renkus-Heinz
Just because a component is a BGA does not automatically mean it is a Moisture Sensitive component. If the parts were soldered into place without any incidents and passed X-ray inspection and electrical test, they are most likely OK. Attempting to remove them and replace them would only add further risk of damage leading to scrap.
Richard Stadem, General Dynamics Mission Systems
Better way to check all component condition before use on PCBA.If its affected definetly by moisture its must to bake before use on board.While using there must be CDC(Check do check) monitor the batch wise production.There is scope for check & finetune reflow profile for this special batch of component/PCB & after getting better result can go for mass production use.Even after complete product must to use X-Ray on sample basis or 100% to get confident on FG product to avoid customer & field failuare.
I will suggest to to identify the defective suspected component or PCBA condition before use on line.After confirmed its got affected go for baking before use in process.Even after baking use the parts on line on controll manner with small small qty by doing CDC(check do check) to avoid the full batch production defective.Also can get the scope for finetune reflow profilw for affected parts.After confirmed good quality product is coming out from line can go for mass production.Monitoring must be followed with X-Ray inspection to avoid pass the defect & can also change the process parameters as per inspection result.Also need to do X-Ray inspection on sample basis or 100% after completed assy to avoid customer end or feild failuare.
Sandip Raut, Jabil circuit india Pvt ltd
The "Assembly Brothers" are absolutely correct. Any possible damage was done in the reflow solder process. Baking the Assembly will accomplish nothing, possibly further damaging the assembly or components. And the damage may be in shortened life of the assembly, if not an immediate failure. If the components were not properly stored, or the HIC card indicates over exposure, or it is simply not known, the components MUST be baked per table 4-1 of J-STD-033D BEFORE it is reflow soldered.
As for proper storage: the components are typically safe in the original moisture barrier bag (MBB) sealed from the manufacture for one year. After that they can be resealed in a new MBB, with new desiccant, annually or stored in a dry storage cabinet at 5% or less indefinitely. As for maximum exposure time (floor life) of the component before solder, this depends on the %RH and Temperature of your factory, the component package thickness and the MSL level of the component per table 7-1 of J-STD-033D.
Paul Austen
Well, the truth is "they're probably fine." (Every quality or process engineer reading this just grimaced.) But you can't know for sure, so the conscientious thing to do is remove and scrap those BGAs. Or if they're valuable enough to warrant saving, I would send every last one of them through C-SAM and X-ray inspection before having them reballed. (Just don't forget to bake the boards before the rework, of course.) A spent humidity indicator card (HIC) does not necessarily mean the parts have soaked up fatal amounts of moisture, only that some (or all!) of the floor life has expired.
Since you can't know how much floor life clock remains (if any!), you have to assume the worst and follow the prescribed baking protocols to reset the floor life....and then be sure to properly store them (in dry pack or nitrogen) until use. The JEDEC guidelines for MSL (J-STD-033D) are available as a free download from their website: https://www.jedec.org/standards-documents/docs/j-std-033c We keep the Table 4.1 posted on our baking ovens. And as part of your Corrective Action, we suggest posting this Monty Pythonesque sign in your shop, "We apologize for the fault in our MSL Protocol. Those responsible have been sacked."
Alan Couchman, Process Sciences, Inc.
Moisture sensitive devices are a class of their own. The manufacturer stipulated instructions on storage and handling moisture devices have to be meticulously followed. If the moisture ingress is very less and is at a benign site in the package these can escape on baking without affecting the device function and properties. However, if a large quantity of moisture has crept in, it can lead to de-laminations at the interface of die, the lead frame, bond wires, heat sink paddle, the in-built PCB of the BGA with the moulding compound.
Although, the said BGA devices have already undergone Reflow and is reported functioning properly does not imply that the device will work forever. The damages induced can pop up at a later stage.
Hence, I strongly recommend to remove the BGAs after baking the PCB and go for replacing the BGAs which are stored and handled as per manufacturer instructions. Prior to replacing the BGAs, one has to bake the PCB. If the devices have crossed the floor life prescribed by the manufacturer, the devices shall be subjected to baking. Bake out regimes stipulated by IPC-TM-650 shall be followed.
Dr. KUTTIYIL THOMAS OOMMEN THARAKAN, VSSC
I would agree last option R&R, avoid a BGA pad repair. A nice long reflow soak should drive off moisture. I would still consider baking the CCA if it was to be wave soldered (preheat much shorter, shocker).
Rob, Amherst
BGA's can be usually checked using even 2D x-ray (3D oblique and CT x-ray are superior methods, however. You are looking for "voids" if the number of voids is deemed acceptable, the BGA's are likely fine.
Bill Peterson, Hirose Electric
I can only add that I would discuss it with the customer (whether it is your external or internal one). Tell them what happened and try to come up with a reasonable solution (decision) together.
