BGA soldering

Go To Last Post
16 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Has anyone worked in soldering desoldering BGA chips?I think that this is the most difficult thing in electronics,but now the BGA chips are present in most modern electronic appliances and due to non lead solder they are prone to solder cracking.Including all the work that has to be done for reballing the chip with new balls this seems like someone is trying to learn and play a musical instrument
as an expert in only few days.Also the regarding information is like a black box and in many cases information from different sources tend to confict about soldering temperature of the solder alloys and maximal chip soldering temperature.

Last Edited: Mon. Jan 13, 2014 - 02:07 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've used a hot plate with pretty good success to solder BGAs. I've also seen low pin count BGAs (9 pins) done with hot air. Really depends on what you need to do...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

on a pst job I had a colleague who did a 81 ball micro BGA with a paint stripper and a lot of flux.
but it is more of an art to be able to do it without problem.
I have done a couple of re-solder jobs using a hot air tool, but that was just take a board that is mall functioning on the BGA, put a lot of flux on it and heat it until you see the solder melt. had 90% success rate (10 boards done 9 worked, the last non working board seemed to have multiple issues in the end so was trashed)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

In the past 14 years I have worked on a number of projects with BGA-packaged ICs. Most of these were in the 300+ balls category. IMHO the BGA package is one of the dumbest ideas ever conceived by the IC manufacturers.

On the positive side, they pack a lot of connections into a very small PCB area.

But on the negative side:

A. They are completely uninspectable unless you have access to a very expensive XRAY machine. Even then the results are often inconclusive. (Compare this e.g. to a QFP package where all leads, pads and solder joints are visible.)

B. The finished assembly is very fragile and suspectable to failure caused by moderate PCB flexure. (The solder balls crack because the PCB flexes, but the BGA package doesn't.)

C. No reliable manual repair method.

D. Solderability is highly dependent on the BGA PCB footprint and its associated trace fan-out. These metallic heat paths must be "balanced" across the footprint so all of the balls are subjected to a proper thermal exposure during the hot air soldering process. Get this wrong and some balls will barely melt while others will boil. In my experience there are multiple credible, but conflicting, opinions on what constitutes a correct BGA footprint. Plus, there's really no practical inspection method to know you got it right because the end result is hidden by the package itself (See "A" above).

The only hope is that somebody will invent a transparent PCB material! In the meantime, I think these atrocities of electronic packaging will be with us for quite some time. Plan to suffer accordingly!

E. The introduction of lead-free solder (including the balls) has only aggrevated the above problems.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Chuck-Rowst wrote:

B. The finished assembly is very fragile and suspectable to failure caused by moderate PCB flexure. (The solder balls crack because the PCB flexes, but the BGA package doesn't.)

This is not entirely accurate. A BGA will beat a QFN of equivalent size every single time in the face of either rapid (vibration) or slow (temperature cycle) strain situations. The solder balls are much more flexible than the "leadless" QFN pins. For this reason, you can get away with a larger CTE differential (or lower substrate Tg) with BGA packages than with QFN packages, even in the critical cycle region of -20C to +20C. Of course, the idiocy that is RoHS, with its stiff silver solders, negates a large portion of the benefit of these "flexible" solder balls.

Take a look at IPC-SM-785.

That said, I would never want to work w/ a BGA manually (i.e., w/o proper pick and place machines, x-ray machines, and rework stations). Even with a proper rework station, you need to leave a pretty large surrounding area bare to be able to reliably rework the BGA.

Science is not consensus. Science is numbers.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The OP remarked about a confusing variety of soldering temperatures.

The temperature profile is usually spec'd by the package manufacturer. It is their package and they can spec it how ever they wish. The only real conflict is when there are several devices with competing specs on the same board. There simply is no "one size fits all". You simply learn to live with it or change to non-BGA packages.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Chuck-Rowst wrote:
B. The finished assembly is very fragile and suspectable to failure caused by moderate PCB flexure. (The solder balls crack because the PCB flexes, but the BGA package doesn't.)
Underfill revisited: How a decades-old technique enables smaller, more durable PCBs by Michael Yu and Syed Wasif Ali (Nexlogic Technologies) (embedded.com; January 27, 2011)
Reducing PCB flexing is a part of system design (packaging, bonding, mounting, potting).
Chuck-Rowst wrote:
C. No reliable manual repair method.
How to Solder & De-Solder & Preheat BGA's, SMD's, QFN's & More. (Zephyrtronics)
An engineer or technician definitely needs a manual rework process and tools because swapping a CPLD or FPGA is common.
Swapping RAM, flash, and MPUs is common for game consoles, mobile phones, and routers.
Appears to be a lot of interest in low-priced SMT rework tools for mobile phones.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hobbss wrote:
Of course, the idiocy that is RoHS, with its stiff silver solders, negates a large portion of the benefit of these "flexible" solder balls.
For industrial usage lead-free may be OK though for some automotive and aerospace usage it may not.
Lead-Free Consensus Update by Dr. Ron Lasky (Indium Corp.; September 13, 2011)
RoHS has some legs:
"However, we can all purchase lead-free, RoHS compliant products at a reasonable cost and reliability." - Dr. Ron Lasky
Has Lead-Free RoHS Compliant Assembly Reliability and Cost "Arrived?" by Dr. Ron Lasky (Indium Corp.; February 25, 2013)

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I see some irony between the title of your first article and the quote in my signature. That said, the article contains some interesting information regarding the durability of Ag based solder joints that I was not aware of.

However, the greatly increased possibility of whisker development with Pb-free solder is a deal breaker for me. You point out that RoHS may not be adequate for automotive or aerospace usage. I agree. However, having worked for an aerospace electronics company, I can guarantee that aircraft/automobiles you travel in absolutely have RoHS compliant components on them, even in safety critical systems. This situation is due to the fact that many component manufacturers ONLY make RoHS compliant parts, decreasing the availability of "traditional" components. Mixed designs are a problem, because even if PbSn solder is used, it typically does not bond as well w/ the pure tin leads. Furthermore, if the solder does not completely coat the lead (which is most often the case), whiskers can grow above the "solder line", resulting in catastrophic failures.

In addition, I have yet to see a single, repeatable study that conclusively shows that lead in discarded electronics results in a significant increase the lead levels in surrounding ground water. It is my belief that RoHS is a solution without a problem.

Science is not consensus. Science is numbers.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

meslomp wrote:
on a pst job I had a colleague who did a 81 ball micro BGA with a paint stripper and a lot of flux.
It's jaw dropping when one hears a technician state he successfully reworks BGAs with a heat gun.
An artist in action.
Could try a tacky flux for removal (keeps the flux somewhat in place instead of running all over).
BGA-specific flux for re-solder (less sputtering): BGA Flux (tm) (Zephyrtronics)

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gchapman wrote:
It's jaw dropping when one hears a technician state he successfully reworks BGAs with a heat gun.
An artist in action.

I have seen technicians perform magic w/ a soldering iron. It is odd -- at multiple companies, I have always found that the little old ladies are far and away the best solderers, whether or not they have obtained NASA certification. They are able to free-hand solder components w/ the naked eye, when I can barely see the leads under a magnifying glass.

Science is not consensus. Science is numbers.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For example the melting point of lead solder is 185C and for liquifying needed around 205-210C.Lead free is 217C and needed some higher temperature,225-230c or even higher.Some chip manufacturers recommend the soldering temperature not to exceed 220C for ROHS and for some other chip the resellers says 230-245 max 260C.Until now i have soldered not perfectly flat chips to the pcb level and in some other cases chips got blistered or in the top,or in the bottom side.And you can not even trust a ready made soldering profile since different size boards need different profiles.Setting the soldering temperature and watching the IC with my eyes i can not even understand when the chip has soldered correct.
And that is the critical question,how can someone understand when a bga has perfectly sit in its place and the soldering process has fully competed so no heat is needed any furthermore.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hobbss wrote:
In addition, I have yet to see a single, repeatable study that conclusively shows that lead in discarded electronics results in a significant increase the lead levels in surrounding ground water. It is my belief that RoHS is a solution without a problem.
China is a leader in RoHS.
"Workers at e-waste sites are usually migrants from extremely poor areas and are often children. They have little to no access to gloves or face masks and are often too desperate for work or uniformed to care about the health risks. Workers at e-waste sites are prone to skin rashes, cancer, weakening of the immune system, and respiratory, nerve, kidney, and brain damage (3). In China’s Guiyu region, workers have extremely high levels of toxic fire retardants in their bodies and over 80% of the children already have lead poisoning."
Electronic Waste Disposal by Sophie Vos (Nicholas School of the Environment, Duke University; November 15, 2012)
Fume hoods should also be used but these have some expense in purchase and maintenance.
There is lead chelation therapy.
Also may be some therapies by nutrition.
Better is complete responsibility and local recycling.
"Preventing liability is your goal and GEEP assures this. As an electronics recycler with a zero landfill goal and a no export policy, we ensure you will never face environmental liability or negative publicity as result of the way we process your equipment."
Why GEEP (Global Electric Electronic Processing)
Pardon the plug but GEEP is in a next-door industrial area.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

geoelec wrote:
Until now i have soldered not perfectly flat chips to the pcb level ...
Some BGA and other SMD package specifications will state a required PCB pad surface planarity or co-planarity.
For PCB finish, IIRC, ENIG is one way of several.
HASL may not be adequate.
Required metallurgy compatibility in BGA ball, solder paste, PCB finish.
geoelec wrote:
... and in some other cases chips got blistered or in the top,or in the bottom side.
Another concern is the unseen damage in the form of "purple plague".
The Science of Heat Damage: How Overheating Damages I.C.s (Electronics Manufacturing Sciences, Inc.)
geoelec wrote:
And that is the critical question,how can someone understand when a bga has perfectly sit in its place and the soldering process has fully competed so no heat is needed any furthermore.
An anecdotal observation: gently and slightly perturb the package at liquid solder temperature and confirm the package returns to in-place.
May want some package registration marks in the PCB's silk screen layer(s).

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

nleahcim wrote:
I've used a hot plate with pretty good success to solder BGAs.
Though he states "Just a feeling." the creator of Gadget Factory, Jack Gassett, also has some success with a similar method (pancake griddle, hot air soldering rework tool).
JaxHax - BGA Soldering at Home (Gadget Factory, YouTube; Jun 21, 2011)
A concern is getting burned.
And the flying parts due to one's autonomic response.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hobbss wrote:
I see some irony between the title of your first article and the quote in my signature.
(Pardon the O.T.) Would you try the following on for size?
http://youtu.be/ceFyF9px20Y?t=1h15s "Types of Proof"
Fat Chance: Fructose 2.0
by Robert Lustig, MD (UCSF School of Medicine)
(University of California Television (UCTV), YouTube, time 1:00:15 for about 3 minutes)
Seems both medical doctors and engineers have concerns, more likely worries, about ethics and risk analysis.

"Dare to be naïve." - Buckminster Fuller