Bga

[Loedown]

Greetings,
For all the people in the know, apart from space saving, why would anyone use BGA as a technology? In my limited experience with such, I've always found it to be a pain to manufacture, a pain to rework and generally unreliable at best. What does everyone else think?

[yaqube]

Originally Posted by Loedown

For all the people in the know, apart from space saving, why would anyone use BGA as a technology?

- Higher pin (or rather ball) count. You can find packages with over 1000 balls.
- Shorter leads (lower inductance) = higher speed.
- Lower thermal resistance (allows higher power dissipation of the IC).

[Hewitson]

Why is thermal resistance lower than other packages?

[Loedown]

Originally Posted by yaqube

- Higher pin (or rather ball) count. You can find packages with over 1000 balls.
- Shorter leads (lower inductance) = higher speed.
- Lower thermal resistance (allows higher power dissipation of the IC).

Intel 1156 & 1366 use a socket, still very fast, the inductance I can agree with.

[FOL]

I have come across, a few home electronics BGA's and all have been crap. Simply cause they were BGA and never had heatsink / fans. They would heat up so much that the solder balls would ment and cause massive problems.

Sony were (proberly still are) the main culprit.

[jbenam]

Originally Posted by FOL

I have come across, a few home electronics BGA's and all have been crap. Simply cause they were BGA and never had heatsink / fans. They would heat up so much that the solder balls would ment and cause massive problems.

Sony were (proberly still are) the main culprit.

Same here.

If you've heard about those Xbox360s and PS3s out there failing, that's the reason. BGA. The solder balls screw up, and the console dies

How dearly I miss ye, QFP (I know, I know, pin count...! ).

[whiteb]

Originally Posted by jbenam

Same here.

If you've heard about those Xbox360s and PS3s out there failing, that's the reason. BGA. The solder balls screw up, and the console dies

How dearly I miss ye, QFP (I know, I know, pin count...! ).

Now that makes sense, I have never had a 360 or PS3, but I have heard of them having a HIGH fail rate, now we know why. You would have thought Micro$oft or $ony had the intelligence to figure that out.

[FOL]

Originally Posted by whiteb

Now that makes sense, I have never had a 360 or PS3, but I have heard of them having a HIGH fail rate, now we know why. You would have thought Micro$oft or $ony had the intelligence to figure that out.

The problem is the airflow and heat removal. Something thats processing will always generate heat. These console designers must be so thick, as to design something in that fashion. All it would take is 1 heatsink and fan.

What they need is more people on the front line, who deal with these problems on a daily basis. They could say to the designers, "hang on there, that will cook and cause problems". Damn sure we would never see a console go faulty due to design flaws.

An old saying fits this best, "all for the sake of". In this case "All for the sake of a £1 heatsink and £1 fan".

[Zetr0]

If I am not mistaken, I remember reading up on the biggest design flaw of the current Gen of Consoles was more to do with the solder mix for RoHS compliancy.

The original design (XBox360) called for a mix of leaded solder, but with european and somewhat national legislations they had to change the solder mix to one without lead, or they wouldn'y be able to sell the product in that country.

As FOL describes these BGA chip solutions heat up very effectively - unfortuantely the solder mix was too britle when it expanded, as such it snaps, creates a fracture in the weld.


This mostly is microscopic fissure but with heat-expansion over time this increases untill there is literally no connection. A common method to repair these faults was to re-flow the solder sealing the fissure - if anyone has seen or heard of an XBox360 with a towel round it now you know why.

Personally I find it deeply disturbing that those chips get THAT hot, enough to reflow solder - thats about 260c - I am surprised that these chips even work after 10 minutes of that temperature!

[Paul_s]

From what I've seen the 360 is very low quality in terms of manufacture. I've seen better plastics on a hairdryer and don't get me started on the motherboard!

[FOL]

Originally Posted by Paul_s

From what I've seen the 360 is very low quality in terms of manufacture. I've seen better plastics on a hairdryer and don't get me started on the motherboard!

My brother reversed his fans and made a bluepeter style funnel and all his XBOX 360 problems went away.

[jbenam]

Originally Posted by Zetr0

If I am not mistaken, I remember reading up on the biggest design flaw of the current Gen of Consoles was more to do with the solder mix for RoHS compliancy.

The original design (XBox360) called for a mix of leaded solder, but with european and somewhat national legislations they had to change the solder mix to one without lead, or they wouldn'y be able to sell the product in that country.

As FOL describes these BGA chip solutions heat up very effectively - unfortuantely the solder mix was too britle when it expanded, as such it snaps, creates a fracture in the weld.


This mostly is microscopic fissure but with heat-expansion over time this increases untill there is literally no connection. A common method to repair these faults was to re-flow the solder sealing the fissure - if anyone has seen or heard of an XBox360 with a towel round it now you know why.

Personally I find it deeply disturbing that those chips get THAT hot, enough to reflow solder - thats about 260c - I am surprised that these chips even work after 10 minutes of that temperature!

Right on spot, Mr. Z!

I reflow 360s every now and then using hot air

They work pretty well after that (Only to die again months/years later)

[Schoenfeld]

Originally Posted by Zetr0

The original design (XBox360) called for a mix of leaded solder, but with european and somewhat national legislations they had to change the solder mix to one without lead, or they wouldn'y be able to sell the product in that country.

So it's not a design flaw, but a manufacturing flaw. The design stays the same for RoHS compliant manufacturing, but you cannot just exchange the solder and run the same process. The expansion of board and chips must be adapted to the updated temperature curve in the reflow process, so it mainly comes down to the PCB material and the pre-pregs that hold the multi-layer cards together.

Originally Posted by Zetr0

As FOL describes these BGA chip solutions heat up very effectively - unfortuantely the solder mix was too britle when it expanded, as such it snaps, creates a fracture in the weld.

...and this doesn't happen at the customer, but it happens right when the board comes out of the reflow oven. However, the micro-cracks are not seen until the customer has used the product for a few weeks. You'd get around this kind of early-life-failure by stress-testing all units in a climate chamber. For devices that are to be used 24/7, this is common use: The devices come pre-aged and therefore have an extremely low failure rate in the field (and this has nothing to do with RoHS, but with common sense).

Originally Posted by Zetr0

Personally I find it deeply disturbing that those chips get THAT hot, enough to reflow solder - thats about 260c - I am surprised that these chips even work after 10 minutes of that temperature!

That's not correct - the chips don't heat up that much - they barely cross the 100 degrees barrier in normal operation. BGA has the big advantage that the solderballs have a low thermal resistance due to their mass (compared to QFP pins). The solderballs make large-field-contact with the copper, which in turn has a good contact to the prepreg, which acts as a heatsink. All it takes for the PCB to become a good heatsink is enough VIAs into the inner and bottom layers, and some copper area to distribute the heat.

The 100 degrees are not a problem in operation - they are a mechanical problem, if the PCB and the chip don't expand at the same rate. If they don't, a certain amount of flexibility is required for the link between the two components, and leaded solder can do that job. However, it's always better to adjust the WHOLE production process to RoHS requirements - that's why Individual Computers and E3B have already started researching new RoHS production processes in 2004, almost two years before it became the law.

If Sony has not understood all the implications of RoHS, it's a pity. Short-sighted managers obviously tried to save money. If they would only spend some time thinking about the *chances* that RoHS is giving, they would have found that RoHS does not reduce, but actually INCREASES production quality, if it's done right. With all the new requirements, you need to look at every single spot that may be affected, and bring it all together. You re-adjust the whole production process, you define new ways of glueing the copper to the pre-preg (here's where the flexibility is - well, if it's done right), and all of a sudden you realize that RoHS even SAVES MONEY, because with the higher quality, you hardly have any returns.

Please think twice next time you blame things on RoHS or BGA. Both are excellent technologies, they are a true advancement over anything we've had in electronics production before. Only if the process is messed up, you get in trouble, but that's also true for any other industrial process.

Jens

[Zetr0]

@Schoenfeld

Thanks for the informations =)

Kinda fills in the gaps from what I have read and experienced.

[yaqube]

Originally Posted by Schoenfeld

Please think twice next time you blame things on RoHS or BGA. Both are excellent technologies, they are a true advancement over anything we've had in electronics production before.

If the RoHS is so excellent why isn't it used in the military, aerospace, medical, automotive, industrial, telecom infrastructure, test & measurement equipment?

[Schoenfeld]

Because those industries are powerful enough to push through the "never change a running system"-approach. Imagine a supplier of military equipment would have to admit that they could have done better in the past, what would that look like?

Also, medical and aerospace equipment is going through extremely extensive testing and cross-analysis by insurance companies. Their cost in migrating to a whole new production process would be extremely high. There's extreme amounts of money to be made in those industries, and it's kind of a "closed club" of companies who deliver to that market. What a nice coincidence that they all agree that non-RoHS is better! Don't invest, but harvest good gains.

You know what those big companies and especially insurances are like: The easiest way to go is to say "it's not guaranteed". Once you have that answer, it trickles down the offices, a little twist here and a little twist there, and all of a sudden, an uncertainty (which is essentially is) becomes a danger, and politicians trust their "experts" that exceptions to the rules are required.

It's like FOL wrote: You need more people from the front lines reporting to the decision makers. I'm currently developing a VoIP router that's supposed to work 24/7 - no crashes, no flaws, emergency calls must be placed any time and absolutely reliable. Requirements are no less than military - and still we use BGA and a RoHS compliant production process.

Just taking one or two examples from the mobile gaming industry and use that as your reason to call RoHS and BGA "crap" is kinda... silly. Look at all the other stuff that's being manufactured since summer 2006: Mobile phones/smart phones that are at least equally punished like a PSP, and still they last for years. The bad examples here are single events where a company messed up. And yes, it's a lot easier to blame it on "too strict rules" rather than admitting that they were trying to save money on proper research.

Read between the lines, follow the stream of money - that's how you find true answers.

Jens

[Zetr0]

I dont think I blamed RoHS or BGA, the latter being a great technology, but as will all techy it needs to be used correctly, and for the most part I would say it is.

Like DDR3/4 and 5 Memory =)

To be honest, there was a huge cluster muck with both the initial productions of the X360 and the PS3 (with PS2 compatability) abliet design and or manufacture, I wouldn't know which so I can blame them both equally =)

However my personal View on RoHS compliancy is perhaps more skewed than most.

Personally I believe that RoHS compliancy is about as much as restricting a market as it is in reducing heavy metal waste.

Thus for the most part enforcing developers to use European manufactures with larger production cost compared to much cheaper Asian manufactures - to sell within europe.

I dont dissagree there needs to be some concern about our refined heavy metal waste, its most certainly a global concern, but I wont sing and dance as to the RoHS tune as pretty much limited to only consumer electronics - i.e. a clever way to dress up market divider/provider.

Fortunately Asia is catching/caught up =)

Just my humble opinion of course

[Schoenfeld]

Originally Posted by Zetr0

Personally I believe that RoHS compliancy is about as much as restricting a market as it is in reducing heavy metal waste.

And what's that opinion based on? Look at the materials that are forbidden: It's only six of them, and each of them is not just "considered" to be poisoness, it's common knowledge. Lead is just one. Chrome-6 is *known* to damage DNA (i.e. harm your offspring). Certain flame retardents are *known* to cause cancer. I think that RoHS is an extremely cut-down list of substances that should be expanded. I welcome this political start, even if it's way under-regulated in my opinion.

Originally Posted by Zetr0

Thus for the most part enforcing developers to use European manufactures with larger production cost compared to much cheaper Asian manufactures - to sell within europe.

Actually, Asian manufacturers were pretty quick to catch up. When I started the cooperation with Castlenet in 2007, they were already 100% RoHS with their own RoHS testing equipment - and that was during a time when the German customs had bought hand-testing equipment that didn't work properly, and many manufacturers were trying to slip away with their old stock and existing production processes.

It took me a while to come to this opinion. I hated RoHS back in 2006, because I had to throw away so many components. When that company picked it up, I was about to quit the market. I've only seen that a political regulation had cost me a five-digit amount of money. Only after analyzing my figures of 2006 and 2007, I realized that RoHS had brought my business up to a whole new level of quality management. My opinion made a 180-degree turn, because I have the facts right in my books.

So I ask one more time, what's your opinion based on?

Jens

[Zetr0]

@Jens

I didn't mean to upset you, my opinion is based only on my experience with RoHS compliancy within the UK =)

Please understand that I really appreciate your thoughts and sharing your experiences here, you have an unique insight into managing your business from idea -> design -> developement -> product. The work that you do is nothing short of inspirational.

Alas from my point of view I have not had that experience with RoHS compliancy, I have seen companies and factories close and people (friends and family) lose jobs - although to be fair most of these companies were on shakey ground to begin with.

To be honest my personal opinion is that my countries interpretation of RoHS when it became a european standard was a complete joke - as such things like the above happened.

The area I live in is quite rural although lots and lots of manufacturing from large plants to micro elctronics, alas the mentality is still the same here, very entrenched conservative view, this as anyone knows is a recipe for disaster when it comes to change.

I completely agree that the RoHS is a GOOD THING, however as you say that list needs to include a lot more and it shouldn't just be limmited to heavy metals, but all carcinogens involved in manufacturing as well as the whats in the product.

Alas as you have experienced - a lot of components / toolings had to be replaced and the old being responsiblly (expensively) recycled, for the most part business had a fair warning, alas I beleive a lot of the smaller manufacturing plants just couldn't afford the finacial shift at the time.

I also hope that that RoHS gets the push it needs to encompas more than just the consumer electronics sector, I cannot agree more with you on RoHS and its need for compliancey within all sectors that use such carcinogenic materials in their products although as you probably know the military are a rule unto themself.

[Schoenfeld]

Did I sound upset? Didn't mean to. There are other things to get upset about ;-)

Jens