BGA packages in high vibration environments

Hi

Anyone have any experience of BGA's (especially fine pitch types) in high vibration environments? Is there a more appropriate newsgroup for this topic?

TIA

Geoff

Reply to
Geoffrey Mortimer
Loading thread data ...

I'm not having experience. The pcb must not flex due to vibrations. Given the high number of balls in relation to the tiny mass of the chip above, I'd worry about other more massive components first. How many g's at what frequencies were you thinking at ?

Rene

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net
Reply to
Rene Tschaggelar

"Geoffrey Mortimer" wrote in news:bmjk16$ngmof$1@ID-

163942.news.uni-berlin.de:

A fellow that designs testing equipment to be attached on rotating machine elements told me that he cannot use BGA. They tend to fall off, so he prefers SQFP. He thinks that the long pins of LQFPs reduce the forces on the solder by their flexibility.

cheers Gunther

Reply to
Gunther Mannigel

Couple of choices- try to get CGA instead of BGA- the additional solder of the column might save the day while not increasing the footprint, or get a BGA-PGA (or appropriate form factor) adaptor and make sure tha the adaptor card has exactly the same coeffiecient of expansion (or as close as is reasonably possible) to your BGA to ad least minimize the thermal stresses that accelerate stress fractures.

--
Skipper Smith                         Helpful Knowledge Consulting
Worldwide         Microprocessor       Architecture       Training
 Click to see the full signature
Reply to
Skipper Smith

It strongly depends from your vibration requirements.

I currently use electronic boards installed into railway equipments which use BGA components, and they are conforme to the European Union railway equipment manifacturing specification

Regards

/Alessandro

Reply to
Alessandro Strazzero

I wonder why a simple mechanical fixture wouldn't be enough to solve that problem. I.e. if you want to prevent a thing from falling off, why not just nail/screw/press it into position? If all else fails, you can still flood the PCB casing with epoxy, turning it into a virtually indestructable blcok.

As to that rotating gear, that should be even simpler to fix: make sure the BGA chip is mounted on the *inner* side of the PCB, seen from the rotation axis, so the centrifugal pseudo-forces press it into place instead of trying to rip it off.

--
Hans-Bernhard Broeker (broeker@physik.rwth-aachen.de)
Even if all the snow were burnt, ashes would remain.
Reply to
Hans-Bernhard Broeker

The BGA is flat, not curved to match the orbit it's following. So the force is unevenly applied along the length of the chip, meaning that it's constantly being flexed (to try and conform to its orbit, basically) when the device is in motion. It relaxes as the device slows down and tenses as the device speeds up. It seems easy to understand why the joints are failing, especially when you consider that the BGA is laid down on a substrate of relatively flexible fiberglass PCB (to match thermal expansion coefficient of the PCB on which it lives) and the solder balls are quite fragile.

The QFPs are by comparison rigid bodies that don't change size/shape as much when the same uneven force is applied. And any change in shape is more readily absorbed by the springy wires that tie it down to the PCB. Think of the leverage exerted on the QFP's solder fillet by the package changing shape; the force is concentrated in a tiny area of flexible wire which fights a relatively large, solid bond area. Compare to the force exerted on a solder ball being subjected to shear. The wire can bend; the solder ball won't do so as readily.

As an experiment, take any PCB containing a QFP and a BGA, and clamp it down firmly. Grasp the chip package and apply enough force to move it 0.5mm along any vector parallel to the board. The QFP will survive, albeit with bent leads. The BGA will take a lot more force to move, but once you do move it, it will fail.

Potting would probably help, true.

Reply to
Lewin A.R.W. Edwards

Hi Alessandro, are there any specifications on the vibration ? Frequency, amplitude, pattern ? Perhaps a defined testing procedure ?

Rene

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net
Reply to
Rene Tschaggelar

This thread might do well in comp.arch.fpga. There are several FAEs there who love digging into just this sort of issue.

I have crossposted there.

--
Rick "rickman" Collins

rick.collins@XYarius.com
 Click to see the full signature
Reply to
rickman

Xpost 2 cae and caf, no Fup.

Hallo,

"Geoffrey Mortimer" wrote:

Actually that's a very hot topic as BGA seems to get usual in the world of FPGAs and ASICs. I know that our mechanical engineers allready research on this topic, as we are very likely to have some fine pitch BGA in a high vibration environment in future. I would guess, that you should ask in some mechanical newsgroups as well. A big problem using FBGA is the test, wether you connected all balls proberly [1], as you have no chance of easy visual inspection.

bye Thomas

[1] in a mechanical aspect. Of course you get a quick answer if one IO has no electrical connection.
--
Don't answer to the email in from, use thomas[at] for
PM.
Reply to
Thomas Stanka

I recently saw a product that allows visual inspection of the solder balls on a mounted BGA. It is a fiber optic microscope and has tiny fiber probes that can run between the balls. I'll look for the info if anyone is interested.

--
Rick "rickman" Collins

rick.collins@XYarius.com
 Click to see the full signature
Reply to
rickman

Followup to: By author: rickman In newsgroup: comp.arch.fpga

A lot of people seem to do X-ray inspection, which I guess could be considered "visual" in some way.

-hpa

--
 at work,  in private!
If you send me mail in HTML format I will assume it's spam.
 Click to see the full signature
Reply to
H. Peter Anvin

Rick,

I'd certainly be interested in more info on the fiber microscope you mentioned. Debugging designs with lots of big BGAs is tough enough without wondering whether it's an assembly issue or not, and traditional xray techniques are good for showing shorts, but no so good for opens ...

----- R> >

high

Reply to
Ron Huizen

There's also JTAG tools that can read and write arbitrary values to I/O pins. Roughly $1K for benchtop systems, $10K for a production tester. If the BGA is hooked to other chips with JTAG, you can make a rather complete test.

And of course there's traditional bed-of-nails, not used much due to cost of implementing on proto hardware.

Dave Kinsell

Reply to
David Kinsell

Sorry I did not get back to you sooner. The original contact was ASG at

formatting link
They make the IS-1000 which gets under the BGA from what I can see. So you can see each and every ball. But you should get a demo since the sales pictures don't clearly indicate if they are looking at the edge row of balls or an inner row.

With a google search I found this -

formatting link
They seem to make a similar product, but the web page is not too clear if they are just looking at it from the outside.

R>

--
Rick "rickman" Collins

rick.collins@XYarius.com
 Click to see the full signature
Reply to
rickman

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.