Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


BGA Picture Evaluation

Kevin Facinelli

#24054

BGA Picture Evaluation | 8 April, 2003

I am tuning a BGA process for a board assembly and would like to know if someone with a depth of experience can look at some Ersa Scope picture I can email out.

Thanks in advance,

Keivn

reply »

#24058

BGA Picture Evaluation | 8 April, 2003

We'd be pleased to look at your pix and offer an opinion. Email them to us.

reply »

Kevin Facinelli

#24067

BGA Picture Evaluation | 9 April, 2003

http://picturecenter.kodak.com/share?invite=ZExr4tzrmmYeI5Uozhlo

Here is a link to the pictures.

This is a HASL board with ws-609 paste. Lead parts look good but would like to see a shinner joint in the BGA area. Thermal profile looks good 150-155 flux activation 210-218 45-60 second reflow

Any comments aprreciated.

Thanks

reply »

#24068

BGA Picture Evaluation | 9 April, 2003

Nice pix. The pebbly texture that you're seeing is aptly called "orange peel".

Temperature appears to significantly influence the appearance of solder spheres during soldering. Indeed, when the reflow temperature is high, tin continues smelting longer, so solder spheres are further oxidized. This oxidation causes the �orange peel effect. We recommend optimizing your thermal recipe to decrease the time over liquidous plus 25*C.

Given that your recipe seems reasonable, there is another possible explanation. We have seen orange peel solder balls on BGA at incoming. We believe that the orange peel on these balls just get worse with solder attach. It's like the oxide has been imprinted on the balls and the flux that we're using is not BIG enough to strip the oxide.

We'd also complain that the collapse on the lower portion of the ball is excessive. This could be related to: * Reflow recipe is too long over liquidous plus 25*C. * Design of the BGA.

reply »

Kevin Facinelli

#24069

BGA Picture Evaluation | 9 April, 2003

Dave,

So if liquidus is 184 are you saying I should be at 209 for how long???

Kevin

reply »

#24072

BGA Picture Evaluation | 9 April, 2003

In order to get reliable reflow, you need to be at or higher than [liquidus + 20*C] for about 5 to 10 seconds. [I said 25* in a earlier post, but in an effort to stay on message, I'll stick with 20*]

Don't forget that your liquidous is probably NOT 184*C. That's probably the liquidous of your WS609 [ummm, good stuff]. The liquidous you need to to think about is the liquidous of the alloy of the solder connection. That, in your case, includes materials from: * WS609 * HASL from your bare board * BGA solder ball * Solder used by the BGA fabricator * Solderability protection soldering by-products [eg, gold, nickel, etc from the ENIG] used on the BGA interposer.

reply »

Kevin Facinelli

#24077

BGA Picture Evaluation | 9 April, 2003

Dave is there an easy way to calculate liquidous temp? or just a ball park for Hasl boards / BGA's. I placed a thermal couple with kapton tape over a BGA to get the approx profile.

Thanks,

Kevin

reply »

#24078

BGA Picture Evaluation | 9 April, 2003

Kevin: We know of no handy approach to developing recipes. * Certainly programs provided by profiler suppliers give you a good first pass, based on your oven and solder paste. * Second pass in developing your recipe comes through trail and error changes to refine the initial pass recipe by improving soldering of components and pads with unusual solderability protection or heat dissipation. * Probably the more structured you are in your development process and record keeping, the more you can rely on and find that information useful and available.

We are not real excited about attaching thermocouples with Kapton tape. Roger Saunders, President, Saunders Technology, Inc. says, "High temperature adhesive tape such as Kapton tape is appealingly easy to use. However, care must be taken to ensure that the thermocouple junction is preloaded to maintain firm contact with the surface to be monitored. As there is no surrounding conductive material, (like solder), if the junction is even a thousandth of an inch off the surface, it will be reading primarily ambient air temperature, though somewhat influenced by thermal radiation. One effective installation method is to bend the thermocouple wires into the shape of a small hook. Then tape the wire down just in back of the hook, so the junction is preloaded onto the surface. PRO: * In most situations tape is quick and easy to use. Thermocouples may be taped to any type of surface. CON: * Adhesive grip of tape weakens with increasing temperature. Consequently, at reflow temperature it relaxes the preload on the thermocouple. This can allow the thermocouple to lose contact with the surface being monitored, and read ambient temperature instead of surface temperature. * It can be difficult or impossible to tape a thermocouple down reliably in tight places, such as between components."

KIC Thermal Profiling gave a paper at Nepcon West 1999 that compared various thermocouple attach methods and found that Kapton tape was the least reliable.

We have no relationship nor receive benefit from the companies referenced in this post.

reply »

MA/NY DDave

#24079

BGA Picture Evaluation | 9 April, 2003

Hi,

"Dave is there an easy way to calculate liquidous temp?"

Just one comment for now.

The supplier should know this temp or it's range. They do it with the bulk material.

YiEng, MA/NY DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

MA/NY DDave

#24080

BGA Picture Evaluation | 9 April, 2003

Hi

I know what DaveF is saying about the Kapton situation, yet from what I remember also at a KIC presentation the Kapton technique wasn't given as bad a billing as Dave describes.

You do have to make sure you get a good firm contact and verify it after all readings. Personally I like adhesives since I have seen the lifting effect, yet Kapton will do if one is careful.

YiEng, MA/NY DDave

P.S. By the Way, You probably have already read the following:

http://www.alphametals.com/products/solderpaste/PDF/WS609.PDF

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

#24108

BGA Picture Evaluation | 10 April, 2003

Kevin, One point that I'd like to make is that in order to get an accurate reading of the solder joint temp's you need to have the thermocouple right at the ball/PCB interface. I'd recommend precision drilling a mechanical sample of the unpopulated PCB (#76 drill bit), from the top side, printing paste, then placing a mechanical sample of the BGA, then reflowing it. You then mount the drill bit in a pin vise, set the depth so that it can only drill 25-50% into the ball, and very carefully ream-out the hole (this needs to be done as some solder will "bleed" into the hole during reflow). One then feeds the TC into the hole ensuring it slides down to the interface. This is the only way that you know exactly what the paste performance is at the ball. Somewhat time consuming, but extremely accurate I'd also recommend TC'ing the smallest device on the PCB as well.

cheers,

Tony

reply »

MA/NY DDave

#24116

BGA Picture Evaluation | 11 April, 2003

Hi

Wow, Quite elaborate. My Mechancial Engineering, Thermo friends would be proud of you. I am not kidding.

The only problem I have with your technique is that quite often internal lands and structures steal the heat during soldering. By your technique you seem to me to have removed some of these internal PCB structures.

Using your same technique I might go on up beside and come in contact with the host land. The TC should also be stealing some heat so I would be reluctant to penetrate the actual solder joint.

Hey CORRECT Me, if you think otherwise

YiEng, MA/NY DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

Daan

#24118

BGA Picture Evaluation | 11 April, 2003

I'd say your profile is about right, assuming these are values that you measured in the balls of this BGA and not elsewhere on the board. If you want a more shiny joint than use a nitrogen oven or experiment with various solder pastes, but why the heck would you want the balls to be shiny anyway ?? Due to pad geometry, board variations, component variations, and oxidation you'll never get 100% perfect looking joints, and it's not worth striving towards it.

Just an opinion....

Daan http://www.smtinfo.net

reply »

PCB Soldering Tools

HeatShield Gel- thermal PCB shield during reflow