Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SMTnet
»
Electronics Forum
»
BGA rework using Tacky Flux

BGA rework using Tacky Flux

Subscribe to Thread

I am reworking some BGAs and just trying to find the easiest... - Jan 13, 2006 by

Flux only works fine. ... - Jan 13, 2006 by RDR

Just tacky flux then? ... - Jan 13, 2006 by

Also is tacky flux the same as gel flux? ... - Jan 13, 2006 by

As far as I have determined, the Gel and Tacky are the same ... - Jan 13, 2006 by RDR

Russ is right. Tacky alone works best. I've done dozens li... - Jan 13, 2006 by URL

Can anyone provide reliability data on this process (flux on... - Jan 13, 2006 by

Hi Ken, Good question, and I did search for what you are... - Jan 14, 2006 by

We are doing a lot of CSP-BGA rework with tacky flux from Se... - Jan 14, 2006 by Mike ZIG

I can't see why it would be any different, You are melting ... - Jan 16, 2006 by RDR

Ken Tacky flux reworked BGA will be perfectly reliable in... - Jan 16, 2006 by davef

Dave, Why does standoff height affect reliability? ... - Jan 16, 2006 by RDR

Dunno. 4% higher standoff results in 16% higher reliability... - Jan 16, 2006 by davef

Thanks for the links! ... - Jan 16, 2006 by RDR

Dave, I would also like to thank you for the links. As ... - Jan 17, 2006 by

What about doing this method on Micro Bgas? ... - Jan 17, 2006 by

Agreed. Some research sugests the "flexibility" of the join... - Jan 17, 2006 by

There are several life studies (see for example ... - Jan 19, 2006 by

Hi James, You can use the same flux for micro BGAs.The 529... - Jan 21, 2006 by Mike ZIG

james

#39047

BGA rework using Tacky Flux | 13 January, 2006

I am reworking some BGAs and just trying to find the easiest way to do it. Can you use tacky flux without solder on the pads to reflow? or do you have to have solder on the pads before placing the BGA to reflow? Any info is appreciated. thanks

reply »

RDR

#39048

BGA rework using Tacky Flux | 13 January, 2006

Flux only works fine.

reply »

james

#39050

BGA rework using Tacky Flux | 13 January, 2006

Just tacky flux then?

reply »

james

#39051

BGA rework using Tacky Flux | 13 January, 2006

Also is tacky flux the same as gel flux?

reply »

RDR

#39054

BGA rework using Tacky Flux | 13 January, 2006

As far as I have determined, the Gel and Tacky are the same animal.

Just flux works well, I did forget to mention that this flux only method will not work with a CCGA (Ceramic Column Grid Array) since these columns are not meant to be reflowed. So if your BGAs are either 63/37 or SAC 305/405 the flux only method will work.

I preffer this method because shorting is almost impossible unless the package warps or it is placed VERY poorly.

Russ

reply »

URL

#39059

BGA rework using Tacky Flux | 13 January, 2006

Russ is right. Tacky alone works best. I've done dozens like that.

reply »

KEN

#39068

BGA rework using Tacky Flux | 13 January, 2006

Can anyone provide reliability data on this process (flux only)?

reply »

DannyJ

#39073

BGA rework using Tacky Flux | 14 January, 2006

Hi Ken,

Good question, and I did search for what you are asking, but unable to find anything conclusive about using tacky flux vs. solder.

In reliability, it depends on which section of the process you are talking about. Mainly (and there are several papers to back this up...) is the solder added to the pads is mainly viewed to help with any warpage on the board or the part itself. As others, I have used tack flux to do BGA repair, but this was mainly on ceramic substrates where warpage was not that big an issue from the pcb aspect. If there is any warpage, tacky flux is not very forgiving, unless the balls collapse tremendously (not loooking forward to trying this with Pb-free...).

From a connection aspect, if you have good adhesion from the BGA to the pads themselves, I cannot find anything that disputes the use of tacky flux.

Not sure if that helps answer your question or not... Hopefully, someone has access to more information than I and will post.

Dan

reply »

Mike ZIG

#39074

BGA rework using Tacky Flux | 14 January, 2006

We are doing a lot of CSP-BGA rework with tacky flux from Senju Metal.Their flux is very good.We are buying directly from Japan 529D flux. We have reworked BGA processor plastic sockets of Fujitsu Laptops. What kind of flux you are using? Not all of them are good. What reflow equipment you are using?Are you rebaling? Hope that this will help. Mike Gipos

reply »

RDR

#39087

BGA rework using Tacky Flux | 16 January, 2006

I can't see why it would be any different, You are melting the balls into reflow to the pad. In my opinion, paste is only useful in this situation to keep the part on the board during placement along with some planarity concerns in some instances. When calculating post reflow solder volume of paste vs. ball, you will find that the paste is about 4% of the total solder joint volume. The only difference I see between reapplying paste and reflowing with flux only is the difference in ball size after reflow. I could be wrong but have used this method for 9 years now with no issues.

reply »

davef

#39091

BGA rework using Tacky Flux | 16 January, 2006

Ken

Tacky flux reworked BGA will be perfectly reliable in most customer use environments. We do not have data, but the tacky flux reworked BGA will be less reliable than a BGA reworked with paste [assuming all other factors are the same], because the standoff of the tacky flux reworked BGA will be lower than the other method.

reply »

RDR

#39096

BGA rework using Tacky Flux | 16 January, 2006

Dave, Why does standoff height affect reliability?

reply »

davef

#39100

BGA rework using Tacky Flux | 16 January, 2006

Dunno. 4% higher standoff results in 16% higher reliability [�The Effects of Solder Joint Voiding on Plastic Ball Grid Array Reliability� Banks, Donald R., et al.; Proceedings of the Technical Program - Surface Mount International; September 10-12, 1996].

Fluxing without paste does come with disadvantages [ intel.com/design/packtech/ch_14.pdf ]: * Ball reflowed with only flux will have a smaller solder volume than the initially attached ball. Therefore, the package will result in a slightly (usually 0.001-0.002 thou) lower standoff height from the PCB. * Thermal cycle reliability of BGA component could be slightly compromised because of the reduced standoff height, especially if solder balls are under the die area as in full array BGA types. * Self-centering ability decreases because of the smaller solder volume. * Formation of the fillet between the ball and the board can be hindered if the ball is misshapen or there is some minor imperfection with the pad area.

For more, look here: * http://www.amkor.com/products/notes_papers/Reliability_of_Lead-Free_Solder_Connections_Whitepaper.pdf

* http://www.ieeexplore.ieee.org/iel5/ 6144/29862/01362802.pdf?arnumber=1362802

* http://www.chipscalereview.com/archives/ES/issues/1198/technicala1.htm

* http://www.integrati.com/content/ersa/bga.pdf

* http://www.cooksonsemi.com/tech_art/pdfs/Solder%20Joint%20Similarities.pdf

* http://www.ieeexplore.ieee.org/iel5/ 9210/29211/01320303.pdf?arnumber=1320303

reply »

RDR

#39103

BGA rework using Tacky Flux | 16 January, 2006

Thanks for the links!

reply »

DannyJ

#39110

BGA rework using Tacky Flux | 17 January, 2006

Dave,

I would also like to thank you for the links. As well, I was not aware of this phenomenon. Excellent reads!

Dan

reply »

james

#39130

BGA rework using Tacky Flux | 17 January, 2006

What about doing this method on Micro Bgas?

reply »

KEN

#39148

BGA rework using Tacky Flux | 17 January, 2006

Agreed. Some research sugests the "flexibility" of the joint(s) is reduced with flux only resulting in premature failure as compared to standard paste + ball reflow processes.

However, this does not mean it is not a reliable method. Relibility requirements must be defined then measured agianst.

reply »

bwet

#39200

BGA rework using Tacky Flux | 19 January, 2006

There are several life studies (see for example http://www.solder.net/stencilquik/articles.asp) which point to higher reliability of reworked BGAs using solder paste vs flux.

The theory on the mechanisms which allow this to happen are as follows:

1. Greater solder volume (obviously up to a point) allows for a greater standoff distance between part and board allowing for greater air flow underneath part for better heat dissipation 2. Greater propensity to have an interconnect shape that is more conducive to withstand the shock and vibrations of real use (think of the increased reliability of solder columns). There is a paper out there describing this effect. 3. Solder paste helps compensate for any real world co-planarity problems of the board or part making sure the interconnections are made and first pass rework yield is higher

BWET

reply »

Mike ZIG

#39249

BGA rework using Tacky Flux | 21 January, 2006

Hi James, You can use the same flux for micro BGAs.The 529D we used for rework of high freq. CSPs and even w/out cleaning it still worked on 2.5GHz.We have not washed the flux for 6 months and the RF features of the circuits remain the same. No changes to be seen on HP8753D VNA. I do not have any benefits from SENJU Metal. Rgrds Mike

reply »