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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Plating for aluminum wire bonding

ramanandkini

#23591

Plating for aluminum wire bonding | 2 March, 2003

Recently, one of our new vendor supplied us PCBs with Electroplated Nickel and Gold. We have the other regular source giving us ENIG plating. In both the wirebonding quality is OK. How do we benchmark this? Which one is suitable for production with less cost and less problems for wire bonding? Both vendors claim that their process is good. With our experience we have been having some trouble with ENIG and this is the first time we have electroplated PCBs for wire bonding.

reply »

MA/NY DDave

#23592

Plating for aluminum wire bonding | 2 March, 2003

Hi

It sounds like you are in a different country or in the USA a long time ago.

I have no idea of the process either is using, so they can be comparable or one can be better than another and lop sided compared to other places.

Benchmarking I think is the wrong term since you are only using two suppliers. Benchmarking applies to the entire industry.

Ask each of them to slightly vary the process or give you parts from the low and high side of their outputs. Compare results in your shop. Now this isn't really statistical yet it give you a sense of each supplier's process as run through your shop.

YiEng MA/NY DDave

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

reply »

#23605

Plating for aluminum wire bonding | 3 March, 2003

Electroplated nickel and electroplated gold is the more �traditional� finish for aluminum wire bonding. That surface was an extension of gold thermosonic wire bonding that requires a thick [50 thou min] gold surface. For years fabricators recommended a thinner [~25 thou] gold for aluminum wire bonding.

Separately, with the drive for flatter solderable surfaces, ENIG became a desirable surface for fabricators, because they could command high mark-ups from their customers that were used to HASL pricing. Assemblers found ENIG acceptable for aluminum wire bonding, which is good because to increase the maximum imm gold plating thickness [<12 thou] to the traditional thickness requires an additional electroplating fabrication step.

So, your suppliers are correct. Either surface works.

We don�t think either surface is superior. * Electrolytic [electroplated] nickel plating is a less demanding process that electroless nickel * Electrolytic require bussing and electroless doesn�t, but if you�re running electroplated gold the boards require bussing and maybe accomplished using the same racking, just prior to gold plating. * Electroless nickel has traditionally been used where no bussing is available. * Lack of ENIG process control can result in black pad [search fine SMTnet Archives for background] that affects wire bonders as well as solderers.

So, price differences may have more to do with individual fabricators set-up and process preferences than underlying materials or structural costs.

Purity of plated material is very important for any type of wire bonding process. * For aluminum wedge bonding the nickel becomes critical. Always have a minimum of 50 thou nickel. You want the nickel plated as soft as possible, no contaminates or voids. It requires good control of pH and plating chemicals in the plating bath. * For aluminum wedge wire bonding, you want the gold thickness to be a minimum of 3 thou and a purity of 99.99% (no thalium).

What problems are you seeing with your ENIG boards? * Wire bonds don't stick when bonded * Wire bonds fall off, or break during service

reply »

ramanandkini

#23805

Plating for aluminum wire bonding | 15 March, 2003

Thanks Mr.Dave. Sorry for the delay. One of our major prolems in ENIG plated surface is that wire does not stick. For many years we were getting a bond strength of 14 grams in both ENIG and electroplated surfaces. Now one vendor supplies ENIG boards with just 5 grams and says it is acceptable as per MIL standards. We tried with AT&S, and they gave ENIG plated surfaces with Atotech chemistry that had just 4~5 microns of Nickel and 0.03 microns of gold. That bonded well with 10 grams.

Our regular vendors plate minumum of 7 microns of Nickel and then 0.03 microns of gold. They are mainly from Hongkong.

I saw a difference in the surface plated by AT&S. It has a very fine grain structure while our vendors surfaces are a bit shiny. I understand from AT&S that they use pumice powder spary for copper clad preparation and our vendors use traditional brushing. Is there any realtion between surface roughness and wire bonding? If so what should be the "Ra" value?

I am able to bond AT&S plated surface even after two months but on my regular vendors PCB, I should do immediate bonding.

I want to do a small research work on the ENIG, Electroplating, surface roughness and the time of bonding etc., It takes some time. Kindly send in your inputs/references.

reply »

#23810

Plating for aluminum wire bonding | 17 March, 2003

Ramanandkini

14 gm versus 5 gm: 14 gm [even 10 gm] sounds better than 5 gm. We get good bond pull XBAR (high teens) and low sigma (<0.75).

5 gm is OK per MIL spec: What MIL spec? We�re unaware of a MIL spec that says 5 gm is OK. And even if there was such a specification, the requirement that wires stick to pads takes precedent.

Just 4~5 microns [150+ uin] of nickel: This amount of well-plated nickel should be OK.

AT&S uses pumice powder spray for copper clad preparation: Bingo!!! That's IT!!! You�re very close in wanting to measure surface roughness. And you could be correct, but we�d bet a box of donuts that pumice is imbedded in the copper. [We�ve seen this previously, both on wire bonds and on soldered surfaces.]

[Ask your supplier to do some cross-sections and check for pumice inclusions.]

References: Check in the SMTnet Newsletter for a review of �Wire Bonding in Microelectronics: Materials, Processes, Reliability, and Yield� George G. Harman, IMAPS; 2nd edition (August 1997); McGraw Hill Text; ISBN: 0070326193. Every wire bonder should own this book. Since it�s been around for a while, you probably can find a gently used copy.

reply »

ramanandkini

#24495

Plating for aluminum wire bonding | 13 May, 2003

Dear Sir, Sorry for the delay in reply. This is regarding MIL standard for bond pull strength: MIL-STD-883C dt.25/08/83 Bond strength (destructive bond pull test); refer table-1 Minimum bond strengths for various wire composition & diameter given.

regards

reply »

StikNPeel™ Rework Stencils

SMT in-printer dispensing