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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Gold Surface Finish on PCB's

Views: 11317

Mike_Kennedy

#38283

Gold Surface Finish on PCB's | 8 December, 2005

Our new PCB supplier can give us a 20% price reduction if we convert from an ENIG finish to a FLASH gold finish. My uderstanding between the two is as follows: ENIG - Nickel is plated over copper and then immersed in Gold which forms a self limited finish over the top of the Nickel until all Ni is covered. FLASH GOLD - Nickel & Gold are electro deposited and the thickess of the deposits relates to the time in the bath. This process is not self limiting and can lead to higher amounts of deposit in some areas. My concern is with the reliability between these two finishes. Has anyone had any experience with FLASH gold PCB's and found with BGA's or LPP's the solder joint life is reduced? Thanks, Mike Kennedy

reply »

#38287

Gold Surface Finish on PCB's | 8 December, 2005

Your understanding of the situation is correct.

In both cases, you solder to the nickel. The gold [Au] protects the nickel from oxidation. When soldering, the gold moves into solution in the solder and forms an intermetallic compound [IMC] with the tin [Sn] in the solder. The primary phases are AuSn4 and AuSn2. Other intermetallic phases can form in this system. The gold-tin IMC is strong, but brittle. So, when the IMC reaches a certain portion of the solder connection, the IMC can enbrittle the connection. Failure occurs at the boundry between the solder and the IMC.

Effects of Gold (Au) on Solder Properties Concentration Range in Weight Percentage w/o Au in Sn63/Pb37 Solder [A Sugarman/Loranger Int'l. Corp] 0 - 0.3% Solubility limit of Au in solder reached and AuSn4 intermetallic forms (1) 0.2 - 0.5% Contamination level at which solder should be renewed (2), (3), (4) 0.3 - 2.0% No effect on solder properties (1) 2.0 - 3.0% Spreadability and fluidity of solder reduced (1) 3.0% Highest acceptable concentration of Au in solder before deleterious effect can occur. (1) 3.0 - 7.0% Sheer strength and ductility of solder decreases (1) 7.0 - 10.0% Rapid increase in hardness of solder (1) (1) CA Harper "Electronic Packaging Handbook"; McGraw-Hill; 1997, 2e, p5.50-5.53 (2) ANSI J-STD-001B Reqmnts Soldered Electrical & Electronis Assemblies (3) El-B. Hannech and CR Hall "Diffusion Controlled Reactions in Au/Pb-Sn Solder System"; Materials Science and Technology; Vol 8, No 9, p.817, Sep 1992 (4) RN Wild "Effects Of Gold On Solder's Properties"; Electronic Packaging & Production; Vol 8, No 8, p.27, 1967

So, you need to be spending all that money that you're saving on making sure the boards that you receive don't have too much gold on them.

reply »

#38290

Gold Surface Finish on PCB's | 8 December, 2005

Hi,

We have recently moved to flash gold as well, as we had solderability issues with emerson gold, and this was supposed to be a solution.

However I did not know flash gold would be thicker, and could have other issues. Is flash gold a stable board finish?

Regards,

Grant

reply »

#38292

Gold Surface Finish on PCB's | 8 December, 2005

Grant

The solution to soldering problems with immersion gold boards is: find a supplier that can control his/her process.

Electroplated gold over nickel is a very reasonable process. Billions of wire bound boards use it evry year.

reply »

Mike_Kennedy

#38322

Gold Surface Finish on PCB's | 8 December, 2005

Dave, Excellent information thank you. What I cant seem to establish is how other manufacturers have found the FLASH gold in terms of reliability especially considering that the flash process is more difficult to control.

reply »

#38324

Gold Surface Finish on PCB's | 8 December, 2005

Mike:

You're correct, but in better shops there are guys with little black books in their back pocket. They know how to contol their variables and give you a good product.

I worked at this plant where we cooked dirt. It glowed orange-red and flowed like water from the kiln. I'd ask Joe Henry how he knew and he'd say "Dave, I just know."

reply »

Mike_Kennedy

#38328

Gold Surface Finish on PCB's | 9 December, 2005

Grant, I dont quite understand the implications my self yet on gold deposit using FLASH or ENIG. However the important part of the gold finish is that it is kept under 3% of the total solder joint, so normally its safe. With the FLASH process it is more difficult to get an even coating so there can be a risk in this part. But at the moment, the main concern I have is with the Gold plating process of FLASH and the porosity of the surface left after the process. With the ENIG process the GOLD grabs onto the Nickel until there is no more gold left, so in this way it is self limiting. This also suggests that if the Nickel was corroded, the Gold wouldnt stick, so you would get an indication there is a problem with a particular baord. With FLASH it is electro deposited and my worry is that with variation in the shape of the track or pad, you may get different thicknesses. Also you may get some areas where to gold layer is too thin and may result in Nickel oxidation later. Not sure about this by this is my concern.

reply »


RDR

#38337

Gold Surface Finish on PCB's | 9 December, 2005

We used flash gold once by mistake and we did pay for it. A very high fallout of BGAs and some of the leaded devices for latent failures from fractured solder joints. I do not know how the other types of components faired in this process but I would imagine that the joints at the other locations were probably just as brittle if not more so.

reply »

Mike Kennedy

#38440

Gold Surface Finish on PCB's | 13 December, 2005

Russ, After a bit more research, I have found that the FLASH gold process doesnt use Phophrous in the process so should not have the black pad problems of ENIG. I am now surprised that you had problems. What did you experience? Was it an oxsidisation problem with the Nickel or a solder reflow problem, mabye you didnt solder long enough? Thanks, Mike Kennedy

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

reply »

Mike Kennedy

#38441

Gold Surface Finish on PCB's | 13 December, 2005

Dave, I have read an article by Vern Solberg who said that Electro plating Ni/Au in not recommended for PCB's. I have tried his email but get bounce back. Do you know him or his email? I knows its a long shot to ask you, but it never hurts to ask....unless you come to my house and punch me of course. I have also read an article by George Dalich who describes the process difference between ENIG & electrolytic(flash) gold. My opinion now is that there are advangages & disadvantages for both but as far as I can tell, they should both give a similar result as far as long term reliability is concerned. What do you think? Thanks, Mike Kennedy (Somewhere is Australia)

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

reply »

#38451

Gold Surface Finish on PCB's | 14 December, 2005

Mike Ozzy

If I came to your house, I wouldn't punch you. I'd make you buy me beer.

Vern Solberg; Tessera Technologies Inc., 3099 Orchard Dr, San Jose, CA 95134; 408 568 3734 F408 894 0768 vsolberg at tessera dot com. If that doesn't work try: Tessera, Inc.; 3099 Orchard Dr, San Jose, CA 95134; 408-894-0700 F408-894-0768 sales at tessera dot com

Comparing immersion gold with electroplated gold: * A 1-3 microinch thick electrolytic gold plating will be porous, but a 1-3 microinch immersion gold surface finish is not porous, because of differences in the nature of the plating processes. You can look at a properly controlled immersion gold surface and see pore free coverage with a scanning electron microscope. * Thick electroplated gold will cause embrittlement problems. * With ENIG your trace and all surfaces are completely covered with ENIG. This will not be the case with electrolytic processes. With electrolytic plating you will have exposed copper on the sides of the traces. The trace is plated ONLY on top with nickel and gold, with the sides being exposed copper. The potential for some serious galvanic corrosion is high. * Electrolytic gold will have differentials in thickness due to circuit geometry, which can be fairly drastic depending on the board design.

Think about this ... Specification: MIL-G-45204 Gold Plating, Electrodeposited [This spec has been cancelled, but is still in wide use. It was replaced with ASTM B 488 Standard Specification for Electrodeposited Coatings of Gold for Engineering Uses] * Type-I 99.7% gold minimum; hardness grade A, B, or C [For general-purpose, high-reliability electrical contacts, solderability, and wire wrap connections.] * Type-II 99.0% gold minimum; hardness grade B, C, or D [A general-purpose, wear-resistant gold. It will not withstand high-temperature applications because the hardening agents in the gold coating will oxidize.] * Type-III 99.9 % gold minimum; hardness grade A only [For semiconductor components, nuclear engineering, thermocompression bonding, and high-temperature application.]

Hardness Grades: A 90 knoop, maximum; B 91-129 knoop, inclusive; C 130-200 knoop, inclusive; D 201 knoop, minimum

reply »


RDR

#38458

Gold Surface Finish on PCB's | 14 December, 2005

We experienced brittle solder joints that did not withstand thermal cycling as I was told. I am positive about the fractured joints but was informed by customer that it was thermal cycling of the part when they failed so they were latent failures. I do not know or remember if it was normal operation thermals or a burn in type of testing. It wasn't black pad the solder joints were great as far as the way they looked etc...

reply »

SMT in-printer dispensing

ii-feed SMT Intelligent Feeder