Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SMTnet
»
Electronics Forum
»
soft gold vs immersion gold

soft gold vs immersion gold

Subscribe to Thread

Hi How does one distinquish between hard gold and immersi... - Aug 11, 2004 by

Q1: How does one distinguish between hard gold and immersion... - Aug 11, 2004 by davef

This is interesting, I was once told by a PCB supplier that ... - Aug 12, 2004 by RDR

Russ: Immersion gold [imm Au] is "soft" and it's "gold", but... - Aug 12, 2004 by davef

Hi Guys thanks for the replies. Dave, the spec is no... - Aug 16, 2004 by

You bet our reponse changes. Q2R: Hard gold thickness is ... - Aug 16, 2004 by davef

what about areas that are soldered ? This is def not a E... - Aug 16, 2004 by

As we mentioned in an earlier response in this thread [Q5], ... - Aug 16, 2004 by davef

ok I agree there has been a confusion with terminology Wh... - Aug 16, 2004 by

First, we didn't say "electrolytic soft gold with a minimum ... - Aug 16, 2004 by davef

thanks all ... - Aug 19, 2004 by

Kris

#29998

soft gold vs immersion gold | 11 August, 2004

How does one distinquish between hard gold and immersion gold ? thickness is 4 to 10 u" for Au and 300 to 400 u" for nickel

what happens if the gold drops below 3" and the nickel is above 350"

what are the typical failure modes associated with hard gold surfaces ?

can somebody direct me to literature concerning soft gold ?

thx

reply »

davef

#29999

soft gold vs immersion gold | 11 August, 2004

Q1: How does one distinguish between hard gold and immersion gold? A1: Distinguish between hard gold and immersion gold by measuring the hardness of the gold. We�d expect the hardness of hard gold to be >90 Knoop.

Q2: Thickness is 4 to 10 u" for Au and 300 to 400 u" for nickel A2: This looks like a ENIG spec. Although, the Ni is quite heavy, 150 uin is sufficient for most applications. [IPC-4552 ENIG specification: * Gold thickness of 0.075 - 0.125 microns [3 - 5 microinches]. * Nickel thickness of 3 - 6 microns [120 - 240 microinches]].

Q3: What happens if the gold drops below 3" [sic 3uin]? A3: 3uin electrolytic gold is porous and allows corrosion of Ni underplate, resulting in lousy soldering. Although not as necessary, we like to see immersion gold thicker than 3uin, also. [just old fuddy-duddies talking here though.]

Q4: What happens if nickel is above 350" [sic 350uin]? A4: If Ni is >150uin, the assembler wastes money.

Q5: What are the typical failure modes associated with hard gold surfaces? A5: When we hear �hard gold�, we think of wear * Could be solderable surfaces, but usually the people that specify �hard gold� for solderable surfaces don�t know what they�re talking about and have just tweeked their wear surface spec to become something that they reason to be a solderability spec. See, platers create hard gold by alloying cobalt [or maybe nickel] with fairly pure [soft] gold. Obviously, if you're soldering, you don't want Co alloying with your solder. [Let us know when you find a phase diagram for that.] * Wire bonding is soft gold, no discussion.

So anyhow, we�re talking wear. Common "hard gold" failures are: * Discolored gold. * Gold that is poorly attached to the nickel. * Poor thickness control. * Low thickness. * Pitting. * Damaged solder mask.

Finally, both 'hard' and 'soft' gold pertain to electrolytic gold. Neither applies to immersion gold, which is a completely different process.

Q6: Can somebody direct me to literature concerning soft gold? A6: When we want to refresh on board fab issues, we scamper over to the bookshelf as fast as our stubby legs can carry us and grab Coombs' "Printed Circuits Handbook" [McGraw-Hill 5e ISBN: 0071350160 ]

Plating books are: "Electroplating: Fundamentals of Surface Finishing" Lowenheim, FA (0070388369 McGraw Hill) (1872422039 Electrochemical Pub) "Graham's Electroplating Engineering Handbook" 4Ed, Durney, LJ Chapman & Hall (0412741105) "Modern Electroplating : Electrochemical Society Series" 3Ed, Lowenheim, FA John Wiley (0471549681) 'Metal Finishing Guidebook and Directory'. Published annually by Elsevier Science Publications, New York, New York, USA. ISSN 0026-0576.

Lecture notes on plating: http://www2.ece.jhu.edu/faculty/andreou/495/2002/LectureNotes/Plating.pdf

Rudy Sedlak�s take on non-electrolytic plating processes: http://www.pcbfab.com/connecto.html

We've observed that you've been working this gold issue for a couple of weeks now. Is there a specific issue that we should be discussing?

reply »

RDR

#30008

soft gold vs immersion gold | 12 August, 2004

This is interesting, I was once told by a PCB supplier that "soft gold" was immersion and the "hard gold" was the plated. But I agree with Dave F., Chris, what's going on?

Russ

reply »

davef

#30011

soft gold vs immersion gold | 12 August, 2004

Russ: Immersion gold [imm Au] is "soft" and it's "gold", but it is NOT "soft gold". Here's the question for your board house: If imm Au is "soft gold" and imm Au self-limits <12uin, how does he/she plate the 30+uin of "soft gold imm Au" necessary for wire bonding?

That's why it's fortunate that Kris don't ask how to tell the difference between soft gold and immersion gold in "Q1" [above]. We would not have gotten away with a slick answer, like the one that we gave above.

One piece of minutia on this topic, platers lay-down a coat of immersion gold prior to beginning their electrolytic gold plating process.

reply »

Kris

#30044

soft gold vs immersion gold | 16 August, 2004

Hi Guys

thanks for the replies.

Dave, the spec is not for enig but for hard gold

does your response alter if thats the case /

reply »

davef

#30048

soft gold vs immersion gold | 16 August, 2004

You bet our reponse changes.

Q2R: Hard gold thickness is 4 to 10 u" for Au and 300 to 400 u" for nickel A2R: This is a hard gold spec?? It looks like a ENIG spec. Although, the Ni is quite heavy, 150 uin is sufficient for most applications. The gold seems quite thin [IPC-2221 says something like ... For edge connectors and areas not to be soldered: * Nickel - 2.5um [~100uin] min [Class 2 & 3]. * Gold - 0.8um [~30uin] min [Class 2], 1.3um [~50uin] min [Class 3]].

reply »

Kris

#30056

soft gold vs immersion gold | 16 August, 2004

what about areas that are soldered ?

This is def not a Enig process and the specs are as they are not sure why

thanks a lot for your help

reply »

davef

#30057

soft gold vs immersion gold | 16 August, 2004

As we mentioned in an earlier response in this thread [Q5], you do NOT want to solder to hard gold. Hard gold is a wear surface. [That's IT, that's the LIST.]

If you want to solder to gold, then either use: * IPC-4552 ENIG specification, mentioned earlier in this thread [Q2] ... OR * Specify "electrolytic soft gold" thickness 5 to 8uin. Electrolytic nickel thickness of 120 - 240uin

In specifying "hard gold", your spec writers may have been trying to make sure they don't receive imm Au. They probably wanted to avoid the "black pad problems" [hypercorrosion] occasionally associated with imm Au, but just weren't clear about the hard versus soft gold issues. [Doesn't make them BAD people.]

reply »

Kris

#30059

soft gold vs immersion gold | 16 August, 2004

ok I agree there has been a confusion with terminology

When you say electrolytic soft gold with a minimum thickness of 5-8 of Au, is there a standard that you spec this from ?

reply »

davef

#30062

soft gold vs immersion gold | 16 August, 2004

First, we didn't say "electrolytic soft gold with a minimum thickness of 5-8 of Au". We said, "'electrolytic soft gold' thickness 5 to 8uin". FOCUS. ;-)

Second, we're unaware of an industry accepted specification that you can reference. IPC-2221 states words to the effect: Gold [max] on areas to be soldered: 8um [31uin for the metrically challenged, like us]. So, that's no help.

Anyway, here's your rationale for the gold:

LOW END: We know that 3uin of electrolytic soft gold is too porous to protect the nickel. So, it has to be more than 3uin. So, let's pick 5uin to cover standard slop.

HIGH END: We know that 10uin is pushing the allowable gold content in smaller connections. So, it has be less than 10uin. So, let's pick 8uin to be within standard slop.

Finally, seriously consider sitting with your board fab and negotiating a specification that makes sense to both of you.

reply »

Kris

#30132

soft gold vs immersion gold | 19 August, 2004

thanks all

reply »