Nickel-gold plating a whole board

When I worked on the avionics and cal lab benches at the Navy base in Jacksonville, FL, we had tons of equipment failures that were easily traced to mechanical joint failures caused by brittle solder-gold contact. The gold on the component leads would delaminate and leave the leads floating open or intermittent inside the soldered connection. As a result, the microminiature soldering training program there issued a requirement that on all repairs requiring replacement of a component having gold plated leads, the leads must be tinned with solder, then the solder wicked off with a fluxed braid or dipped into a solder pot several times to leach all the gold from the leads. I remember that a lot of HP equipment having Ni-Au plated boards had this problem, resulting in a lot of the older equipment going to the salvage yard to avoid the cost of troubleshooting and repair.

Oh-oh, that doesn't bode well then. The gold is nowadays extremely thin so wouldn't likely saturate anything. But who knows, I sure don't want to see a repeat of that HP issue.

My favorite would be nickel-only but they don't seem to do that anymore. So that would make it a boutique process.

Have you looked at IPC-4552? Art

Sure, and it talks about solderability and all that:

However, statements like these (page 17, under aging) didn't exactly instill confidence in me, quote "These samples were now 240 plus days old since plating, (I knew there was a reason I don?t clean my office)".

Now the gold layer thickness these days is in the tens of nanometers. Not sure if the HP boards Jeff was mentioning had much thicker gold. If you have too much gold dissolving into the solder then all sorts of evil things can happen.

There are gold intermetallics, but that's what's supposed to happen. Gold boards are the bee's knees for solderability. Use 2% silver solder to protect the gold if you feel a need.

Cheers

Phil Hobbs

For mil/aerospace stuff, we de-guild everything... I guess you can't do selective gold then....

Have had no problems with a number of batches (SMT reflow + a few parts hand soldered, Sn63Pb37). Gold got a bad reputation because they actually used to put some on, now that the dollar is worth 5.5E-4 oz of gold, it's just a protective layer on the nickel.

This has to work with regular leaded solder. So you think ENIG for the whole board will be ok then?

$1,782 per ounce yesterday. Depressing :-(

That's what I thought, maybe HP had the really lay it on there. Nowadays you get 50 nanometers or so. Like whatever flakes off of a nugget flying across the fab at supersonic speed.

I don't even want to think about what the next gold crown will cost should I ever need one ...

Unfortunately I can't on this one, it's either all ENIG or none of it.

IIRC the problematic phase exists only above 0.8% gold content, which won't happen with ENIG ultrathin 0.2um plating. All the surface gold gets easily dissolved, so no pb with the intermetallics transition, and no pb with the joint too thanks to the very low gold content.

I've a customer that uses exclusively ENIG process for his sensors (SMDs down to 0201) with no pb at all 'til now.

No pb with refusion SMDs, but the solderwave guys don't like gold plated boards because that makes gold concentrating over time, rendering their bath unusable.

-- Thanks, Fred.

There are a whole lot of gold plated boards out there, and have been for a long time. I suspect that the Navy problem was something unusual, e.g massive temperature cycling.

Cheers

Phil Hobbs

Good to know that there's a lot of those boards out there. Once assembled it becomes impossible to see because the solder will have eaten up the ENIG gold layer. It's more of a whiff than a layer.

"Fred_Bartoli" a écrit dans le message de news:4e4c3f69$0$27001$ snipped-for-privacy@news.free.fr...

IIRC, I found that documented on Xilinx site, in some appnote about fine pitch BGAs

You don't even get as much as 0.2um anymore these days.

If the gold price keeps climbing at the current rapid pace maybe some day they'll stop complaining :-)

Thanks, guys, that helps me a lot. Looks like we can go the ENIG route then. It also makes the board look rather posh but that's not the reason.

We do a few of our boards that way, specifically boards that have BGAs and must be ROHS. The surface is very planar, so BGAs like it.

You have to practically lock the boards and solder in separate rooms; the solder almost leaps onto the boards. The gold is only around 10-15 micro-inches, so it quickly dissolves into the solder. That leaves the solder sticking to the nickel, which I don't really understand.

But it works great, and the bare boards look cool.

ftp://jjlarkin.lmi.net/Gold_T564.jpg

John

It is not just that it delaminates. There is more :) Gold makes brittle intermetallic alloys with solder that are prone to cracking. That is why there are special solders for soldering gold. In50/Sb50 e.g. one of those solders and I do even have a roll...

Yes, it looks like a million Dollars. I also have two SMAs on there, gold plated.

The solder up there at Y1 on your board looks very shiny. If it was gold-choked it would look dull and cracked. Well, I just fired it all off, we'll get golden boards out of Colorado.

's

As far as I know, all my crowns are chrome-cobalt alloy. It doesn't make them cheap. I don't think I've ever needed to have one replaced. There's also a titanium implant, put in (at their cost) when the dental hospital made a hash of doing a crown and post restoration on my lower left canine, but that goes down into the jaw-bone.

-- Bill Sloman, Nijmegen