soldering to thick gold plating

Subscribe to Thread

We are trying to apply solder to boards with 80 microinches ... - Jul 11, 2006 by

Not reflowing completely. Slower = hotter. Try a slower co... - Jul 11, 2006 by

We agree with Chunks. You need to turn-up the heat [slow th... - Jul 11, 2006 by davef

Hi, I believe you have classic gold imbrittlement here. ... - Jul 12, 2006 by Flipit

Hi, One other note. You don't want gold to be more than ... - Jul 12, 2006 by Flipit

mattkehoe@sipad.com

#42660

soldering to thick gold plating | 11 July, 2006

We are trying to apply solder to boards with 80 microinches of electro plated hard gold. No components involved, just print 63/37 paste and reflow. The results are very poor on some boards, not so bad on others? Pictures at http://www.sipad.net/thickgoldsolder.htm Looks very grainy and no flowing. Brand new paste, standard profile. Boards before and after without he thick gold look fine.

Is there something special that needs to be done to get the solder to flow on this tick gold? Hotter? Slower?

Matt Kehoe SIPAD Systems

reply »

Chunks

#42661

soldering to thick gold plating | 11 July, 2006

Not reflowing completely. Slower = hotter. Try a slower conveyor speed.

reply »

davef

#42665

soldering to thick gold plating | 11 July, 2006

We agree with Chunks. You need to turn-up the heat [slow the conveyor] because you have significantly changed the melting point of the solder alloy by adding so much gold. Relative to ENIG you should expect: * Similarly smooth surface, possibly a little less smooth * Higher wetting angle * Duller finish on solder connection

If components are soldered to this substrate, we'd expect significant reliability loss [from ENIG or other common solderability protection] due to the large portion of gold in the alloy.

reply »

Flipit

#42682

soldering to thick gold plating | 12 July, 2006

Hi,

I believe you have classic gold imbrittlement here. With 80 microinches of gold you are way over the limit. You can try to reflow longer time and at a higher temperature. The gold does not melt into the solder joint. The gold dissolves into the solder joint. By raising the reflow temperature or increasing the reflow time above liquidus, the dissolution rate of the gold increases. You want to dissolve most of the gold into the solder joint and reflow or attach to the nickel below the gold. You do in essence create sort of a new alloy. You have a solder joint embrittled with tin/gold needles. I suspect that your customer will have a difficult time reflowing the solid solder deposits. No matter what you do, you will have gold embrittled solder joints. In fact the solder joints will probably look even worse if you raise the temperature or increase the reflow time. This will dissolve more gold into the solder joint and gold embrittle the solder joint even more.

If you print more solder paste or increase your stencil thinkness, your solder joints will look better. This is because you decrease the gold content dissolved into the 63/37 paste by increasing the 63/37 volume. This is also the reason ENIG (thin gold) looked better. The thin gold was probably under 10 microinches thickness.

If you want to create a good quality solder joint that looks good and performs well, decrease the gold thickness. Go to selectively plated gold. Use Indium / Lead alloys if you must solder to thick gold. That presents an entire new set of problems and processes.

Good luck,

Chris

reply »

Flipit

#42685

soldering to thick gold plating | 12 July, 2006

Hi,

One other note. You don't want gold to be more than 1% to 2% of the the solder joint. Above this percentage you will have embrittlemant issues. Lots of information on gold embrittlement on this site and on the Internet.

I estimate your product to be approximately 25% gold by weight.

Here is a great paper on gold embrittlement.

www.semlab.com/GoldEmbrittlementofSolderJoints.doc

Thanks,

Chris

reply »