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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SN100C Selective Solder Voiding

Views: 5339


SN100C Selective Solder Voiding | 5 November, 2008


We have recently installed a selective soldering machine and are seeing some voiding/cratering in random locations. We have done some basic experimenting with fluxing parameters, preheat and dwell times but haven't been able to eliminate it. We've seen this issue on several different boards.

I've seen this with tin/lead wave soldering when there is porus copper plating in the barrels causing outgassing but that's not the problem here (had some boards cross-sectioned and confirmed proper plaing thickness).

We're planning to do a DOE on this but wanted to solicit input from those with experience. Is there some inherent issue with LF alloys or SN100C in particular that makes them suceptible to this or do we have a process issue or maybe both? I've attached a couple photos. Thanks.


reply »


SN100C Selective Solder Voiding | 8 November, 2008

SN100 was designed for selective soldering and producing low levels of voiding. But thin barrel plating is a common cause of voids in wave and selective soldered connections, as you say. As a result, you checked the plating thickness and it was acceptable. So, the issue that you see is curious.

We're stuck on the barrel plating issue. So, please bear with us. We're not thinking barrel thickness from the fabricator. We thinking about the thickness after your ~100% tin solder has had a chance to dissolve the copper, bubbling away on that shiney new solder machine. Questions are: * What was the solderability protection on the boards? Can't be ENIG, correct? * Did you measure the plating thickness on voided boards or unsoldered boards? * When you measured the plating thickness, was the knee thickness acceptable on soldered boards?

From a Celestica presentation:

reply »


SN100C Selective Solder Voiding | 10 November, 2008


You've got me worried - All of the boards we've tested to this point are LF HASL. Why do you say "can't be ENIG"? Am I missing something?

We have not sectioned soldered boards as yet but plan to shortly.


reply »


SN100C Selective Solder Voiding | 10 November, 2008

Hello Hoss,

try with the flux again. I would first change the flux. Alcohol based for me is always better. But you can try water based too. 1. Change the flux 2. Go with more pressure on your atomizer(or whatever they call it on your machine). Just may the s very small. I had simular problems before. Most probably the main reason is problem with the board finish, but flux is you partner in finding the right way.

Good luck, Emil

reply »


SN100C Selective Solder Voiding | 10 November, 2008


On the earlier ENIG comment: No, you're not missing anything. ENIG board are less likely to have copper dissolution issues that other solderability protection.

reply »


SN100C Selective Solder Voiding | 10 November, 2008

Hello Hoss

Voidin/blow holes can also be exacerbated by residual flux. Have you experimented with dwell times and separation speed parameters with your selective ?

What related parameters are you currently running processing under ?



reply »


SN100C Selective Solder Voiding | 11 November, 2008

Thats the problem residual residue from HASL fluxes Lead free especially are worse.Very hygroscopic and out gass and produce dewetting as well over wave soldered boards especially. Leave iron on blown joint and will after several seconds fizz and pop and spit out solder if particularly bad. You will alos note a sweet acrid smell/fume as the residue wicks towards the iron. We also produce Sn/Ni/Cu/Sb alloys with no problems on selective so not alloy. everyone blames poor plating but is not this at all I have commented on previous threads about this seen it alot in earlier Lead Free days. Get PCB's saponified cleaned using HOT solution rinse properly then dry and retry. hope it helps Cheers Greg

reply »


SN100C Selective Solder Voiding | 12 November, 2008


We are experiencing voids, also at Selective, on a transformer with known flux residue issues. The supplier told us they were using a OA, or RMA flux which has to be cleaned, and we accepted these parts anyway.

How does the "separation speed" parameter affect voids?

reply »


SN100C Selective Solder Voiding | 12 November, 2008

If its an OA flux then this too acts similar to HASL flux in that they are very hygroscopic and would attract moisture and that would blow out the solder causing a void. Remember these fluxes have very high boiling solvents that do not get volatised during normal soldering temperatures

reply »


SN100C Selective Solder Voiding | 12 November, 2008

I attended SMTA's copper dissolution webinar this morning and am almost convinced that this is not our problem. Still planning to do cross sections to verify but our dwell times are nowhere near levels that could cause a problem assuming good plating thickness to begin with.

Jim - We honestly haven't run that many boards through the equipment so we haven't had much of a chance experiment with parameters. Machine isn't released to production yet.

We did start with the following settings:

Pot temp - 300C Process speed - 10in/min Preheat Dwell - 1-2sec (with nozzle ~.100" below solder target) Dwell - 2-3sec Flux Pressure - 3PSI Flux Atomization Flow - 6 CFH

We're going to try Greg's suggestion of a saponified wash for the LF HASL boards and also run some ENIG samples and let everyone know how we made out.

Thanks for all of your input.

reply »


SN100C Selective Solder Voiding | 13 November, 2008

try Flux Atomization Flow - 15 CFH

reply »


SN100C Selective Solder Voiding | 11 December, 2008

Update - We've tested a board previously having voiding issues with ENIG versions and the voiding has completely disappeared. No other variables were changed. Looks promising.

I have an order of LF HASL boards arriving soon that we plan to split into groups of "straight from the pkg", "saponified wash" and "saponified wash with bake out".

More when I have it.

reply »


SN100C Selective Solder Voiding | 11 December, 2008

thanks for the update Hoss

reply »


SN100C Selective Solder Voiding | 13 February, 2009


Received cross-section analysis for ENIG plated boards and am happy to report that we have no voiding issues. Intermetalics look great and no copper dissolution.

No process parameters were changed, only the PCB plating. For now, we are only running ENIG boards and will do a more in depth study on the LF HASL plated PCBs in the near future.

reply »

PCB machines

JTAG Live PCB Debug Software