SMT line reflow

[limct]

Hi all,
I'm not sure if this is the right place to ask this question, but here is how it goes:

I have two SMT sub-con houses who are equal in terms of their equipment. My product is a two layer PCBs which the top and bottom copper are OSP treated. The issue I'm facing is that one of the sub-con always deliver perfect goods where the bottom copper of the PCB just look "stainless" like it never go thru' the reflow, but the that's the other sub-con who prouce boards with tarnished bottom copper, and they always claim it's the nature of their so called "no clean" process.

I'm from Mechanical background, so my first impression on tarnished copper is there must ne some unwanted chemistry reaction somewhere. As I do not see any sort of marks that resemble a finger print, I think the chance of mishandling is very minimal.

I'd really appreciate if someone can shed some light here.


Best regards,
ct

 

[ScottyUK]

For those wondering, OSP = Organic Solderability Preservative.

At the moment I'm not sure what you mean by 'tarnish'. I assume that any copper protected by the solder mask will be unaffected, and that any copper not masked should be tinned. Are you saying that there is additional copper which is unmasked but which also not required for component placement and therefore has not been tinned during the reflow process, and this is what is affected? If so it sounds like it could be related to thermal breakdown of the OSP during reflow.

What detail can you give of the two assembly processes? I assume screen printed solder paste then pick & place, pre-heat, reflow, cooldown? No washing / drying process? Are the two assemblers using the same process and the same materials or have you left the selection of OSP, solder paste and reflow temperature profile to them?

Post a couple of photos if you can - there is an option on this site to link to a file already on the web or you can upload to a file host.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[limct]

Hello Scotty,
Yes, the affected are is actually with OSP finish and not covered with solder during the reflow.

The two sub-cons are running the process in the same manner, no washing no drying process. Both use lead free solder paste but of different brand and are given the liberty to tweak the reflow profile (if necessary).

Here are the photos

http://www.fileshost.com/en/file/23957/good-pdf.html

http://www.fileshost.com/en/file/23956/bad-pdf.html

Thanks for your time on this.


Best regards,
ct

 

[ScottyUK]

First a couple of questions which might make some bits of what is below irrelevant:

Have those boards actually been through reflow yet? I can't see any sign of components or solder wicking through the vias, but with a lead-free solder that isn't much of surprise.

Do the boards originate from the same PCB house prior to assembly or do the assemblers make their own boards?


I'd say that this was surface contamination present on the copper prior to the OSP being applied, possibly an oily / greasy substance by the appearance. I seem to remember OSP is incredibly fussy about surface cleanliness to perform properly. What cleaning process is the 'bad' plant using immediately prior to applying the OSP? Do they have access to a rotary brush machine similar to

http://www.circuit-supplies.com/board_cleaner.htm?

That company was Hibass when I was in the industry - times change!

I'd be paying a very close eye to their board manufacturing process from the moment it comes out of etch until it is ready for assembly. Pay special attention to the hole plating operation, especially the rinse and dry phases to ensure that all residuals are removed from the vias.

There is a suspicious concentration of the marks around the two tiny vias. Why are they so much smaller? Looks like an avoidable complication based on what is visible. Is a component present directly on the other side - i.e. could this reasonably be a result of the flux wicking through the vias?

Request a QA audit and trace a batch through. Run a batch that you have personally witnessed being polished and degreased and watch it all the way through OSP print & cure to final assembly. Inspect at each stage. Don't let it out of your sight.

Request a batch be run using the same solder paste type that the 'better' subcontractor is using and see how that behaves. No-clean fluxes range from basic rosin-based types through highly selective synthetics. This could just be a simple incompatibility problem between the OSP and the flux.


I would be very tempted to tell the subby that their process is giving poorer results than their competitor and that they have a vested interest in resolving the cosmetic issues. In such a competitive industry an unhappy customer usually becomes a former customer pretty quickly.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[IRstuff]

I suggest that you use the Engineering.com file upload, as that has none of the annoying ad stream.

TTFN

FAQ731-376

 

[ScottyUK]

Oh, I forgot to mention that. What a horrible file host - damned singing adverts and all sorts of crap. IRstuff's suggestion is a good 'un. You can get to engineering.com from a link just above the 'Submit Post' button.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[limct]

Sorry for leading you gys to the lousy file hosting site. ya, in fact, i wasn't happy either with the too much of the ads.

Scotty,
Perhaps I should give you more info in order for you "investigate" further.

1) The boards are made by one PCB manufacturer, a same batch of boards run by the two assembly may turn out differently, so I'm sure we can rule out the suspicion on the OSP quality.

2) Yes, the components are mounted on the other side of the board. The two smaller vias are half surrounded by solder with 30 mils width solder mask bridge around it.

3) The assembly houses are given the liberty to use their own prefered solder paste. will it helpful to you know the details of the paste?

IN fact, the prime suspect that I can think of is the amount of flux content in the solder paste, but I'm not sure, will flux be so corrosive to OSP surface?

Again, thanks for your time on this.

Best regards,
ct

 

[ScottyUK]

Sounds like you're on the right track with a fluxing problem. Could be a compatibility issue with the OSP, or just a more aggressive flux. It's a long time since I was doing this sort of thing for a living so I'm probably not the best person to discuss modern no-clean solder pastes with. RMA fluxes and water-wash synthetics from the early 1990's is more my field. Are the pastes of roughly similar composition, i.e. both a rosin based type or whatever? The paste types and brands might be of interest, if only from a composition point of view. If you know them feel free to list them or post a link the datasheets. We used to get pastes 'brewed' to our spec with non-standard solder:flux ratios to suit our process, but our sister division making industrial and automotive products was a pretty big customer and the paste manufacturer was prepared to invest time with us to get the best out of the process.

Initially I'd request that the 'bad' subcontractor runs a batch using the same solder paste as the 'good' subcontractor. Solder pastes are easy enough to get hold of and not hugely expensive. The subcontractor might need some time to play with the process to get the screen printing working properly and to get the reflow profile right so you'll have to be prepared to supply some boards you're prepared to write off while they do their process development. If they believe that they might get a better process out of it then they should be more than willing to help without billing the engineering time if you are prepared to free issue the materials.

One daft thought: they're not doing anything naughty like adding solvent to the solder paste to modify its printing characteristics are they? It certainly won't be officially documented in the procedures - I occasionally used to find an operator adding toluene to a jar of paste which had been open too long in order to thin it a little for the printer. Magically QA would then descend to bring down the wrath of god upon the production supervisor! Adding anything to a paste usually has a negative effect on the behaviour of the paste.


----------------------------------

If we learn from our mistakes I'm getting a great education!