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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Converting to 'No-Clean' process

Murad Kurwa

#4419

Converting to 'No-Clean' process | 14 April, 2000

Hi All:

Can you tell me specifically if you converted an exiting product made with Clean process using WS609 or compatible to 'No-Clean' process. We have to convert assemblies from current clean process at a CEM to No-Clean process in our own factory. Customer is concerned about long term reliability. There are approximately 2000+ components with minimum PTH (no wave - solder pot instead), mostly SMT with PQFPs, some PBGAs. Chip component only on the bottom side. All boards go through ICT, Board level FCT, Final System Assy and Test.

Have you addressed reliability concerns from the customers.

Product Quality and Reliability specifically related to process parameters that you addressed - Can you please comment for each:

Screen Printing issues / modification Stencil cleaning (frequency) Stencil / aperture design Paste type Flux type No-Clean residue Reflow profile ICT and use of probes Touch-Up/Rework cycle Customer Qualification Others

I WOULD APPRECIATE YOUR THOUGHTS.

reply »


RDR

#4420

Re: Converting to 'No-Clean' process | 17 April, 2000

Murad, I have just completed a switch over in our facility to no-clean. We used Alpha 609 and 857 paste and wave flux respectively. We currently use Alpha UP78 paste, NR330 wave flux, NR205 liquid flux for touchup and cleanline 7000 wire solder as our no clean chemistries. As far as reliability goes the manufacturers have the SIR,electromigration, etc... data that you will need to supply to your customers for approval. So far the only products we have not yet been able to approve are RF applications. your specific questions:

Stencil cleaning -Make sure you have a solvent that will clean the rosin. We have found that the stencil life, printability, and inprocess cleaning of no-clean is much better than with the W.S.

You will most likely need to redesign stencils to eliminate "midship" solder ball from caps and resistors.

Residue -should be very minimal after both wave and SMT, wave fixtures will show residues along edges, You will also inherit solderballs most likely from the wave process

REflow profile - same as W.S. usually, ensure that your soak time is minimal and not excessively long and hot.

ICT - not much affect unless you are probing solder joints, about 5000 hits before replacement

rework - operators will hate no-clean

As a note, I saw you use a solder pot instead of wave, this process is not very good for no-clean and leaves alot of residue and solder balls at least in my experience.

Good luck

Russ

reply »

#4421

Re: Converting to 'No-Clean' process | 17 April, 2000

Murad: Your customer is smart to be concerned about the affect of changing production processes on product reliability. Your Company is smart to inform customers and keep them in the loop about these changes. In developing your demonstration of the affect of the change, three angles to consider are:

1 CUSTOMER IS RIGHT. After all, your customer drives the bus, even if they don�t have a road map and should not have a license. Your customer may read over the specification and conclude that Appendix B testing is required in order to be compliant to J-STD-001C. Then again, your customer may have different ideas of what an appropriate demonstration would include. In this case, we have all been in situations where the customer is dead wrong, but we still have to do it his way. It may be necessary to educate your customer on when it is appropriate to require the Appendix B protocol and when it is more appropriate to use other test data.

2 MATERIALS COMPATIBILITY DEMONSTRATION. J-STD-001C, 4 tells you, the assembler, that all of the materials and processes you use should be compatible with each other. A good, sound policy. 4 says:

"The materials and processes used to assemble / manufacture electronic assemblies shall be selected such that their combinations produce products acceptable to this standard. Objective evidence of this compatibility shall be maintained and available for review. When major elements of the proven processes are changed, (e.g. flux, solder paste, cleaning media or system, solder alloy or soldering system) validation of the acceptability of the change(s) may be performed and documented in accordance with Appendix B. These process changes can involve a change in one of the process steps. They can also pertain to a change in board supplier, solder mask or metalization."

3 CHOICE OF THE FLUX. J-STD-001C, 4.2 reads:

"Flux shall be in accordance with J-STD-004. For Class 3: a) Flux shall conform to flux activity levels L0 or L1 of flux materials rosin (RO), resin (RE), or organic (OR), except organic flux activity level L1 shall not be used for no-clean soldering. b) When other activity levels or flux materials are used, data demonstrating compliance with testing of Appendix B shall be available for review.

Note: Flux or solder paste soldering process combinations previously tested or qualified in accordance with other specifications do not require additional testing.

c) Type H fluxes may be used for tinning of terminals, solid wire and sealed components when performed as part of an integrated fluxing, soldering, cleaning, and cleanliness test system."

� and 4.3 reads:

"Solder paste shall be in accordance with J-STD-005 or equivalent. Solder paste shall also meet the requirements of paragraph 4.1 and 4.2."

First off, to be compliant to J-001C, you have to use fluxes and paste fluxes (solder pastes) that meet the minimum qualification criteria of J-STD-004. Since this is a commercial document, there is no Qualified Products Listing (QPL). You would have to obtain a test report from the flux vendor or an independent laboratory showing the flux to be compliant to J-STD-004, Amendment 1. Most flux vendors should have such a report on file. If not, push them. I would suggest, in the strongest possible terms, that you use only a fluxes or solder pastes qualified to this standard. If the vendor won�t spend the bucks to qualify it, don�t consider it.

In J-STD-004, Amendment 1: * There are four flux types: Rosin (RO), Resin (RE), Organic Acid (OR), and Inorganic Acid (IN). This generally refers to the primary constituent of the solids in the flux. * Fluxes are further classified by activity: Low (L), Medium (M), and High (H), which is an indicator of the aggressiveness of the flux. Low activity fluxes are usually benign, while high activity fluxes are very corrosive. * Fluxes are also classified as halide free (designation 0) or containing halides (designation 1). For example: a common NC flux, such as Kester 951, might have a designation as ROL0, which indicates that the solids are rosin based, low activity, and contains no halides.

So, what does this mean relative to Appendix B? For most manufacturers using NC, it means you don�t have to do Appendix B to be compliant. If you�re using a flux or paste from this list, the material is assumed to be benign enough to be used safely (a somewhat shaky assumption to my way of thinking).

ROL0 - Rosin, Low activity, no halides ROL1 - Rosin, Low activity, some halides REL0 - Resin, Low activity, no halides REL1 - Resin, Low activity, some halides ORL0 - Organic Acid, low activity, no halides. Example: Adipic Acid ORL1 (if you clean it) - Organic Acid, low activity, halide containing.

If the flux you are using is not in the above list, then you have to do Appendix B testing. No way around it. Most water soluble fluxes and many of the RMA/RA fluxes are Medium and High activity levels, so Appendix B testing is required. The higher you go in flux activity, the more dangerous the flux residues become, and the greater the need to demonstrate that you can handle the materials well.

The good news is that if you are going to a NC process, most of the NC fluxes and pastes on the market today fit into the first four designations above.

While you�re ordering standards from IPC, consider TP-1115, "Selection And Implementation Strategy For A Low-Residue, No-Clean Process."

Next, continuing with the processes you listed: your Product Quality & Reliability Department is prudent to steer you towards thinking through the potential impact of converting to NC on those areas. For instance, there was a thread about NC flux residues gumming-up ICT probes on SMTnet within the past six months. Further, there was thread similar to yours about converting to NC on SMTnet within the past six months. There may be some lessons learned in the archives.

Yours sounds like an interesting project. Just remember that the cost benefits of converting to NC will prove to be very elusive. You need to pay close attention to process. Even though it will drive you nuts, it is your friend. NC is a manufacturing engineers full employment product. ;-)

Good luck Dave F

reply »

D.A.Munro

#4422

Re: Converting to 'No-Clean' process | 19 June, 2000

We use both Nc and Ws at wave and reflow, Nc is less forgiving and the board topside must reach aprox.110c prior to wave to be fully activated. Your application of the flux is even more criticale,it's not very agresive for one thing and apply to much and you get to much residue. apply to little a expect poor wetting. Ther is always residue, but if the profile is right it is kept to a minimum. We had excelent results with Qualitek and poor results with Multicore

reply »

Flux-Free Reflow Soldering

Comprehensive Analytical Services and Support