Our pallet manufacturer recommends titanium inserts to allow larger apertures near closely spaced surface mount components. Are they worth it? The Assembly Brothers, Jim Hall and Phil Zarrow, share their insights. Board Talk
Board Talk is presented by Phil Zarrow and Jim Hall of ITM Consulting.
Process Troubleshooting, Failure Analysis, Process Audits, Process Set-up CEM Selection/Qualification, SMT Training/Seminars, Legal Disputes
With over 35 years experience in PCB assembly, Phil is one of the leading experts in SMT process failure analysis.
He has vast experience in SMT equipment, materials and processes.
A Lean Six-Sigma Master Blackbelt, Jim has a wealth of knowledge in soldering, thermal technology, equipment and process basics.
He is a pioneer in the science of reflow.
And welcome to Board Talk. You've got Jim Hall and Phil Zarrow, the Assembly Brothers. So what is today's question?
It comes from H.Z. We use selective wave soldering with aperture pallet. Our new designs have tighter spacing which are causing problems with hole fill. Our pallet manufacturer recommends titanium inserts to allow larger apertures near the closely spaced SMDs. They are expensive. Are they worth it?
Let's just put it this way; wave solder has been around for 60, 70 years now. And it worked great with the through-hole days, but as we approached surface --
And a 100mil pitch with no surface mount parts obstructing the bottom of the board.
And ever since surface mount appeared on the screen here basically we've been trying to adapt it in every way, duct tape, spit, scotch tape.
Peanut oil, spitting in the flux, air knives, nitrogen, the vibrating nozzles. So obviously one of the more, I would say, successful adaptations has been the use of the pallets, the masking pallet fixtures for adapting your wave solder for doing, if you will, a variation of selective soldering. And that's what H.Z. is talking about where you are masking off the previously reflowed surface map components and trying to solder just through-hole remaining areas and mix technology.
Surface mount parts on the bottom side, just wave solder side of the board that you can't glue and wave solder such as QFNs or QFPs and things like that.
It does work, but it is fraught with problems and defect generation. A lot of people try to build their own apertures. There is a lot of engineering involved in terms how close the edge you can locate these things.
What kind of aperture you can have. The density of the lead through you have. The bevel on the edge. It is as simple as it might appear and it is not as air proof as it might appear.
But underlying it all is basic physics. It's just like stencil printing. You've got a big aperture, it's easy to get paste to flow in and out of it, to fill and release.
If you got a big aperture around your through-hole joints it's easy for the wave solder to get the flux, the preheat and the flow from the wave up into the aperture.
But as the aperture gets smaller, it becomes more difficult and more error prone as my brother has indicated. What drives that, well, the close spacing of your SMDs relative to your through-holes. In the old days when things were spread out, your apertures were big, you have plenty of room.
So now you got a surface mount part that's really close, you still have to have the mask cover the surface mount part. You have limitations in the Delrin high temperature plastic materials that they're using for the pallet.
So the answer is, make the wall around these tight tolerances out of a little piece of titanium and set it into the fixture, so we can make that that wall between the aperture for the through-hole and the covered up surface mount part as thin as possible. And that's the strategy. And the answer is, yeah, it can work up to a limit. You know, you can buy yourself a little more aperture opening and for some things it will be successful, but --
Don't fight a fact, deal with it, you're basically doing an adaptation of an adaptation. Don't fight the fact, deal with it, face the music, and consider using a process ...
A process more designed to deal with this problem. We are talking about point to point or multi wave or fountain selective soldering.
And that's assuming you can't do intrusive soldering going to this newer concept.
Which allows you to customize the process on a point-to-point level down to a solder joint. And what's the ultimate fallback? Solder it by hand, you get a real pro in there with a tiny little soldering iron and they can get the heat and the soldering flux from the cored wire in a tight space without damaging the adjacent surface mount.
Well, that's what the selective soldering is designed to do and it's why it is the fastest growing technology in our industry because things are getting tighter, spacings are getting closer and closer. And it's putting more and more strains on wave.
So yes, maybe for HZ, maybe for your application you might be able to get away with titanium inserts depending on the ultimate spacing, all the characteristics of the joints, how difficult it is to get the flux, preheat and solder into that aperture.
But, long term things aren't getting better, spacings are getting tighter. We feel that most assemblers should be considering moving to some form of controllable selective soldering to give you a much greater process window.
So look to the future, HZ. And on that note, let's say, we hope we answered your question.
Don't solder like my brother.
And don't solder like my brother.
In the 'thru-hole only' days, each board wavesolder profile, utilized an algorithm composed of "number of layers; density of the ground plane; population density; barrel size; top/bottom preheater temp; conveyor speed; wave height; fluxer/air knife profile; proprietary calculations necessary to handle the boards as they exited the machine, to prevent mortality due to physically shocking the boards".
When possible, we ordered boards from the board house, to be pre-palletized. The benefits of easier handling, an exponentially lower loss/scrap rate, and boards which required far less touch-up, far outweighed the additional cost, and board material wasted during the depalletization step in the process/procedure.
I can't tell you how many times boards which were well-locked into the pallet fixture, floated out of the fixture while traveling over the wave. I once noticed the wavesolder operator washing Â¾-inch stainless nuts, which she placed atop 2.5-inch round display driver boards to help stop them from floating out of the fixture.
My advice is this: ask your wavesolder operators for their opinions/preferences in answer to the problem. Translating their "gut instincts" with regard to the design, will most likely prove beneficial.
Tim Hufnell, H. F. Engineering Systems of PA Llc.
1st off, well done Phil & Jim! This is a very challenging topic to discuss in an open forum.
I agree with Edward, it all boils down to the design of the circuit board, otherwise known as DFM, "Designed For Manufacturability". Many years ago when I started in this Solder Pallet industry, about 90% of the folks I dealt with on a daily basis were PCB designers, the people that laid out the card and prepared to make a million copies of the same thing, step & repeat had to be their, parts needed to be held flat, bowing of the pcb due to heat was taken under consideration, sensitive SMD's were placed on the top surface and away from PTH barrels, and on and on it goes!
In most cases, CM's cannot control DFM, they can influence, but eventually they get what they get! Get your tooling people involved early in the MFG process, before you take on the project. Always keep in the back of your mind that TITANIUM CAN BUY SPACING.
Eric Miller, Stentech
DFM!! Engineers and layout techs often ignore the realities and hardships of manufacturing. Do a design review, I'll bet more often than not the problem can be designed away.
Edward M Llorca, E2M Logistics
All true* - the process window for selective wave solder is 1/4th the window of a easily solderable PCBA. Aperture balance between small and large apertures is a tool design challenge. The process is dependent on flux application, even top side pre heat, controllable wave parameters/conveyors & PATIENCE.
* NOTE - Reading public - Do not use Delrin/Acetal for your wave pallets unless you want to fill your shop with acrid smoke. The materials are Durastone/Durapol and a few others, glass filled composite.
Its an ROI analysis - Alternate process/post wave touch up, etc VS pallet cost divided by PCBA volume.
David Duke, Stone Mountain Tool
There is one wave pallet manufacturer (Stentech) in the US that offers wave solder pallets made out of a different material called Ricocel. This material is very strong and can be machined down to wall thicknesses of .010". This has eliminated the expensive titanium inserts in all of our pallets.
Yes, titanium inserts are invaluable when they are truly required, however, we rarely use them on our wave solder. If we must go to a titanium insert than the physics of soldering are what is pushing us to do so. Going to Titanium may solve some issues, but normally, if it is required for a traditional wave solder, then selective soldering will almost always be a better solution. We use extremely complicated titanium-insert based selective fixtures with our selective soldering machines with fantastic results. For each of the assemblies, I would not even consider attempting them across our traditional wave.
Andrew Williams, CSMTPE, PRIDE Industries
Yes! They are most definitely worth it. In the past couple years I had to educate both our EMS shops about using Titanium Inserts in Selective Wave Solder Pallets. We had a PCBA at each shop with layouts that had a few THT pins to close to some SMD devices and the yield was terrible with the standard Pallet material. The standard material needs a thick wall for 2 reasons, the initial machining of it and the long term durability. However, a "Titanium Insert" can be machined to a very thin wall and have basically infinite durability.
For both our PCBAs at the 2 suppliers, once we added the Titanium Insert the yield issue was totally eliminated. This is because the thinner walls in the Titanium Insert allowed the opening for the THT pins to become a recommended proper width. The Titanium Inserts we used were only about 3cm x 3cm in area and we used 2 short Flat head Stainless Steel Screws to hold the Insert into the standard pallet material. These Titanium Inserts were in the low hundreds of dollars and are reusable for all future Pallets. When the rest of the standard material Pallet degrades and needs to be replaced, you just remove the Titanium Insert and reuse it in a new Pallet. So, it is just a one time cost of a few hundred dollars. They are one of the most effective and lowest cost solutions in the PCBA manufacturing world!