Problem with Solder Paste for DIY Reflow

[dchisholm]

I have started experimenting with a small-scale solder reflow technique for building SMT circuit assemblies in small quantities. (That's anywhere from one-offs, to a batch of a dozen or so copies.) I am encountering problems when working with the solder paste.

I am trying an electric skillet as a reflow oven. This is described in several online forums and blogs, e.g. "Simple Skillet Surface-mount Soldering" at http://www.instructables.com/id/Simple-Skillet-Surface-mount-Soldering/?ALLSTEPS . I have both a bench magnifier, and a low-end toolmaker's microscope ( http://www.amscope.com/stereo-micro...croscopes/20x-stereo-boom-arm-microscope.html ). Last week I purchased a 35 gram syringe of Kester EP256 solder paste ( http://www.kester.com/download/EP256 Solder Paste Data Sheet.pdf ), a standard lead-based solder in a no-clean flux. I am working with common FR-4 circuit boards fabricated by a respected U.S.-based board house. ( I suspect they are actually fabricated in India, but this supplier has a five-year history of providing boards with no known errors or shortcomings.)

Here's my problem:
I am manually applying the solder paste to the circuit pads, then placing the components onto the bed of solder paste, then reflowing the circuit assembly in the electric skillet. I'm having great difficulty getting the solder paste to adhere to the pads on the board. When I put the small dab of solder paste on the board it almost always sticks to the tool I'm using to place the solder paste rather than to the board. I have tried placing the solder paste with wooden toothpicks, small screwdrivers, pointed stainless "dental probe"-like tools, and small bits of plastic-insulated hookup wire - as well as directly from the solder syringe tip. No matter how I try to press, roll, spread, or smear the solder paste onto the circuit pad the result is always the same: when I pull the tool away, all of the solder comes away with the tool rather than sticking to the board.

Several incarnations ago I did a small amount of work with solder paste, doing repairs and rework on previously assembled boards. I don't recall ever having a problem like this. Quite the contrary, it seemed like the solder paste wanted to stick to the board, or at least leave a smeared mark, at the slightest provocation.

Has anybody else observed this behavior? What am I doing wrong?

Dale

 

[JonSnell Electronic]

I use flux as a cream, (Amtech NC-559-ASM-UV and solder balls. Works like a dream if one tins the PCB lands first, if they have never been soldered before.
Temperature is the key. Use hot air or IR as a heat source above the board.

 

[richie00boy]

Is this just one batch of boards? Sounds like faulty plating process.

Also, how old is the paste or has it been opened before? It does go off after just a few months. But gone off paste tends to clump and go hard(er).

 

[dchisholm]

I use flux as a cream, (Amtech NC-559-ASM-UV and solder balls. Works like a dream if one tins the PCB lands first, if they have never been soldered before.
Temperature is the key. Use hot air or IR as a heat source above the board.

My boards already have a hot-air solder-leveled finish (HASL), which - as far as I can tell - is the same as solder tinning.

Is your comment about temperature a reference to the actual reflow step? I know that I must make some adjustments there, but the problem I'm asking for help with is the steps BEFORE I turn on the electric skillet I'm using as a reflow oven. Specifically, I'm having trouble getting the solder paste to stick to the (tinned) lands; and then getting components to stay in place on the lands until I move the board into the skillet.

Dale

 

[dchisholm]

Is this just one batch of boards? Sounds like faulty plating process.

Also, how old is the paste or has it been opened before? It does go off after just a few months. But gone off paste tends to clump and go hard(er).

So far I have experimented with just one batch of boards. They were produced a little over a month ago, and kept in their sealed packaging until I started populating them this week.

The sticker on the paste syringe says it was loaded into the syringe from larger OEM containers. It gives the date of manufacture as Jan 2016. That puts the paste as 3 months into a 6 month shelf life - not as fresh as I'd like, but still well within Kester's guidance.

Dale

 

[AndrewT]

I wonder if the solder pads have a corrosion resistant temporary coating?
That may need to be scrubbed off.
Try dish washing detergent and a toothbrush.
Followed by a thorough rise in clean water and a final wipe with an alcohol dipped cotton bud.

 

[richie00boy]

Usually such coatings are applied to bare copper.

To confirm if it is the board rather than the paste, try tinning some pads with an iron and solder wire. I reckon faulty plating has led to an intermetallic alloy formation.

 

[tonitonitoni]

i am curious, when paste liquifies then solidifies inside syringe due to heat from outside , does it loose its properties?

 

[AndrewT]

Only changes properties if some of the volatiles evaporate and escape from the syringe.

But shelf life indicates that there is some longer term change that is possible.

 

[marce]

Generally solder paste will be kept in a fridge until required once a package is opended in any way.
HASL is not the best finish for SMD it has a meniscus, but as said it should match the solder. I would suspect though that it is lead free HASL unless you have specifically requested a lead finish, as leaded solder breaks the rohs rules...
It is best to screen the paste then it will stick to the pad, in your case the pad is slippier than the application tool so the paste adheres to the tool, some plastics are very slippery and if you can make a very small trowel you could try scraping the paste onto the pad. Try an ENID finish if the cost is not to great, better adhesion and a flatter surface for the SMD pads.

 

[Adam Reed]

Dale I've been using a similar technique for a while now.
I like to use a low temp solder from Chipquick, makes this process a lot easier.
The material is PN SMDLTLFP. Available from DigiKey, Mouser, Amazon, etc.
It also helps a LOT to have the board slightly warm and to scrub with a toothbrush and a dilute dish soap solution.

And do not hesitate to use both heat from below and heat gun from above (carefully)

 

[dchisholm]

I wonder if the solder pads have a corrosion resistant temporary coating?
That may need to be scrubbed off. . . .

That was one of my first thoughts. Or, a thin film of corrosion had formed on top of the HASL finish during the month since the boards were manufactured, even though they were kept in the sealed plastic wrap as shipped from the board house.

I buffed the surface with about half a dozen strokes of a Scotch-Brite pad, followed by scrubbing with a cotton swab dipped in isopropyl alcohol, and air-dried.

It certainly didn't degrade the adhesion of the raw solder paste to the pads, but any improvement was barely noticeable. (And the abrasive pad scuffed the glossy surface of the solder mask.)

Dale

 

[dchisholm]

To confirm if it is the board rather than the paste, try tinning some pads with an iron and solder wire. I reckon faulty plating has led to an intermetallic alloy formation.

The HASL finish on the exposed pads accept solder - either flux-core 63/37 wire solder, or the Kester EP256 paste - as easily as any I can recall, and probably easier than most. As soon as the solder melts, it flows evenly and self-levels on cooling.

Dale

 

[dchisholm]

i am curious, when paste liquifies then solidifies inside syringe due to heat from outside , does it loose its properties?

Only changes properties if some of the volatiles evaporate and escape from the syringe.

But shelf life indicates that there is some longer term change that is possible.

When I was reading up on this process before trying it I came across a web page which essentially said that all of the warnings and instructions about storing and handling solder paste, shelf life, temperature profiles, etc, are MUCH less critical for bench-top assembly of small batches of boards, than for similar assemblies mass-produced on automated lines. As I recall, they found that shelf life was at least a year when refrigerated.

(I thought the page was on the SparkFun site, but I can't find it right now.)

Dale

 

[AndrewT]

All my solders are way over a year old.
Some of my solder paste is over 30years old. They all work.

 

[dchisholm]

. . . HASL is not the best finish for SMD it has a meniscus, but as said it should match the solder. I would suspect though that it is lead free HASL unless you have specifically requested a lead finish, as leaded solder breaks the rohs rules...

I don't know whether my HASL is leaded or lead-free, but I suspect it's leaded.

The company I'm doing this work for is still using leaded solder exclusively. (On the day I first showed up for work I asked whether solder was leaded or lead-free. My foreman didn't know the difference, but a quick glance at some labels showed it was definitely leaded solder.)

They also insist on getting the absolute cheapest version of any production part. The PWB drawings I produced had a fab note stating "Lead-free materials preferred but leaded materials acceptable." The drawing that went to the board house had that note removed, and the boards received from the board house did NOT have any markings indicating lead-free construction.

It is best to screen the paste then it will stick to the pad, in your case the pad is slippier than the application tool so the paste adheres to the tool, some plastics are very slippery and if you can make a very small trowel you could try scraping the paste onto the pad.

By "screen" are you referring to what we call a "stencil" on this side of the pond? At this time I have instructions to produce about two-dozen units for evaluation, field tests by preferred customers, and demonstrations at trade shows. I don't have authorization to spend US$100 (est) for a set of temporary (plastic) stencils and related tools. If demand lives up to the expectations of the company president , the whole assembly process will go to a contract assembler. (And, the company president will be driving a new BMW. I doubt that we'll sell enough to get him a used Kia.)

Results have improved over the week that I've experimented with this process and materials. The solder paste coming out of the syringe seems to be a little stickier. I have no way to really measure this, though I can accept the theory that the paste at the very end of the syringe had become a little dried out; or solder and flux started to separate due to how the syringe had been positioned in storage. As I dispensed the material from the very end of the syringe, and started using paste located higher up in the syringe, the paste may have been stickier.

Or maybe I just developed a more effective technique as I worked with the stuff. I eventually decided that a pair of wooden toothpicks made the most effective "tool" for applying paste to pads. I found that a non-obvious "rolling" motion was the best way to transfer the paste from toothpick to pad, with the second toothpick used to separate the paste from the first toothpick. (It's easier to do, than to describe.)

Dale

 

[JonSnell Electronic]

My NC559 paste is five years old, the expiry date is Jan 2013 and I have no issues with Xbox 360 or GPU re-balling. Mind you I am working with pre tinned lands.

 

[kamis]

I solved a similar problem by mixing gradually a small amount of rosin based flux paste with dried solder paste up to desired level of tack ability.

 

[Neil Davis]

I solved a similar problem by mixing gradually a small amount of rosin based flux paste with dried solder paste up to desired level of tack ability.

...or maybe swipe the board with a flux pen prior to dabbing on the paste with the syringe. I've never had a problem applying directly from the syringe to the pads, but the solder I'm using is fairly thick and gooey.

Here's two boards I did last week: