Tuesday, 11 June 2013

Stencil Success!

So I'd managed to knock up a pair of solder paste stencils, and now it's time for the moment of truth:  can I actually manage to lay down pastes and successfully solder some parts on?

Plastic Not-So Fantastic

In my last past, I made mention that I'd only intended to test the plastic stencil if the metal one failed.  Well, I got in-touch with my inner feminist and decided to start with the plastic stencil after all.  Why?  Well, the plastic sat flush with the boards, and I was thinking I'd probably not need the top frame in my Jig.

Sitting Pretty

But Before that...

The jig I'd built last time had a few issues - I only had the one tooling pin at the time, and now that I have the three I planned, I found out that my holes were too tight.  I couldn't slide the Jig frame over the pins - I could tap them in place, and they would hold really tight, but what's the point of that if you want  to actually separate the layers?

So V2 was quickly spun out, with the holes enlarged by 0.1mm.  Love that mill!

V2 Milling

Base Plate and Three Tooling Pins

A lesson learnt here - I'd overdone it with the mounting holes.  I was thinking that I' need to screw the whole assembly together with just the top frame only located by the tooling holes.  But even with the 3.1mm holes and 3.0mm pins, it still sits together well - the extra mounting screws are not needed, so if anyone copies my method, feel free to skip these :)

Mid Plate

Tin Can Stencil and Frame

Plastic Stencil and Frame

Note that I'm reusing the frame for both the Tin Can and Plastic Stencils.  I was planning to glue the stencils to individual frames, but after this test I don't think it's warranted for small build numbers.  In fact the frame was still sticking even with 3.1mm holes, so I might take to the frame lightly with a rat tail file loosen things up a bit, but I'll need to experiment with steeping up the holes in 0.05mm  increments, until I'm happy with the fit, for future builds.

Nice Fit!

One thing I did fluke, was how good a fit the frame is to the plastic spatula I had lying around in the garage  (I have small kids, I'm forever patching plasterboard in the house....)

The Bends

Before getting started, the above shot shows how my metal stencil was still curled - when you use a round can as stock it's an issue - and this combined with my 'gut feeling' that 0.1mm was going to be too thin, I decided to use the plastic stencil for my first attempt.

First Time for Everything

So I loaded the jig with a pair of boards, and crossed my fingers.

Plastic Ready to Rock

Its hard to tell in the above pic, but I'd fitted the plastic stencil and was ready to start with the paste.

Only the Best for Me :)

The solder paste I got from eBay, and at $3.50 for a little pot, I probably won't bother trying to keep this in the fridge once opened, but rather just get a new pot for each run of projects.  Love eBay! 

Starting Small

Remembering that the plastic sheet is only 0.25mm thick, I started small.  I used a zip tie as a paste stick, and smeared a small amount of paste at one end of the stencil.

Too Little

After my first attempt at spreading the paste, I knew I had too little.  So I loaded up some more paste and took a second pass.

More Paste


And from the initial inspection, it looked like I'd nailed it.  So I tentatively separated the frame and stencil, and on closer inspection....

Lots and Lots!

... it was pretty obvious that I had way too much solder.  At only a quarter of a millimeter thick, the pads managed to stand out like Mount Everest, and with small clearances adjacent solder paste pads smeared into each other.  

Thin is Best

So, that's it for my easy to use plastic stencil.  Time to get cracking with the Tin Can version.

Cutting Corners

As the Tin Can stencil was still a bit curly, I decided to try stenciling by holding it in place with my hand.  I guess that this would be easier than using the jig,  You know, the whole reason I made the jig was to clamp the stencil down, but my brain fart decided to go free form on this. Held down with the left hand I added some paste, took a swipe, lifted and took at good look at the result.


Argh, it was worse than the plastic stencil result.  You can see that the pads are smeared, so it was obvious that just holding the stencil in place by hand was not enough to stop it squirming around when working the spatula.


So after cleaning everything (with isoproply alcohol) I loaded the stencil in the frame, applied a little paste and swiped.

Fail. Again.

Okay, this time it wasn't a wobbly hand letting me down here.  What I think I did wrong was to try and make the paste spread as far as possible.  I used a similar amount as per the plastic stencil, but on my first pass not all of the stencil holes were filled.  To 'fix' this I took a few extra swipes to move the paste around and this combined with minor buckles in the stencil, seems to have forced paste under the stencil.  So that didn't work either. 


Not to give up, I took another tact.  This time I spread extra paste gently over the stencil - using the end of a zip tie and carefully not applying any pressure to the stencil.  Only then, when I was happy that a full covering of past was present, did I swipe the spatula, and this time only once in the one direction.

Single Pass

Not shown was the fact that a heap of paste stuck to the spatula. Once done, I carefully lifted the stencil and frame, and half wishing, took a peep at the result.


Camera Flash Off

The two pics above (one with flash on, the other off) show the paste on the test board.  You can see the separation between the larger 2 pad device and the 4 pin device is quite small, and the paste is separate. There are a few pads on my 0603 packages (to the rear of the board) where the paste bridges the gaps, but I went with a punt that when flowed, the solder would separate - got to love soldermask.

Test Board One

Test Board Two

Let's Cook

Before jumping right in, as I only have enough correct parts to complete a limited set of this design, and that thanks to Seeed I have quite a few extra boards, I decided to load some 1k 0603's (I have heaps of these) and try out the new toy in the workshop lab:

Reflow Oven.  Woot.

You'll note that I was pretty hap-hazard at how I placed the resistors above.  I'd read before that when molten the surface tension of the pads should pull the parts into line, and I wanted to test that theory.  I also loaded a few parts in the board I stenciled using the plastic stencil, just in case I was wrong about the amount of solder needed.

Ready to Roast.

I loaded the boards into the oven, and originally I was thinking that the wet paste might allow the parts to slide around, but to my surprise the paste is quite sticky and the parts stayed put.  I even inverted a board to check - I won't try that with large package parts - and everything stayed put.  Sweet.

Default.  FTW. 

So I took the oven on its maiden run, with the first default profile.  In only 7 1/2 short minutes it was time to pull the boards from the oven.


Half Baked comes to mind.  Yep, the default profile didn't seem to heat the board evenly, and only the smaller parts seemed to reflow.  The larger pads didn't melt, and neither did any of the pads from the plastic stencil.  So that profile didn't seem to apply enough heat.  

Half Baked.

Raw Dough.

So it seemed to be an easy enough heat - pick a profile with more heat!  Option 6 had the most area under the curve, and I think it's got a higher peak temperature that the default option one as well. 

Turning Up the Heat

Before getting too serious when using this oven I'm going to have to read up a few datasheets to get an idea of the temperature profile I should be using, but for now as I'm only experimenting with resistors, this will do.  Also, as some of the pads had already flowed, I added some liquid flux to those pads, just in case.


Tin Can Stencil

So, another 7 1/2 agonizing minutes later, the following shots below show my first successful reflow boards!  Whoo Hoo!  Apologies for the crappiness of the shots, I'm not camera wiz.  But a tripod is not on my shopping list.

Yes, It Worked!

Look Carefully....

What I thin is impressive is that the solder has flowed well, and formed nice fillets on the pads. I was amazed at how much solder there is when using the 0.1mm stencil, but when you look at a pad, and then realise that the device you are soldering into place displaces most of the surface area, you soon realise where all the solder comes from.

And the parts were all pulled into nice alignment.  Yay Physics!

Plastic Stencil

To compare, here's some beauty shots of the board from the plastic stencil.

Blobs and Blobs

Extra Excess

Even though I proclaim that I can hand solder small parts, compared to the brilliant techs I work with, my hand soldering is usually on the heavy side, and I use a lot of solder-wick cleaning up my joints.  And even thought I'm proclaiming that the plastic stencil also has excess solder in the joints, it's still better than any of my hand soldering efforts.  The fact that the tin can stencil can produce an even better result is a bonus, and they are never hard to find.  So my future stencils will be can based. 

The only downside is the maximum usable area I'm getting from a typical can is 10 x 10 cm - perfect when using Seeed services, but if larger is needed, I might have to switch to imported beers to get larger cans.

Side by Side

And for comparison, the plastic stencil board is on the left, tin can version on the right.

What's Next

The parts for this design should be here soon so I'll assemble a pair for testing and spill the beans on this board's purpose once finished.  And I might even get the time to make a stencil so I can finally assemble this.

No comments:

Post a Comment