Solder Paste expiration and optimising hand soldering SMT

[ramonest]

Hello, I've started working in PCB design and assembly.

I do the layout and the component assembly. The PCB itself while in prototyping phase is manufactured "in-house" by other technicians. They use substractive method with a CNC milling machine. There are good things and bad things on doing it in-house. We get to do relatively fast and cheap prototyping.
Drawbacks:

• Unreliable vias (I use thin wire or resistor leads if the trace has to carry power)
• No solder mask
• The tolerance on the milling tool makes it so that we can't have really thin traces, and pads get thinner than designed

Any recommendations to mitigate the problems that this drawbacks create? I think that when I bridged to legs of the smallest component when reflowing was do to the lack of solder mask and the pads being too thin and taking too little solder on them so it went upwards on the legs.

We'll need to do at least 20 PCBs and hopefully when we finish prototyping and get the design right we'll send it to manufacture outside, so it'll be proper professional PCB.

The PCB is SMT and double sided. Passives down to 0603(imperial) and the smallest IC has a 0.6mm pin pitch (TPS6212x). I can solder all of them by hand using soldering iron, but I hoped to get better results and for the assembly to be faster with reflow, be it using a hot air station or a reflow oven (we've one).
I looked into it and in some places i found that you can actually reflow the board two times (one for each side) and if the components are not too heavy they won't fall when you reflow the other side. Does that actually work or is it unreliable? And is it the best way to solder them or is there a better way for what I've to do? I guess when we have them made professionally we could even get the stencils for faster solder paste application (as Dave shows on EEVblog #415).


While I was soldering and testing parts of the prototype one prototype board, apart from the legs getting a solder bridge (as i explained earlier) I also found another problem while reflowing (with the hot air station). Some pads were already properly reflowed and had a nice solder joint but others even after a much longer time under the hot air didn't reflow properly. The solder was "soft" i could deform it with the tweezers.
Could it be due to the solder paste being expired (1 year expired or so), or would it be more likely due to me doing something bad? Maybe they had bigger thermal mass? But I really stayed quite longer on those after some already reflowed... I think that all we use is lead free, if that makes any difference.

Where can I get proper information whether solder paste goes bad or not and how to identify it?

Thanks much in advance.

[tautech]

Yeah paste goes off and while you might choose to freshen it for hobby work, for manufacturing ....nah.
It should have an expiry date and be kept in the fridge to get the longest usable life from it.

Have a look at member paulca's hot air reflow:
https://www.eevblog.com/forum/beginners/so-surface-mount-it-is/msg1442590/#msg1442590

[james_s]

Solder paste expires in a fairly short period.

For hobby work though you can just add a bit of fresh liquid flux and that brings it back. Mine has been expired for years and it still works fine. I wouldn't do that for commercial production though.

[mrpackethead]

Seriously, i'd say ditch the milled pcbs, and all the limitations and issues that come with them.   You can get low cost pcbs in a couple of days, compelte with proper stencils for nothign.     been there done that, and its just not worth the drama.

The easiest paste i've ever worked wiht and most forgiving is Henkel GC10.

[Kjelt]

Any recommendations to mitigate the problems that this drawbacks create? I think that when I bridged to legs of the smallest component when reflowing
was do to the lack of solder mask and the pads being too thin and taking too little solder on them so it went upwards on the legs.

Do you clean the pcb? Little rub with very fine 3M scratch pad and good clean with IPA ?
I even get bridges with commercial boards with solder mask when the pitch is too small. Mostly too much paste (stencil problem or squeegee pressed too hard).
My solution only put paste on a few outside pins of the component before reflowing, this will make it stick in the correct position.
After reflowing handsolder each pin with solder paste. Put a tiny string of paste on the pad and a bit on the leg, put the small solderiron to the leg of the component while pressing a tiny bit and in one/two seconds your done per pin. Skip each pin for thermal issues, or let it cool off for ten seconds before going to the next pin. Visually inspect each pin under the stereo microscope while holding it under a corner.

[KL27x]

My solution only put paste on a few outside pins of the component before reflowing, this will make it stick in the correct position.
After reflowing handsolder each pin with solder paste. Put a tiny string of paste on the pad and a bit on the leg, put the small solderiron to the leg of the component while pressing a tiny bit and in one/two seconds your done per pin. Skip each pin for thermal issues, or let it cool off for ten seconds before going to the next pin. Visually inspect each pin under the stereo microscope while holding it under a corner.

That's... interesting. My take was to remove these fine pitch gullwing parts from the past layer/stencil and drag solder them with a big fatty iron after the reflow in one swipe per row of legs.

FPC connectors. Now there's a real PITA.

[mrpackethead]

All of the above pretty much goes away if you use a decent stencil on pcbs that are properly made.

My solution only put paste on a few outside pins of the component before reflowing, this will make it stick in the correct position.
After reflowing handsolder each pin with solder paste. Put a tiny string of paste on the pad and a bit on the leg, put the small solderiron to the leg of the component while pressing a tiny bit and in one/two seconds your done per pin. Skip each pin for thermal issues, or let it cool off for ten seconds before going to the next pin. Visually inspect each pin under the stereo microscope while holding it under a corner.

That's... interesting. My take was to remove these fine pitch gullwing parts from the past layer/stencil and drag solder them with a big fatty iron after the reflow in one swipe per row of legs.

FPC connectors. Now there's a real PITA.

[Kjelt]

That's... interesting. My take was to remove these fine pitch gullwing parts from the past layer/stencil and drag solder them with a big fatty iron after the reflow in one swipe per row of legs.

Yes that is another possibility but personally I leave that to the showboat youtube vbloggers 
It does work, but if you're too slow or unexperienced you are left with a flux floating thermally destroyed component.
You only heat one side of the component and all pins at once.
I guess that will give more thermal stress inside.
Reflowing is heating the entire component gradually.
Pin soldering is heating a small part of the component and leaving it too cool off again before the heat goes inside.
So personally I rather go safe and see what I am doing, pin by pin. But perhaps you solution is better. 

All of the above pretty much goes away if you use a decent stencil on pcbs that are properly made.

Even happened on stainless steel stencils and boards from an expensive european manufacturer such as eurocircuits.
And it is not their fault. Just that some pads are not implemented correctly so the stencil space is just too large, eg too much paste, or you press too hard or sweep too slow. These only occur with the very small pitch components, normal pitch components are zero problem.
I often had to do a redesign of the stencil in order to make smaller spaces because the Eagle pads were not ok.

[mrpackethead]

All of the above pretty much goes away if you use a decent stencil on pcbs that are properly made.

Even happened on stainless steel stencils and boards from an expensive european manufacturer such as eurocircuits.
And it is not their fault. Just that some pads are not implemented correctly so the stencil space is just too large, eg too much paste, or you press too hard or sweep too slow. These only occur with the very small pitch components, normal pitch components are zero problem.
I often had to do a redesign of the stencil in order to make smaller spaces because the Eagle pads were not ok.

Yes, all that! i shoudl not have spelled it out.  I spend a LOT of time looking at solder apertures to make sure they are optimised.  It makes a massive difference.     Decent stencil = properly designed apatures, correct thickness,  stencil in tensioned frame, or under tension in a proper frame.  It also needed to be electropolished.     Applciation technique is also really important, along with having solder paste that is still good.

[tautech]

All of the above pretty much goes away if you use a decent stencil on pcbs that are properly made.

Even happened on stainless steel stencils and boards from an expensive european manufacturer such as eurocircuits.
And it is not their fault. Just that some pads are not implemented correctly so the stencil space is just too large, eg too much paste, or you press too hard or sweep too slow. These only occur with the very small pitch components, normal pitch components are zero problem.
I often had to do a redesign of the stencil in order to make smaller spaces because the Eagle pads were not ok.

Yes, all that i shoudl not have implied.  I spend a LOT of time lookign at solder apatures to make sure they are optimised.  It makes a massive difference.     Decent stencil = properly designed apatures, correct thickness,  stencil in tensioned frame, or under tension in a proper frame.  It also needed to be electropolished.     Applciation technique is also really important, along with having solder paste that is still good.
[/quote]
One I saw recently was placed on a milled steel frame from 1.6mm panel steel so to be the same thickness as normal FR4 and the milled relief was the perfect fit for the PCB.

[CatalinaWOW]

My experience with expired paste is that it solders just fine.  If you get the right amount of paste down.  When hand applying with a syringe this isn't hard with fresh paste, but very difficult with expired paste.  I don't know what about the chemistry changes, but it doesn't stick to the pads the same way fresh paste does.  I haven't tried freshening it with new flux since it already seems too runny.  Adding more than a couple of percent seems like it would make it more liquid than paste.

The same problem applies when using masks, but anyone using masks is not going to cheap out and not use fresh paste.

[mrpackethead]

If you are new to using stencils, and only doing small volumes, i would highly recommedn Henkel GC10 paste.  Its very very forgiving, can be stored at room temp, and has a long expiry.   It is easy to apply and cleanup.  it is about twice the price of other pastes, but in the big scheme of thigns, its not an expensive part of your set up.

I do find it doe'stn work well in a syringe though.   Foruantly i do'nt do much of that, as stencils are so cheap.