Heat Gun/Solder PAste SMD soldering

[stevie1556]

I just use a cheap Chinese paste for my stuff and find it works great, although it's only in a pot and I need to use a small screwdriver to apply some to the pads, and use a hot air soldering gun with the air flow set on a very low setting, which seems to work really well. If you do use the heat gun route, then be careful as the air flow maybe too powerful and blow the components off the board.

There was a video on YouTube of someone using an electric skillet for it, but I can't seem to find it anywhere.

However, you can get some great results with the paste and heat gun method.....

[pickdropper]

If I am going to be using a new PCB... and clean the pads... Would I need to use flux on each pad prior to laying down the solder paste OR can I just put the paste down on the pads and start placing components?

If the paste is new, you probably don't as flux is present within the paste.  I'd recommend spending a bit of time reading here:

http://www.kester.com/knowledge-base/

[Govmnt_Lacky]

Cheers gents! Thanks for the inputs.

I am going to try the skillet route however, I still have reservations whether the PCB will stand up. Particularly the traces and vias 

[wgc]

You should be fine.  PCBs usually go through an oven in production assembly processes, often twice. 

You CAN burn a pcb but that takes a much higher temp than melting solder.

Also, watch out for solderballs, especially underneath components.  Try to keep as much of the solder paste off the solder mask as possible.

[pickdropper]

You should be fine.  PCBs usually go through an oven in production assembly processes, often twice. 

You CAN burn a pcb but that takes a much higher temp than melting solder.

Also, watch out for solderballs, especially underneath components.  Try to keep as much of the solder paste off the solder mask as possible.

This is basically true, with the exception of the occasional PCB with a really crappy solder mask.

[rullywowr]

If I am going to be using a new PCB... and clean the pads... Would I need to use flux on each pad prior to laying down the solder paste OR can I just put the paste down on the pads and start placing components?

While it will probably work without it, as there is flux in the solder...a little extra flux can never hurt!  I like to use a Kester flux pen with "no-clean" flux.  Either 951 or 952 type work well.  Griddle it up!

[Govmnt_Lacky]

While it will probably work without it, as there is flux in the solder...a little extra flux can never hurt!  I like to use a Kester flux pen with "no-clean" flux.  Either 951 or 952 type work well.  Griddle it up!

Yep! Getting close to obtaining all of my parts I need. Just trying to source a good candidate for the hot plate 

[copachino]

https://www.youtube.com/watch?v=0XENpPtisnM

this looks quite easy

[alanp]

From Muffwiggler's MI Braids thread:

http://www.muffwiggler.com/forum/viewtopic.php?p=1771205#1771205

Three problems with solder paste and a reflow oven or hot plate:

a) reflow techniques are intended for high volume industrial production, where dozens of test boards will be thrown away while establishing just the right temperature parameters and curves for a particular board and set of components. With 0603 parts and smaller, I am told that "tombstoning" can be tricky to avoid without high-end reflow ovens that guarantee completely uniform temperature at all times. There is a risk you will ruin all three of your boards just getting the reflow process right.

b) unless you have a solder paste stencil, applying the paste by hand and then positioning components by hand will take as long as soldering directly by hand. Even with a stencil, you still need to use tweezers to place all the components by hand. And the solder paste tends to obscure the PCB silkscreening, so you need to refer to a reference diagram or another identical board to make sure you are putting the right part in the right place. And you need to do all that in one sitting, because the paste dries out if left for too long.

c) because 0603 passives are unlabelled, you have to be super careful not to mix them up. Easy enough when soldering just one value at a time by hand, but with reflow, you need to place all the values on the board all at once. Can be done, but you need to be super careful - again, the proportion time probably outweighs the time saved in soldering all bits at once.

And then you still need to solder all the bits containing plastic by hand. Doesn't seem sensible to me. But I'd be very interested in hearing how you go, because I have a whole stack of SMD boards yet to solder, and doing them by hand certainly takes a long time.

I want the SMD guru's opinions on this -- the Mutable Instrument boards use pretty much ALL 0603 for the passives, SMD opamps and DACs and stuffs, and a CPU chip with pins on all four sides that are really weeny.

[wgc]

I want the SMD guru's opinions on this -- the Mutable Instrument boards use pretty much ALL 0603 for the passives, SMD opamps and DACs and stuffs, and a CPU chip with pins on all four sides that are really weeny.

I prefer to think of myself as an eternal student with some deep experience, not a guru, and I hope my comments can be helpful.  Good news is that "notify of new posts" thing works great.

I don't disagree with some of the points, but there are other factors at play which, having been overlooked could also cause some issues. 

To say that your boards will be ruined is probably not true, you can rework, it won't be that much fun, but it can be done, likely easier than through hole if you had redo one of those boards to the same extent.

Three problems with solder paste and a reflow oven or hot plate:

a) reflow techniques are intended for high volume industrial production, where dozens of test boards will be thrown away while establishing just the right temperature parameters and curves for a particular board and set of components. With 0603 parts and smaller, I am told that "tombstoning" can be tricky to avoid without high-end reflow ovens that guarantee completely uniform temperature at all times. There is a risk you will ruin all three of your boards just getting the reflow process right.

This is true to an extent. You don't start to see it much until you get to 0402, where it happens sometimes, and 0201, where it can happen a lot.  Reflow efficiency can have an effect but I wouldn't expect much on a double sided pcb.  It shows up a lot more where you have lots more layers and buried ground pours.  Easy enough to fix by reprofiling.  It only takes 2-3 pcbs for someone who understands how to do it, and usually can be fixed, once identified by adjusting zones, conveyor speed, nitrogen environment, or crazy as it sounds, even inserting the pcb into the oven in a different orientation.

More often in my past experience, tombstoning happens when the part placement is off.  With older p/p equipment, smaller parts are pushing the limits of the machine and placement can be slightly off.  In this case, surface tension of the liquid solder will pull on one side more than the other, and the part will tombstone.  There is/was a stencil aperture design called "home plate" used for avoiding solderballs under passives.  This could also contribute to tombstoning related to placement issues since your placement target for getting both side of the part into an equal amount of paste is reduced.  So how the paste is applied could cause some issues, though again, this is more for 0402 and 0201 than 0603.

b) unless you have a solder paste stencil, applying the paste by hand and then positioning components by hand will take as long as soldering directly by hand. Even with a stencil, you still need to use tweezers to place all the components by hand. And the solder paste tends to obscure the PCB silkscreening, so you need to refer to a reference diagram or another identical board to make sure you are putting the right part in the right place. And you need to do all that in one sitting, because the paste dries out if left for too long.

See my comments for a) above.  A stencil is definitely handy, as are tweezers.  I can probably solder them just as quickly by hand, but either way, its not prohibitive IMHO.  Paste will dry out so one sitting is recommended.  Tupperware and a freezer can help if you have to set it aside for a little while.

c) because 0603 passives are unlabelled, you have to be super careful not to mix them up. Easy enough when soldering just one value at a time by hand, but with reflow, you need to place all the values on the board all at once. Can be done, but you need to be super careful - again, the proportion time probably outweighs the time saved in soldering all bits at once.

And then you still need to solder all the bits containing plastic by hand. Doesn't seem sensible to me. But I'd be very interested in hearing how you go, because I have a whole stack of SMD boards yet to solder, and doing them by hand certainly takes a long time.

If your parts are 0603, it will be a bit more challenging but still doable for sure.  Magnification and some patience will go a long way.  The biggest risk IMHO is getting all the parts in place and then dropping it on the way to the skillet.  :-)  We used to call that a "wipe down" pcb.  If you know where the parts should go, you're in luck.  Otherwise, you pick off the ics and semis, scrape away the passives, and wash the pcb in isopropyl till there's no more solder paste or flux on it.  Start over.  You'll learn a whole new way of swearing.

That said, if you do have problems after reflow, you can reheat and pull the parts back off while hot.  I would not do that more than 2x but it can be done, all is not lost.

For the fine pitch ic's, there is a technique called drag soldering that is pretty cool once you get the hang of it.  If you're doing skillet or toaster oven reflow, you'll just need some solder wick and flux to clean up any bridges.  It should self center pretty well during reflow due to surface tension.

The lesson here is surface tension can be your enemy, but with a little mutual respect, it can be a very dear friend.

Either way, I'm sure you can do this Alan!  Looking forward to seeing another cool build and hope this helps.  If you have other questions, feel free, I will help as able.

[pickdropper]

You can get 0603 resistors that are labeled.  0402 is usually the point where the resistor labels drop off.

MLCC caps generally don't have labels, so mixing them up is always a potential problem with those.

[Govmnt_Lacky]

After a lot of back and forth... I think I am going to make my initial attempt with my MGChemicals paste and my Stinel adjustable heat gun. The low-flo hot air setting on the gun is pretty tame. I "think" it will be able to do the job without sneezing all of my 0805-sized components off the PCB.

For the ICs.... should I tack 1 or 2 pads down by soldering iron and then lay down the paste -OR- should I just paste the pads, place the ICs, and gun it?

[pickdropper]

After a lot of back and forth... I think I am going to make my initial attempt with my MGChemicals paste and my Stinel adjustable heat gun. The low-flo hot air setting on the gun is pretty tame. I "think" it will be able to do the job without sneezing all of my 0805-sized components off the PCB.

For the ICs.... should I tack 1 or 2 pads down by soldering iron and then lay down the paste -OR- should I just paste the pads, place the ICs, and gun it?

I'd just paste, place and gun it.  It's probably less susceptible to blowing off the board than the smaller components and leaving it free should help it self-center on the pads.

[Govmnt_Lacky]

Its funny because I have seen a lot of replies on here about the smaller bits blowing off the PCB but, I have seen a lot of Hot air gun soldering tutorials online and it seems like all of them just crank the air gun on and the bits never seem to blow off the board.

It sounds like they put their hot air guns on full blast! 

[wgc]

Dude, no guts no glory.  Not the end of the world if you screw up.

With all the variables involved, including the output of your heatgun, its hard to say what will happen until you try it.  For sure, your best bet is to try and hit it straight on, and not from the side or an angle.

Agree with Dave's comments.

Like I mentioned before, try some passives on a piece of copper clad and see how it goes.

You can do it..!