Reflow vs. Just soldering it...

[alank2]

I've been trying to move from PTH hand soldering to SMT/stencil/reflowing.  I've hand soldered 4 of the board I'm working on (about 40 smd parts) with not too bad a speed and 100% success.  The smallest part is a 0.8mm TQFP 32.  I tried stencil/reflow tonight and by time I get everything setup and go through the process it seems to take a lot more time.  What has been your experience?  I would think the stencil/reflow would be faster...

[zapta]

The more identical boards you are making, the more advantage the smd/stencil approach has. If I have a stencil, I will use it even for a single and small board. Seems to be easier and nicer solder joints though not necessarily faster.

How do you pick and place? A vacuum pump makes it must easier than with tweezers.

[alank2]

I started with a vacuum tool and then moved to tweezers as I wasn't getting the hang of it.  It just wasn't a good night for reflow.  I don't have a good desk to work at for it, my solder paste is old, my first try at a metal stencil seemed to have too much paste on the board (perhaps 0.15mm is too thick).  First the paste was too thick so I added a little 100% ipa, then it seemed like too much paste was on the board.  It was pretty discouraging tonight.  Luckily I did one board at a time so I only have two boards with solder bridges in the usb port and tqfp instead of a lot more...

[Fat]

Punched out two boards tonight with the oven. If I have more than a few components I use the oven. Single chips I just use the iron.  Spent three hours one night doing two SMD boards with an iron. Did the same board in the oven in 25 minutes. No errors or bridges.  Need to learn how to make stencils.

[tggzzz]

What has been your experience?

See https://entertaininghacks.wordpress.com/2015/03/25/assembling-pcbs-with-surface-mount-components/

[unitedatoms]

After 5 boards made by air gun I switched to reflow oven for boards 6,7,8. And finally got the perfect board after 7 previous ones. The board is 100 TQFP 0.5mm, 128 TQFP 0.4mm, and 40pin PFC 0.5 mm.
All boards work, but only the last one did not require anything to do after it got out of oven. This is my notes
http://unitedatoms.com/blog/soldering-notes/

1.Screw in the mounting screws (to have smallish gap between bottom of board and oven pod) before reflow to have minimal contact to the oven pod. The air will be hotter than oven pod, so the temperature from sensor will match the temperature of board with better accuracy. Also there will be no cold joints
2.When printing it is OK to have a single side of stencil fixed with wide mask tape. No need to have 2 points
3.Make sure there is no sticky tape at the bottom of stencil So when stencil is lifted the lifting speed can be controlled
4.The excess solder can be put back into jar, but with condition that it spent no more than few minutes in open air. <5 min
5.The board before printing must be wiped with isopropyl alcohol on coffee filter soaked in diluted remains of older solder paste. That was most successful reflow. The blank board is HASL Pb-free and solder is Pb-Free as well, but different mixture. Having the diluted (few nanometers ) of no-clean flux all over the HASL helps to merge the solders with different formulas.
6.There is no need to clean the board after reflow
7.Watch out for silkscreen hole: one spot got over the fine pitch pins. Scratch it off with steel needle before alcohol treatment
8.Move FPC connector 1 mm down to reduce area of contact for pins. This will let more solder get wicked into between pins and pads and lessen the cold joins.
9.Reflow one board at a time, since oven is so small and weak and uneven
10.Place the board on pad with high pitch part being in the center of oven pad, not the actual board center
11.Lock the capton FPC cable while the connector is still warm right after reflow. Use mini hand tool.

Thing I did not try yet.
12.Use hand vacuum pen with remote control to pick and place.

The board is on the picture:
http://unitedatoms.com/media/1063/img_0064.jpg

The coffee filters are best wiping material, because it sheds nearly no fibers.

[zapta]

The coffee filters are best wiping material, because it sheds nearly no fibers.

And so are Kimwipes.

Regarding the wiping of the HASL with solder paste, is it a common practice or something you figured out yourself?   (I am using leaded HASL and leaded solder paste).

[unitedatoms]

The coffee filters are best wiping material, because it sheds nearly no fibers.

And so are Kimwipes.

Regarding the wiping of the HASL with solder paste, is it a common practice or something you figured out yourself?   (I am using leaded HASL and leaded solder paste).

HASL pre-wiping I figured out myself. The boards are completely new, - started soldering same day as they arrived from PCB house. The packaging was vacuuming in polyethylene with silica gel bags. I made bad print and wiped the board with alcohol to repeat the print. And decided to not use too many paper filters to lower the impact on environment. So the finding is accidental.

I think that the PCB house (China) is never cleaning the boards after HASL and/or silkscreen (I don't know what goes after what). And very likely whole board is always covered in tiny layer of oxides, oils, salts and metal vapour from manufacturer's air.

The wiping of next boards I will try will be more scientific:

1. alcohol wiping
2. no-clean flux+ alcohol wiping
3. alcohol+older solder paste (tiny amount)
4. windex + alcohol, air blasting then alcohol
5. no wiping

[alank2]

Well, last Saturday I rolled up my sleeves and decided to give the whole stenciling/reflowing thing one more really good try.  I made 19 boards and it took me around 6 hours to stencil, pick/place, and reflow the boards.  I was getting the hang of pick/place, but it is tedious (I know why people want a machine to do it), and what I did was stencil all 19 boards and then pick/place reflow them 3 or 4 at a time.  The first 8 boards turned out pretty good, but the last 11 started having small solder balls at the joints that got worse the later in the day it became.  The paste was almost too wet when I first put it out (messy) and too dry on those last boards I did.  I ended up hand soldering the boards that had the tiny solder balls to correct them which took another couple of hours.

The tiny solder ball problem was almost certainly a paste quality issue.  I was doing a test with paste that was a little older by adding some of the same type of flux it has.  Doing this is much better than not, but still not the same as it was.  Perhaps this is why it has the short printing life that was maybe 2 or 3 hours instead of the usual 8 that I think I've read about.  Even if I had good paste though, it is still such a sensitive process.

In the end, I don't think paste/stencil/reflow is going to work for me.  I can hand solder in perhaps 2 or 3 minutes longer per board without all the extra complication.  If I get to a part that can't be done any other way, then fine, reflow or prewarm/hotair.

So reflow vs. hand solder, I thought about the advantages of just hand soldering it

I don't have to wonder if all the joints or good or do a post inspection/resolder session on boards.
I don't have to deal with the mess or difficulty in handling of solder paste.
I don't have to worry about how good or fresh solder paste is.
I don't have to worry about the expense of solder paste.
I don't have to deal with smearing paste or having to use the vacuum tool when placing parts.
I don't have to order or produce stencils or clean them up.
I don't have to worry about running my hand into a part and moving it during pick and placing.
I don't have to worry about whether the parts will melt because they can't take the heat or crack because they are too humid.