Double Sided Surface Mount PCB Population

[youtube=http://www.youtube.com/watch?v=6M3MeADb1dE]

Above is a video detailing one method for populating a two sided surface mount PCB. We covered using a stencil to apply solder paste for a PCB a few weeks ago. In the comments there was a debate about the virtue of using stencils as well as a question about how two sided boards are populated. This was a good question because reflowing a board twice can cause components on the underside to fall off.

[Wim L’s] comment mentions that there are a couple of methods for two sided population. In the video you will see that a stencil is not being used, but instead, paste is applied by a pedal actuated syringe. The paste is applied to the underside of the board first, then a teeny dot of epoxy is added to hold the component in place. Each part is then positioned normally and baked in a reflow oven. This process both reflows the solder, and cures the epoxy. When the board is reflowed a second time, the epoxy holds the bottom components in place as the top solder reaches its melting point.

This method of applying solder paste is slower than using a stencil. But if done correctly, every component can get the amount of solder needed.

20 thoughts on “Double Sided Surface Mount PCB Population

  1. I used to make double sided boards all the time using the syringe method. I did it manually but also used a low cost manual pick and place from Bomir that has a built in electronic programmable dispenser.

    Unless the parts are big, there is no need for glue..especially on little resistors and caps. If you make a frame for the board (the Bomir had an integrated PCB holder) you can put pasted down and the parts, flip it over and place the other parts. Then of course you need a frame in the oven as well, but I never reflowed double sided boards twice.

  2. My firm does a lot of our own prototype boards, and that team just reflows the boards twice. They do the side with the light components first, and then run that side on the bottom when they reflow the side with all the chips and inductors and heavy stuff.

  3. I used to do a LOT of SMT work (avionics, semi-custom actuator controllers) and we did all our repair and short run work by hand. No stencils, no reflow oven, no solder fountain. Magnifier lens, Weller soldering station, Kester solder spool and a steady hand was all we used. I didn’t even know about stencils until I was getting out of the industry. We would double-check our solder joints at X30 on the microscope to make sure the joints were flawless. It did take a couple days of training to get to where your joints were perfect… That didn’t come out right. :P

  4. On larger boards, wouldn’t the epoxy on the bottom make removal of defective or toasted components impossible? It’s tough enough to remove a SMT fuse on a heavily populated board without causing any collateral damage. I’d be afraid that the pads would be ripped up on this design.

  5. I would recommend using the epoxy for the underside as directed, but before reflow, applying paste to the top side and placing those components. Then reflow the board. The fewer thermal cycles, the better.

  6. I work in the electronics industry. On double-sided boards, another stencil is used to apply a small blob of epoxy before the solder paste. The board is then run through the reflow oven, where the epoxy cures and the solder reflows.

    The other side of the board can then be stenciled, populated and reflowed without the epoxy.

  7. There is glue specifically for this purpose. Loctite 348 is one such product.

    Unless the part is really large, I don’t recommend the dual heat cycle. The glue should start curing before the solder reflows and hold the part on.

  8. I used to work in the electronics manufacturing industry myself. The most common method is the one jeff-o mentioned, with the epoxy and solder paste being done with two stencils. Another method that also gets used is to stencil the epoxy, place the components, cure it in the oven, then stencil-place-flow the other side. Then the epoxied side gets soldered in a wave flow solder machine.

    It’s ugly, and tends to result in a lot of minor rework to remove shorts, but it works and does get used.

  9. What a surprise to see my little movie featured in Hack A Day, and a great pleasure too… Thanks!

    The background: this movie was intended to explain our readers (/begin commercial/ Elektor is a monthly magazine for electronic designers and entousiasts, available worldwide in more that ten languages /end of commercial/) how-to-do low cost double sided soldering with a reflow oven.
    The movie comes with an article, that was published in september and describe a bit more in detail the whole process/idea.

    Using a pasta dispenser does not aimed at being faster but more versatile, we have here a lot of different project – what I like to call electro-diversity – but only produce a few protypes mostly used for test (destroying) purpose :).

    So far I’ve tested two different epoxy glues:
    – ref. SMA10SL from Electrolube
    – ref. CB8006-V91 from Loctite
    Everything went well in both cases; the Loctite is maybe a bit more fluid, is it a feature or a drawback, you have to decide. Both are 10 ml syringe, the Electrolube came with a plastic needle and a plunger, we got a Bundle Loctite so no idea if you can have it packaged with this two accessories too, anyway the plunger can be replace by a screwdriver, any kind of “Luer Lock” compatible needle will fit, be careful with medical one they stitch in.
    I believe that with these two brands we have world coverage and you won’t have difficulties to find then anywhere.
    But I guess any kind of “SMD Chip Bonder” (key word for your favorite search engine) will do the trick.

    The board is indeed a USB to 1-Wire/I2C bridge.

    P.-S. Don’t blame for the music please, did not choose for it ;-).

  10. This has LESS of a probability of failure than the hot air gun.

    The hot air gun can blow other parts off the board, and you are heating the board more than twice.

    You are heating it for EVERY component, thus not good.

    Hot air is best done for REWORK.

      1. That’s not a disagreement… He said hot air is best reserved for rework (vs. baking in an oven). He said that the oven is less probability for failure than hot air for manufacturing.

        I believe all of those statements to be correct.

Leave a Reply to jamieriddlesCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.