8
\$\begingroup\$

I'm going to assemble a prototype PCB which have components on both sides. I have access to a reflow oven with profiling control, solder paste and stencils (from OSH-Park)

In the reflow process for the second side, I expect that small components will stick to the board even in the melting temperature, as mentioned in this answer.

But I'm worried about a big component that I used on the board. SEDC-10-63+ is a 3cmx2cmx1cm coupler with weight of 7.3g. I have two of them exactly packed back to back on the layout. Because of the exposed pad at the bottom of package I cannot use a heat gun nor the hand soldering iron to solder the part. My question is that does the bottom part will fall off because of it's size or I will going to get a successful soldering and I shouldn't be concerned so much.

enter image description here

Non-Acceptable Answer

I know that I can use a low-temperature solder paste like this one, or using SMD Epoxy Adhesive but I'm more interested in hearing the limitation of simple reflow process about what packages can and what can't soldered using this method (with exactly dimensions and weight they have successfully/unsuccessfully assembled)

Thanks

\$\endgroup\$
  • 1
    \$\begingroup\$ Because this will have EVERYTHING to do with the surface tension of the solder on the pads, I suspect you'll have to provide a land pattern, and maybe the thickness of the solder paste applied. \$\endgroup\$ – Scott Seidman Jan 3 at 18:27
  • 1
    \$\begingroup\$ 10 yrs ago we always used automated glue dots for bottom side parts. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jan 3 at 18:29
  • 1
    \$\begingroup\$ @ScottSeidman, thanks for your comment, land pattern is already provided in the question. If you click on SEDC-10-63+ link in the question it will directly bring you to the pcb footprint. For the solder paste thickness I use a 100um stencil. \$\endgroup\$ – pazel1374 Jan 3 at 18:36
  • 4
    \$\begingroup\$ I us Kapton tape to hold already soldered components in place that I am concerned about falling off, or shifting, while reflowing the second side. \$\endgroup\$ – CrossRoads Jan 3 at 19:46
15
\$\begingroup\$

You can find some good and relatively modern information in this document.

WEIGHT LIMITS FOR DOUBLE SIDED REFLOW OF QFNS Sasha Smith, David Connell and Bev Christian

Although their tests were for QFN packages, they work with the ratio of total pad wetted area to package mass. It will vary a bit with the solder type as well, in the paper SAC305 (96.5% tin, 3% silver, and 0.5% copper) is used.

They also refer to an older "rule of thumb" formula in unpleasant mixed units:

\$\frac{\text{Weight of the component (grams)}}{\text{Sum of the area of all solder joints (square inches)}} \lt 30\$

Of course you can always glue the parts. It's often possible (and often desirable) to keep all the heavy parts on the "top" and lighter parts on the bottom.

\$\endgroup\$
  • 1
    \$\begingroup\$ WoW, thanks @Spehro. If thats Ok with you I will accept your answer in 2 or 3 day so I can hear from other experts like you too. Thank you again \$\endgroup\$ – pazel1374 Jan 3 at 18:51
  • 1
    \$\begingroup\$ @pazel1374 It's always best to wait at least 24 hours before accepting an answer. \$\endgroup\$ – Spehro Pefhany Jan 3 at 18:52
  • \$\begingroup\$ WRT the formula (nice rule of thumb by the way), I tried to do the math and I got grams/mm2 < 0.046. Not nicer but the units are right(er) at least \$\endgroup\$ – clabacchio Jan 11 at 15:48
3
\$\begingroup\$

There is no equation for max part size. It will depend on your pad geometry and stencil geometry and the resulting surface tension. In production, unless it's a standard jellybean component that the manufacturer knows won't fall off, I see glue pads. Designing a board with massive components on both sides is bad DFM practice.

Also, you can certainly hand solder those parts you show below. I've soldered similarly difficult parts on multilayer boards using a SMT hot plate (example) and a hot air gun from the top.

\$\endgroup\$
  • \$\begingroup\$ I'm not asking about equation. I'm asking about the experience that people had before. like if they successfuly solder a not so small component and if yes or no what is the size of the component they tried. Moreover I definitly can hand solder that! the problem is the exposed pad and have component exaclty on top of each other which using a heat gon directly on top part will melt the bottom part solder joint too! \$\endgroup\$ – pazel1374 Jan 3 at 18:42
  • 1
    \$\begingroup\$ @pazel1374 It doesn't look good when you insult people trying to help you and demean an entire industry at the same time as asking for help. There are rules of thumb but no exact solution. If your design forces this requirement a typical approach would be to run test panels with just these parts on both sides and test. Custom rib type SMT carriers are another possible path to production but require working closely with your fab shop. Most people stick to touch up soldering or glue dots. \$\endgroup\$ – EasyOhm Jan 5 at 22:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.