0
\$\begingroup\$

I need to assemble and test a rather large (40x30 cm) board with single sided SMD parts, Sn Ag Cu solder (officially 217°C melting point). Usually I place group after group of components and locally solder these groups with a "hot air pen" (rework solder station), which works beautifully.

However, this method does have problems when parts are tall and contain a lot of temperature sensitive material like plastic connectors or electrolyte caps. For those parts, the hot air can heat the parts way higher than the 260 °C they are usually rated for, before the solder starts to melt, leading to damaged components.

Is there a reliable way to solder such parts when a proper reflow oven is not available ? In particular, I have a plastic multirow SMD connector with a BGA like landing pattern (Samtec SEARAY) that I don't know how to solder well.

I do have the following:

  • hot plate up to 300°C
  • hot air pen
  • kapton tape
  • a large convection oven which goes to maybe ~210°C (doesn't reflow reliably).

Unfortunately the board doesn't make good contact with the hot plate because it warps upwards when heating the backside

\$\endgroup\$
11
  • \$\begingroup\$ Assuming this is just a prototype, the obvious solution is leaded solder/paste and reduce the temperatures accordingly. \$\endgroup\$
    – Lundin
    Commented May 6, 2022 at 11:28
  • 3
    \$\begingroup\$ @tobalt 200 pins!? Soldering a 200 pin BGA manually without a reflow oven is simply madness. Also these would need xray inspection and so on. Hire a professional contractor who got all this equipment. \$\endgroup\$
    – Lundin
    Commented May 6, 2022 at 12:47
  • 2
    \$\begingroup\$ The integrated thermometer only tells you the temperature inside the heater but you want to know the PCB surface temperature. Put a thermometer in contact with a larger copper plane and monitor the actual surface temperature. Gradually bring the surface up to ~150-200C (or what your paste recommends), then put directly into the hot plate to quickly do the reflow. Might want to practice once on an empty PCB but it's not too tricky once you can see the PCB temperature. \$\endgroup\$
    – user1850479
    Commented May 6, 2022 at 12:49
  • 2
    \$\begingroup\$ Don't do it for the BGA. I worked for a company once and was tasked to run tests to see if they could get away with hot air in-house. These rework staff were pros and these were BGAs with less than 200 pins. Every board failed testing. \$\endgroup\$
    – DKNguyen
    Commented May 6, 2022 at 13:33
  • 1
    \$\begingroup\$ I don't think his plastic connector is literally a BGA part, but probably a standard SMD part with pads on the underside. \$\endgroup\$
    – user1850479
    Commented May 6, 2022 at 15:08

1 Answer 1

Reset to default
2
\$\begingroup\$

Thanks for the suggestions, especially @user1850479 .

I put a glass petri dish above the component to create a kind of hothouse and then set the whole thing on the hotplate.

I set the hotplate to 270 °C after doing some tests with a thermo couple.

The initial buckling prevented full contact so the surface could heat gently to around 150~200 °C. After about half a minute I pressed down on the hothouse, flattening the board against the hotplate. Within about 15 more seconds the area reflowed evenly and nicely.

The result was very beautiful and reliable.

enter image description here

In this side view one can see that several rows of pins have reflowed to a near identical result. Also, I am sure that indeed all pins' charges have reflowed because the connector drops down a bit when that happens as described in its datasheet linked in the question.

\$\endgroup\$
1
  • \$\begingroup\$ Glad you got it. You should mark this as the answer. \$\endgroup\$
    – user1850479
    Commented May 9, 2022 at 19:18

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