2
\$\begingroup\$

I have a couple of questions, plus I'd invite any comments/feedback/suggestions to improve my results.

I used solder paste Chipquik SMD291AX10 (Sn 63 / Pb 37) on the pads and also flux (Chipquik SMD291NL) on the pins/pads of the components. I wonder if this is the recommended practice, or whether the solder paste already has the necessary flux in the mix?

Notice that there is this "amber" touch around the components, and I believe it is the "cooked/possibly-burned" flux. Here is the link to a couple of pictures I took, plus the temperature profile:

https://cal-linux.com/tmp/

enter image description here enter image description here enter image description here

My other question is: could it be that the temperature profile (in particular, how slowly it went) is negatively affecting the results? The image above shows the plot of the profile, as captured by a thermocouple some 5 cm away from the actual board inside the toaster-oven.

What I did is: set the oven to max. When the thermocouple temperature reaches 90 °C, I unplug and leave it unplugged for one minute. Then, when the temperature reaches 185 °C, I unplug it for good. Then, when the temperature passes through 200 on its way down, then I open the oven door to speed up the cooling down.

Does the above make sense to those of you with more experience on this?

Thanks!
Cal-linux

\$\endgroup\$
4
\$\begingroup\$

Paste solder has flux in it. The shape of the recommended temperature profile is to bring the system up to a temperature so as to activate the flux, give it time to work, and then melt the solder before the flux becomes deactivated.

My experience with respect to prototyping is that the stuff is pretty forgiving. Your profile, though, is a bit slow, and could be about twice as fast. Indeed, I think your profile moves along slow enough that its pointless to unplug at 90C, but I could be wrong about that.

I recommend trying to get more beef out of the toaster oven. Certainly, cover any glass with tin foil. Also, I got much better performance by dismantling and insulating with glass wool.

Put the thermocouple on the board, near a big component. It may not be as hot as you think.

In general, I'd say you have too much solder. Don't worry about separating the solder on each pad, and just smear about half that amount across all the pads. Let the surface tension separate the pads and pull the components into place.

Hard to say, but my guess is that some of your joints may be cold. The fact that the resistors didn't get pulled into a centered position by surface tension probably indicates uneven heating. You might expect tombstoned 0402's, but the sizes you're using should work just fine.

Great start!!

If you're going to keep on doing this, you might consider making boards at a house where they give you a free stencil, like pcb-pool.com

\$\endgroup\$
1
\$\begingroup\$

The goal of soldering is to produce good joints. If you're producing them without adding extra flux then you don't need to add any. Same with speed: longer exposure to heat causes oxidation of metal pieces and some of the joints come out bad. However, if you heat up too fast some areas of the board won't reach the temperature and some other joints would come out bad. Count bad joints after each reflow and tweak the process till you get zero bad joints - heating/cooling speed, chemicals, the shelf time of the blank are all factors, as well as several others.

\$\endgroup\$
1
\$\begingroup\$

Wouldn't it be best to just follow the manufacturer's specifications for the solder paste you are using? In the specification they tell you the flux type in the mix and the recommended reflow profile. See: http://www.chipquik.com/datasheets/SMD291AX10.pdf

\$\endgroup\$
5
  • \$\begingroup\$ Of course that would be best, but it's not real easy to achieve with an off the shelf toaster oven with marginal power for the job and without a ramp-soak controller. \$\endgroup\$ – Scott Seidman Nov 19 '15 at 19:20
  • 1
    \$\begingroup\$ You need to get one of those newer computerized toaster ovens with an SMT setting next to the bagel button. \$\endgroup\$ – Nedd Nov 19 '15 at 19:26
  • \$\begingroup\$ sound's funny, but I installed a heat-ramp-soak controller from Omega instruments in my toaster oven. The knob on the oven just turns on power, and I have to manually twist it off when done. I love how it dings when I turn it off. \$\endgroup\$ – Scott Seidman Nov 19 '15 at 19:31
  • \$\begingroup\$ Oops --- I do remember checking the solder paste's datasheet and saw the profile, but I certainly missed the detail about the flux already contained in the paste! Thanks for pointing this out! \$\endgroup\$ – Cal-linux Nov 20 '15 at 3:00
  • \$\begingroup\$ Actually only a partial joke. Some toaster ovens with a bagel button (switches heating from one side to two sides) can be used to roughly simulate the profile of a reflow oven. Other simple toaster ovens might have a Bake/Broil or a High/Low switch. You first bake the board on the lower settings to simulate the slow rise portion of the profile, (or until the flux starts to activate), then breifly turn up the heat for the final reflow action. Using a simple controller on the switches you could likey come close to recreating the recommended profile. \$\endgroup\$ – Nedd Nov 21 '15 at 9:11

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.