3
\$\begingroup\$

I don't have much soldering experience. I soldered 5 level converters, they seem to work and look about the same as this STM32 as I soldered.

However, after a closer look I see not all pens 'shine' and some don't have a good form. Probably used too much soldering, because it was quite hard to determine (using too less cause nothing but heating pins which I also wanted to prevent more than needed).

How can I best test if all soldering connections are ok? I don't have any STM32 experience, so I don't know if there is a general way to do this except for writing programs/testing for each pin?

Or is it visible directly from the pictures that this will not work (I hope not).

(btw it's hard to make pics, they are all from the same STM32, the first pic shows soldering connections which are much darker than the second, the second looks more like the real thing).

Edit: I think I know what I did wrong ... since I didn't have much space the solder tip and solder tin sometimes might have touched directly, also during that the pin itself might not have been heated enough. Next time I really should make sure the tin and tip are on opposite sides of the pin.

pic1

pic1b

pic2

pic3

enter image description here

Update: Used aceton/alcohol

I used a mixture of 50% 96% alcohol and aceton and got the following results: It seems to leave some white residue ... wondering if that is a problem.

enter image description here

\$\endgroup\$
  • 2
    \$\begingroup\$ There are some cleaning solvents that might help remove some of the flux, which helps us "see" them better. But many of those look acceptable enough for hobby work to me. A few look "cold", but it's hard to be sure. Of those, I'm more bothered by A15 and A12. C15 is "different" but I can't tell much from the pics. Learn to quickly clean the soldering tip, dab just a bit of solder to it, move firmly to heat the target, and then quickly use a consistent amount of solder (up to a point, less is more.) Oxidation is bad and the longer you take the more that gets in the way of doing it well. \$\endgroup\$ – jonk Aug 5 '17 at 20:16
  • 1
    \$\begingroup\$ Google better, there are free versions around \$\endgroup\$ – Gregory Kornblum Aug 5 '17 at 20:54
  • 1
    \$\begingroup\$ @MichelKeijzers Yeah, I mean a liquid, labeled for this purpose, in a jar with a brush. (I don't imagine you want to use the old method of a boiling vat of CFC-12 that you immerse the board into.) In some situations, an aerosol can spray is okay. But to make your own, just get some 91% (or better) iso-propyl alcohol and some pure acetone and mix these in a 50/50 mixture. That stuff will work quite well. \$\endgroup\$ – jonk Aug 5 '17 at 21:13
  • 1
    \$\begingroup\$ @MichelKeijzers See, it's a good document :) \$\endgroup\$ – Gregory Kornblum Aug 6 '17 at 4:36
  • 1
    \$\begingroup\$ wiggle the pins close to your ear; squeeks/clicks are bad news. flashlights behind can help. you can also check the continuity of adjacent pins (there usually shouldn't be any) \$\endgroup\$ – dandavis Aug 7 '17 at 9:39
8
\$\begingroup\$

I don't see any obviously bad solder joints on those pictures. However, these pictures aren't great for inspecting the solder joints. To get pictures of small things, you need to use a macro lens, good lighting, and proper exposure.

The main thing to look for is that the solder flowed. That is best seen in the top picture. It seems the solder wetted and flowed on both the pads and the pins, although too much solder is clearly evident on many of the joints. Too much solder makes it hard to tell the difference between a blob that wetted and flowed over the pad, and one that is just sitting on top of the pad with a layer of flux or oxidized solder in between.

The proper procedure for hand soldering such pins is to put the iron in contact with both the pin and the pad. The feed a small amount of solder into a crack between the iron and the pin. The purpose of this little bit of solder is not to make the connection, but to provide good thermal conductivity between the soldering iron and what it's trying to solder.

After a second or two, the pin and pad should be hot enough to melt solder on their own. Feed a bit more solder. This should now melt easily and flow all over the pin and the pad. Remove the iron as soon as the solder has flowed. While the joint is cooling, make sure nothing is wiggling the two parts of the joint relative to each other.

Added

On closer look at the third and fourth pictures, the solder joint for C15 is suspect. At the least, the solder doesn't seem to have flowed over the whole pad.

\$\endgroup\$
  • 2
    \$\begingroup\$ Whoever downvoted this, what exactly do you think is wrong? \$\endgroup\$ – Olin Lathrop Aug 7 '17 at 12:46
  • \$\begingroup\$ I have a macro lens somewhere, so maybe I should try a better pic (not able before the weekend though, since the camera is not near me). Could I try to resolder C15 by heating it up again so it will reflow or will because of the lack of flux make it worse?) I would need solder wick instead? \$\endgroup\$ – Michel Keijzers Aug 7 '17 at 12:47
  • 1
    \$\begingroup\$ @Mich: Try it and see. Simply re-flowing what is that may work. Otherwise, remove most of the existing solder first, then add fresh solder, which includes flux. \$\endgroup\$ – Olin Lathrop Aug 7 '17 at 12:48
  • \$\begingroup\$ Thanks for the remark ... first I will check if it works (since in that case I don't change anything). Otherwise, I will try to reflow it, if that will not work solder wick + new solder. Thanks again! \$\endgroup\$ – Michel Keijzers Aug 7 '17 at 12:51
  • 1
    \$\begingroup\$ @MichelKeijzers Simply heat it up and add a tiny bit of tin. Most often, fixing bad joints is merely a matter of heating it again. When doing so, some superfluous tin also end up on the tip, which is fine. \$\endgroup\$ – Lundin Aug 7 '17 at 13:00
5
\$\begingroup\$

If you use leaded tin then none of them should look matte. (ROHS tin is another story, such joints tend to look rather bleak even when properly soldered.) This wouldn't pass a professional examination, but might work (for a while) for indoors hobbyist purposes.

Overall looks like you use far too much tin and too little heat, several joints look fishy and "cold". Often the problem is that the pad didn't get enough heat. The iron needs to be placed so that it touches both the pad and the pin. You can let the tin quickly nudge the iron first, only to trigger the wetting. Then apply the tin for the actual joint without touching the iron.

There should be roughly enough tin to cover the hole. For professional use in tough environments, it is good if a bit of tin floats out on each side of the hole (assuming it is through-hole plated which should be the case here), to prevent corrosion over time. However, you don't need to "drown" the whole pin in tin as in those pictures - this might hide poor soldering underneath.

You should strive for something that looks like B3.

\$\endgroup\$
  • \$\begingroup\$ Thanks for your answer and some good tips .. I will try it again with my next one (or if the board above does not work to do some rework on this). \$\endgroup\$ – Michel Keijzers Aug 7 '17 at 13:10
3
\$\begingroup\$

You did a fine job there. Those cheap header pins are likely non copper. As such it is best to dial up the temperature a little and out some solder flux on the pins plus the holes before soldering them.

But your soldering is fine.

\$\endgroup\$
  • 4
    \$\begingroup\$ While I agree that the soldering appears to be adequate for the purpose, calling it a "fine" job is going too far. \$\endgroup\$ – Olin Lathrop Aug 7 '17 at 12:06
  • \$\begingroup\$ It's my third soldering things on this scale (except from soldering some wires)... But I learnt from it. I don't have separate flux (except for what's already in the solder), but I ordered it to be added for next times. \$\endgroup\$ – Michel Keijzers Aug 7 '17 at 12:49
  • 1
    \$\begingroup\$ @MichelKeijzers You don't need separate flux for this. With some "solder suck wire" you can remove superfluous tin. Then if you need more flux, just add a tiny bit more tin after that, to use the flux in the tin itself. Just don't heat the tin too long or all the flux will vaporize. \$\endgroup\$ – Lundin Aug 7 '17 at 12:55
  • \$\begingroup\$ @Ludlin I will do that (usering solder wick), but only when needed (I mean when it doesn't work). But good to know I can use normal solder :-) \$\endgroup\$ – Michel Keijzers Aug 7 '17 at 18:31
  • 1
    \$\begingroup\$ If I had to criticize your soldering, I would have dialed up the temperature a little and used less solder. Your looked a little bit too cold. But that borserlines on nitpicking. \$\endgroup\$ – dannyf Aug 7 '17 at 18:52
1
\$\begingroup\$

Whereas the above answers are useful, I don't think they answer the actual question:

How can I best test if all soldering connections are ok?

Even poor solder connections might conduct fine for a while, or even indefinitely if not disturbed. The problem is that you want to have some reassurance that the connections remain good even over changing conditions of thermal expansion/contraction and vibration.

Put the board through a few cycles of heating/cooling through its rated temperature range, all the while applying a low-voltage square wave to each of the pins in question. If you see any interruption or distortion in the output, one of the connections has likely failed. Do the same procedure but rather than temperature variation, run a small motor acoustically coupled to the board, ideally trying multiple different axes of motion.

Edit: if you're trying the acoustically coupled motor, you would also want to test a range of frequencies (rotational speeds) of the motor, because the mechanical properties of the board and the solder connections will vary over this range - from "very low" ~1Hz to the high hundreds of Hz.

\$\endgroup\$
  • \$\begingroup\$ Interesting approach ... about the heating ... I would need to check what is possible (and not having a temperature sensor that can reach that temperature, neither something to heat it up). But the low-voltage square wave is possible. I have a servo motor and small ac motor (very small), but haven't tried even to connect it (I'm just reading more about the cpu etc). \$\endgroup\$ – Michel Keijzers Aug 12 '17 at 17:04
  • 1
    \$\begingroup\$ The heating can be as simple as putting the board in a fridge or freezer, and then taking it out and exposing it to a heat gun at some known (and safe) distance \$\endgroup\$ – Reinderien Aug 12 '17 at 17:08

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.