10
\$\begingroup\$

I'm currently reading an article (in Russian) and the author compares two devices (hard disks) of the same model one manufactured by Company A and the other manufactured by Company B. One of the things he compares is how solder looks like on the boards.

Company A device board has soldered parts looking like this:

"bad" solder

and this is claimed to be "low quality tin-coating". What I see is that solder spots are not shiny and the solder layer is thinner near the spot edges and thicker far from the edges (a bit concave).

Company B device has soldered parts looking like this (that's exactly the same area of the board as on the Company A device):

"good" solder

and this is claimed to be "high quality tin-coating". What I see is that solder spots look shiny and the solder layer looks having uniform thickness across each spot.

So to me the first board just looks neater. From what I know about soldering once the surfaces have been properly degreased with soldering flux and the solder was melted properly the connection will be just fine regardless of how shiny and neat looking it is.

However the author claims that the Company B device is of higher quality and should be preferred because of (among other factors) the "better quality" of tin-coating. How reasonable is such claim? Can device reliability be judged based on such tin-coating analysis?

\$\endgroup\$
4
  • \$\begingroup\$ I agree with you. Board A looks to me like a tradition board done in a commercial PCB solder bath. Unless there is some special properties of the 'tin-coating' on Board B that we don't know about, say ultra-low resistance with ultra low thickness. I would say Board A is better ??? \$\endgroup\$
    – kingchris
    Jun 5, 2012 at 17:21
  • \$\begingroup\$ From what I learned, if solder isn't shiny, it's a sign that it's been reflowed without using flux, and probably has internal oxidation. Plus, aren't those vias rather than parts (could be for through-hole components, but I see no pins)? \$\endgroup\$
    – Ben Voigt
    Jun 5, 2012 at 17:42
  • \$\begingroup\$ @Ben Voigt: Well, looks like those are vias, yet the author uses their looks in his analysis. \$\endgroup\$
    – sharptooth
    Jun 5, 2012 at 17:49
  • \$\begingroup\$ It's hard to tell from the low-res photos, but the second PCB might be Immersion Silver, rather than Hot Air Solder Leveled. Immersion silver is preferred these days because it's a lot flatter, and so better for very small SMD devices. \$\endgroup\$ Jun 5, 2012 at 18:44

3 Answers 3

16
\$\begingroup\$

I've said this before, and I'll say it again: How shiny the solder looks is not a reliable indication of the quality of the solder joint.

Even before lead-free solder the shine wasn't a reliable indicator. More reliable than lead-free solder, but not reliable enough for most people.

Here are some things that I look for when evaluating the quality of the solder:

  • Consistency: Do all the solder joints look the same? If not, it indicates a variation in the solder process and thus it is more likely to have problems.
  • Wetting/Wicking: Did the solder melt evenly and flow onto the mating surfaces, or does it look like the solder beaded up like water on a newly waxed car? Solder that is beaded up could have problems, and hidden cracks under the bead.
  • Smooth finish: I'm not asking if it is dull or shiny, but rather is it smooth or are there lumps in it? Lumps are a sign of uneven or incomplete melting.
  • Conductive Flux: This one is rare, but important. Some types of flux are conductive, but not everyone is aware of this. Sometimes a board will be reworked with conductive flux but the flux will not be cleaned off correctly (water for water soluble flux, etc.). Check that the proper flux was used in the proper way. Note: Some flux leaves a residue and this is OK as long as the residue is not conductive even though it might look bad.
  • Cracked Solder Joints: Often this can only be seen using a microscope, and sometimes not even then.
\$\endgroup\$
1
  • 1
    \$\begingroup\$ +1 in general, but particularly for mentioning conductive flux. We had a local small scale manufacturer (hand soldered prototypes) mess up QFN packages this way. The flux was conductive enough to hold the MCLR line of a processor low despite a 20 kOhm pullup. Eventually we had to unsolder the QFN parts, clean the board, then resolder them using rosin flux. We've meanwhile taught that manufacturer about QFN issues. This was their first time. \$\endgroup\$ Jun 6, 2012 at 11:48
3
\$\begingroup\$

It used to be a lot easier before RoHs soldering processes came into vogue.

To me, all solder joints with lead-free solder look dull, and look quite a lot like a poorly-done conventional SnPb solder joint. (I guess I'm showing my age now)

IPC-A-610E is the 'official' criteria for judging acceptance of solder joints.

\$\endgroup\$
2
  • \$\begingroup\$ As long as you don't mention that solder on the valves of your radio... :-) \$\endgroup\$
    – stevenvh
    Jun 5, 2012 at 18:40
  • 1
    \$\begingroup\$ Connecting tubes with solder is old fashion! I heard they made new kind with pins made of invar so you can stick them into little pads.. yeah, sockets. \$\endgroup\$
    – user924
    Jun 6, 2012 at 2:18
2
\$\begingroup\$

The case can be of comparing 2 completely different technologies: traditional solder paste printing with stencil printer vs electroplating by deposition of solder with help of electrolysis through openings of photoresist mask.

The difference can an order of magnitude in amount of solder deposited. If the board is made with electroplated deposition technology, then it can be much more recent and perhaps superior in quality.

What can be possibly invented in close future is some sort of pick-and-place of individual leafs of solder foil created by some precise CNC chopper on the fly.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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