I am developing a PCBA that is (for me, an amateur) fairly complicated: 4 layers, double-sided, mixed SMT/THT assembly. It is a fairly dense layout with ~150 total components including all the passives on a 100x55mm board.

The top side will hold some fine-pitch SMT devices (narrowest is TSSOP, 0.65mm pitch, 0.24mm pad clearance), which leads me to believe I should order the boards with ENIG finish. There will be some 0603, Kemet tantalum cap Case B (3528), SOT-23-3, and SOD-123 components mounted on the reverse side of the board.

There are a modest number of THT components like multi-turn pots to be assembled as well. I believe the PCBA shop (I use JLCPCB) uses wave soldering for this process. In expectation of that, I have laid out the underside SMD elements all perpendicular to the intended process direction.

Finally, there are certain THT components that I will need to assemble myself after receiving the PCBAs: a tempco resistor that must be attached by long leads and thermally bonded to a particular transistor on the board, and two rows of THT headers that will be soldered "upside down" relative to the mounting direction for the rest of the THT parts. (This board will be mounted underneath another board.)

I have a few inter-related questions regarding these circumstances, which (hopefully?) each have short answers.

  1. Are there any issues with this finish/process combo? I read a blog post that offhandedly suggested that ENIG was not an appropriate finish for wave soldering, and that HASL is preferred -- but did not elaborate further, and I can't find further guidance elsewhere. I am also concerned that the fine-pitch devices being reflow-soldered might not adhere to an uneven HASL surface.
  2. I understand that ENIG is a very solder-rework-friendly surface, so it also seems good for the soldering I will perform myself when I receive the PCBAs. Is this still the case after it has been immersed in a solder wave? Is there any surface prep I should do before the final board rework, beyond washing the pads with flux, or anything else to do differently when soldering?
  3. Are there any issues in general with passing a board through a wave soldering process with some THT mounting holes left empty, for assembly later? Will the wave process close up those holes with a solid thin film of solder? These are 0.9mm diameter holes with 1.65mm diameter pads, 2.54mm pitch. I have added thief pads on the underside of the board on either end of the rows of THT pads. For space reasons, these are "surface mount" thieves, basically 1.5mm test pads on the wave-contact side of the board only, rather than drilled holes accessible on both sides of the board.
  4. I have also seen caution to use "larger" pads on the wave side for SMD components. Is that necessary for all SMD components, or only for very small sizes like 0402 and below? Which is to say -- can I get away with "standard" footprints for 0603 and up? (KiCad's library uses 0.90x0.95mm pads for 0603.) I don't actually know if JLC would wave solder the underside SMD elements, or whether they are reflowed first and only THT parts get bonded during the wave process. Nor have I found definitive guidance on how large these "larger" pads should actually be?

Thank you!

  • 2
    \$\begingroup\$ I hardly believe that JCLPCB will wave solder a small batch of boards (which I assume this is), as it is a process requiring a fair amount of fine tuning for each board design, may require some fixtures etc. More likely, they will either use selective soldering or actually hand solder the THT parts (labour is pretty cheap in China). \$\endgroup\$
    – Klas-Kenny
    Commented Mar 31, 2023 at 8:23
  • 1
    \$\begingroup\$ For high volume production, ask your for your EMS's preference. For anything low volume, ENIG is my go-to option. It "solders more easily" when hand-soldering. \$\endgroup\$
    – winny
    Commented Mar 31, 2023 at 9:14
  • \$\begingroup\$ @Klas-Kenny Yes, it's a small run. I was surprised too -- but as you watch the progress bar through various phases of PCBA, they do list "wave solder" as one of the stations your assembly goes through, and don't mention hand soldering at any point. \$\endgroup\$ Commented Mar 31, 2023 at 9:48
  • \$\begingroup\$ The improved flatness of ENIG over HASL is only really relevant for BGAs and very large QFNs. ENIG is also better for ultrasonic wire bonding, if you use any COBs. But you say you're not using anything finer-pitch than TSSOP, so unless the price difference is tiny, I wouldn't worry about using the more expensive ENIG, personally. Just use lead-free HASL. \$\endgroup\$
    – Hearth
    Commented Apr 1, 2023 at 15:12

1 Answer 1

  1. I can't say whether ENIG or HASL is better for wave soldering. HASL is a cheaper and easier process. The pads will be covered/plated with solder and levelled with hot air. There's going to be imperfections i.e. the pad surface can be uneven. But if they are not ridiculously uneven then there's going to be no problems. The main problems can come from gluing and temperature profiling.

  2. Regardless of being HASL or ENIG, the main material underneath is still copper. So those two are only "plating" processes. The adhesion between the copper and the base material (FR, CEM or whatever) is still the most important parameter. So, for me, it doesn't matter if it is ENIG or HASL for better or safer manual soldering work. If you play with a pad for too long or too much / too many times it's still possible to get the pad peeled off.

  3. Depends on the hole size, type, and location: If the hole (can be a slot as well) is too big then melt solder will flow through these holes and come to the top surface of the PCB during wave soldering, and screw everything up (especially if there are components around the hole). Small holes do not bring this problem but they will quite possibly be filled with solder, so they are going to need some extra work before mounting the TH components. To prevent these issues "peelable masks" are used. They simply cover any potential problem sources for wave soldering or cover the component pads for further use, and can easily be removed after the process.

  4. This is defined in detail under IPC. In any case you'll need larger (longer or wider or both) pads to maximise the possibility of getting the component soldered properly. It doesn't end with this: You'll also have to follow some distance- and positioning-related rules to prevent shadowing (i.e. the component blocked by another bigger one) or any other possible problem.

PS: Klas-Kenny had a good point in their comment. If the quantity is low then don't worry about wave-soldering and any other processes because the company you are (or will be) working with won't bother with wave soldering process for low quantities. But still, it's good to keep the general knowledge for further use in larger projects.


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.