4

I've arranged for a custom shaped strip of pure nickel that I intend to solder to ENIG pads on a PCB using reflow/paste.

Will we have problems getting the solder to adhere to the nickel? Is there a trick to it?

Other (potentially relevant) details:

  • The solder paste is probably lead-free, or at least it will be in production.
  • The nickel strip is 16x6x0.3 mm. This may still change but not by much.
  • The purpose of the nickel strip is to allow us to cold/spot weld LiPo cell tabs to the BMS PCB, which would otherwise be difficult if we were to attempt welding directly to the ENIG pads. The nickel strip will provide a barrier layer.
    • The cell tabs are too big for soldering.
    • One tab is nickel, and the other is nickel plated.
  • I know the ENIG surface finish already involves adding nickel to the copper foil. But then gold is added to help solder adhere. That implies that nickel is not easy to solder to, but in this case, our nickel strip is pure nickel and we've never attempted this before.
ocrdu
  • 8,705
  • 21
  • 30
  • 42
Dan
  • 133
  • 8
  • Side note: I already found this post, but it's about using an iron, not a reflow oven. https://electronics.stackexchange.com/questions/607109/soldering-pcb-with-nickel-strip – Dan Aug 25 '22 at 03:31
  • We used nickel plated PCBs for one application and the nickel did not solder well with ordinary 63/37 Kester 44 solder (not a problem because the nickel was for sliding contacts). Lead free flux may be more aggressive, perhaps you should test before committing. – Spehro Pefhany Aug 25 '22 at 03:47
  • This can probably solved by using a more aggressive flux on the nickel. You may have to wash with hot water after soldering to remove flux residue. A quick search found kester 742 flux and Johnson's J-35 flux. Be sure to read all the fine print. Aggressive fluxes can ruin boards over time if they are not washed off properly. – user57037 Aug 25 '22 at 04:33
  • Thanks. Changing the solder paste and flux would involve our contract assembler changing what they've loaded their machines with. I'd have to check with them, but that doesn't sound doable. – Dan Aug 25 '22 at 05:57
  • I think that nickel plating the whole PCB would make it difficult to solder the rest of the components that make up the BMS. Our PCB fabricator also recommended that we avoid using different plating methods for different region on the PCB, and to pick one for the whole board. – Dan Aug 25 '22 at 06:00
  • 2
    Have the strip tin or gold plated at the soldering end. – Tim Williams Aug 25 '22 at 08:43
  • _"The cell tabs are too big for soldering."_ - how big are they? – Bruce Abbott Aug 25 '22 at 13:31
  • Thanks Tim. I've asked the nickel strip supplier if they are willing to plate the nickel on one side. They are in a different time zone so I probably won't see their reply until Monday at this point. – Dan Aug 26 '22 at 05:29
  • Hi Bruce. The cell tabs are 12mm wide. 1 is nickel and the other is nickel plated. The heat and time required to solder those tabs directly to the PCB would be more than we are willing to risk this close to the cells. However, that's not a problem if we solder the strips to the PCB first, and then spot/cold weld the cell tabs to the strips. – Dan Aug 26 '22 at 05:33
  • I think the idea proposed by tim williams is your best bet. I do want to point out that everyone may refuse to change any part of their process. The supplier won't coat the tabs. The PCB assembly people won't hand solder the tabs then clean the PCB after. If everyone tells you "no, we can't do that" what are you going to do? Give up? In my experience, nickel IS difficult to solder to, even when it is very hot. So I think something will have to change. Or you could do a test run to assess and see if it works. – user57037 Aug 29 '22 at 04:23
  • 1
    Hi Tim, I got a response from the nickel strip manufactures and plating the strip will raise the MOQ from 10,000 to 200,000, and increase the price. I don't know how much the price is increased by because they won't quote that. I don't think this is an option. However, the manufacturer tells me that plating shouldn't be necessary. Our assembler isn't so sure, so I think I'll buy the 10,000 and just try it. – Dan Aug 29 '22 at 04:23
  • Unless there are other ideas. Any thoughts are welcome. – Dan Aug 29 '22 at 04:24
  • I think you should call a flux vendor to see if there is a suitable no-clean flux that will get solder to stick to copper. Are the nickel strips attached as a secondary operation or are they placed onto solder-pasted pads with pick-and-place machinery and reflow soldered? – user57037 Aug 29 '22 at 07:10
  • Thanks mkeith. The nickel strips are going to be placed with a pick and place machine on solder-pasted pads. The paste screening machine, pick and place and reflow oven are all linked together with a belt system. I don't think our assembler will want to load their machines with a different paste or flux. – Dan Aug 29 '22 at 07:21
  • No harm in trying I guess. But I guess I am also wondering what happens if it doesn't work. It sounds like nobody is willing to consider doing any of the things you are asking, so I am wondering if you have the right partners. Maybe if the first samples are a disaster, you can circle back and persuade someone to step out of their comfort zone a bit and actually find a process that works. – user57037 Sep 05 '22 at 07:15

1 Answers1

2

Our assembler originally recommended some special solder paste with a certain flux in it to bond the nickel strip to ENIG pads, but after a bit of trial and error, we didn't end up having to do anything different to reflowing other components. No tinning or any other special steps. It just worked reliably on every board.

Dan
  • 133
  • 8