9
\$\begingroup\$

Pretty noobish at electronics here. That said...

I want to attach a very thin 1080 spring steel/music wire (30 AWG or ~0.25mm) to a pcb with a thru hole, hopefully in an automated way. The choice of spring steel is for the mechanical properties.

It seems like soldering stainless steel is doable, see the below links. I think that applies to spring steel as well.

Though it looks like I'd need special acid (HCl) flux for this.

A few questions:

  • Does this seem doable?
  • Is acid flux necessary here? If yes, can that be done in a reflow oven or would this need to be done totally by hand?
  • I think I want the thru hole diameter to be as small as possible to roughly match the wire (Say 0.5mm), though I'm not super sure how small of a hole diameter I can get. Thoughts?
  • Is there any good way to automate this? I don't think I could do a pick and place type approach here, but is there a machine that could apply the right flux, put the wires in place, and then put the thing in a reflow oven?
  • If this is crazy and spring steel won't work, is there a better springy material?
\$\endgroup\$
8
  • 2
    \$\begingroup\$ Acid flux is incompatible with circuit boards. However, you could solder (even braze or crimp) the wire to a tiny copper or brass eyelet, clean it, then solder that using conventional flux / means. Bonus, connection is more rigid. \$\endgroup\$ – rdtsc May 14 '20 at 19:32
  • 1
    \$\begingroup\$ Yep, needs to be electrically conductive, forgot to mention that. \$\endgroup\$ – apollo901 May 14 '20 at 21:04
  • 1
    \$\begingroup\$ Generally throughole stuff is not soldered in reflow ovens, it's normally either hand-soldered or wave-soldered. \$\endgroup\$ – Peter Green May 15 '20 at 8:08
  • 1
    \$\begingroup\$ @JoelKeene 1080 is a high carbon steel, yes. And it is true that they have quite different properties. Especially the surface passivisation layer, which you would think would have a big impact on the flux required. In practise however neither can be soldered with a flux suitable for using on a PCB, but both can be soldered with a strong acid flux. So the advice works out much the same. \$\endgroup\$ – Jack B May 15 '20 at 19:46
  • 1
    \$\begingroup\$ Does it have to be soldered, or would it be sufficient to have a clamped connection, like around a screw under a washer? Which could make it easily field-replaceable, if mechanical failure can occur. \$\endgroup\$ – Andrew Morton May 15 '20 at 20:06
18
\$\begingroup\$

We solder stainless steel from time to time. It needs more heat than I'm really happy bringing to a PCB, and the flux (we use a HCl flux, as you say) will quickly eat the tracks off the pcb if you leave any residue. Our process is to tin the steel with the acid flux, then clean very thoroughly, then solder to the PCB in a separate operation. If you can buy the wire pre-tinned, all the better.

We can't use conductive epoxy for this for various reasons. If we could, we might consider it. But in my experience it comes a poor third place for mechanical strength, and good epoxies (i.e. Epotek H20E, not "silver paint for PCB repair") can be expensive and a little faffy to work with.

Another approach we've used is to crimp a ferrule to the end of the steel wire, then solder that to the PCB. That avoids the nasty flux, and gives a mechanically very satisfactory join. But it does take up a little more space.

\$\endgroup\$
6
  • 2
    \$\begingroup\$ This, I believe, is the right answer. It requires that you realize that it's not necessary to bond the solder to the wire at the same time you bond the solder to the pcb. If you split it into two separate processes the drawback of the first (acid flux) can be eliminated before the second. \$\endgroup\$ – WhatRoughBeast May 14 '20 at 21:17
  • \$\begingroup\$ Yes, good point about splitting out the processes. I particularly like the ferrule option, seems like it'd be more solid mechanically (though I could be wrong, worth testing both for my application). Any recommendations for a ferrule part? Ideally it wouldn't be too long. The 10mm length on these guys gives me pause: digikey.com/products/en/connectors-interconnects/… \$\endgroup\$ – apollo901 May 14 '20 at 21:31
  • 1
    \$\begingroup\$ @apollo901 - Those are generally for inserting wires into terminal blocks, not what you're trying to do here. Also, the plastic shoulder on them would likely melt (as they're intended only for mechanical crimping). What you probably want here is more along the lines of wire-to-board connectors for a single pin. \$\endgroup\$ – Connor Wolf May 15 '20 at 9:09
  • 1
    \$\begingroup\$ We 1) are small volume for these 2) have a whole lab full of mini-metalworking tools and 3) tend to have lots of copper tube in stock anyway. We just make ferrules by cutting 0.8mm copper tube into ~1.5mm lengths. So I can't recommend anything specific. But I would look at wire-to-board connectors as Connor mentions, or maybe male crimp contacts from connectors where the contacts and insulator come separately. \$\endgroup\$ – Jack B May 15 '20 at 13:38
  • \$\begingroup\$ Thanks, that's helpful indeed. This might risk going a bit far afield, but I need to fab this at scale. Do you have thoughts on how to get there? Seems like soldering tinned wire directly would be great - recommendation for a product or supplier? If we were to do crimping, is there any way to automate that or at least make it scalable? \$\endgroup\$ – apollo901 May 18 '20 at 21:26
8
\$\begingroup\$

Since you seem to be thinking about possible quantity production, consider sourcing plated wire which is solderable. This kind of material is used for such applications as battery holder springs.

You'll have to consider how the high temperatures in reflow soldering will affect the spring metallurgy. Soldering temperature is within the normal range of tempering for carbon steel (and much briefer), so I'm guessing it won't have a huge effect.

I don't have a lot of optimism about how sturdy a soldered carbon steel wire would be stuck through a PCB hole. I think the joint will tend to fracture near or at the wire-soft-solder interface if stressed. For small quantities consider silver soldering the wire to a ferrule (can be done easily with a propane or MAPP torch) and staking/soldering the ferrule into the board.

\$\endgroup\$
5
  • \$\begingroup\$ Good point about the metallurgy. From a quick test in my toaster oven it doesn't seem like it'd be too problematic, but I definitely won't bank on that, ha \$\endgroup\$ – apollo901 May 14 '20 at 21:34
  • 2
    \$\begingroup\$ Why the stainless wire - why not plated beryllium copper - you can have all sorts of platings and according to ngk-alloys.com/berylco_wire.html has excellent solderability. \$\endgroup\$ – D Duck May 15 '20 at 17:44
  • \$\begingroup\$ @DDuck Sure if cost is of little concern. Also, phosphor bronze is nice, and doesn't have the potential toxicity issues of beryllium. But for cost-sensitive applications plated steel strip and wire is very common, both for springs and for soft component leads. The electrical and thermal conductivity isn't as good as the others. \$\endgroup\$ – Spehro Pefhany May 15 '20 at 21:09
  • \$\begingroup\$ Do you have any recommendations for plated wire? I'd ideally be looking for 1080 music wire rather than stainless, for the springiness (or a cheap alternative w/ similar mech. properties, sounds like they're expensive though). A quick search on Mcmaster-Carr doesn't turn up anything suitable: mcmaster.com/music-wire/lubricated-1065-and-1080-spring-steel . Are there other suppliers that would be better for this? \$\endgroup\$ – apollo901 May 18 '20 at 21:15
  • \$\begingroup\$ @apollo901 These guys advertise it, mjwire.com/spring-wire/music-wire but if their MOQ is too high, which would not be much of a surprise, try China. \$\endgroup\$ – Spehro Pefhany May 18 '20 at 23:15
3
\$\begingroup\$

Rather than soldering, consider using a conductive epoxy.

Example.

Steel can be soldered using tin/lead solder and the correct flux. The amount of heat this needs to work is higher than most PCBs can withstand without de-laminating.

\$\endgroup\$
2
  • \$\begingroup\$ Any thoughts on the comment from above (Jack B's answer) that conductive epoxy tends to be mechanically not very strong? The wire is meant to serve as a bit of a poor man's spring, so it's important that it's firmly rooted to the board. \$\endgroup\$ – apollo901 May 18 '20 at 21:17
  • 1
    \$\begingroup\$ The one I linked in as an example may not be the best for you, however, they do make some very strong epoxies for this. Something from the automotive line would be good since they are made to take abuse in harsh environments. If it were me, I would call Henkel (Loctite) or MasterBond, either one of those places will put you on the phone with an engineer that can recommend a good product. I have had really good experiences with Henkel. \$\endgroup\$ – evildemonic May 19 '20 at 14:58
1
\$\begingroup\$

If you have aspirin tablet at home then put tip of the still wire (spring) on the tablet of aspirin and heat it up with soldering iron. Be careful as evaporated aspirin is harmful for health. Once the tip of the wire is tinned, solder spring into the PCB. This approach was used by radio amateurs for decades.

\$\endgroup\$
1
  • \$\begingroup\$ Can anyone confirm that this works? Very interesting. \$\endgroup\$ – Drew Jun 9 '20 at 22:17

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.