8
\$\begingroup\$

Usually my circuits are full of very fine-pitch SMD components. I solder the prototypes manually, which takes a lot of time. Good tools and high-quality solder can speed up the process.

I prefer using leaded solder, as it flows better at relatively low temperatures. This way I can prevent my components from overheating. Leaded solder is not allowed for commercial products, but is okay for prototyping.

There are several types of leaded solder wire on the market. I'm trying to find out which one is "best". Let's define "best" as follows:

  • Low melting temperature (prevents overheating components).

  • Good wetting of pads and pins.

  • Preferably contains some flux, so one doesn't have to apply it all the time externally.

  • Very fine diameter for soldering small components (like LFCSP package, 0402 or even 0201 resistors, ...)

  • Price is no issue.

I have several questions:
 
 

1. Tin - Lead alloys
I read on Wikipedia that the Sn60Pb40 solder is very popular for electronics (I agree, I have used this one so far). Wikipedia also mentions that Sn63Pb37 is slightly more expensive but also gives slightly better joints.

What do you think about Sn60Pb40 vs Sn63Pb37? What is actually the difference?
 

2. Exotic alloys
But these are not the only solder alloys. More exotic combinations - containing tin + lead + silver and even with gold exist.

Will these exotic combinations change the properties?
 

3. Bismuth and Indium alloys
Some of you made me aware of Bismuth- and Indium- based alloys. I've dedicated a new question to cover them: Bismuth or Indium solder - what would you choose?


NOTE: I use a solder-smoke extractor.

\$\endgroup\$
  • 5
    \$\begingroup\$ You forgot one important property: environmental effects. Lead based solders are a no-go in the EU for being toxic and hard to handle in a way that doesn't expose the employees of recycling companies to health risks. Hence lead based solder mostly disappeared in the EU and most of the countries that export to the EU. \$\endgroup\$ – Attila Kinali Sep 20 '17 at 16:53
  • \$\begingroup\$ If you want really low temperature solders, look at indium alloys. There is even a company called Indium Corp that specializes in them. They have solders with melt temperatures down to ~38 C. \$\endgroup\$ – The Photon Sep 20 '17 at 16:54
  • 2
    \$\begingroup\$ @ThePhoton , wait a minute, I get confused. 38°C? Are you serious? I read in the datasheet of the Bismuth-based solder on DigiKey (part no. SMDSWLTLFP32-ND) that the melting point is 138°C, not 38°C... \$\endgroup\$ – K.Mulier Sep 20 '17 at 17:05
  • 3
    \$\begingroup\$ Even if you use lead based solder just for prototypes, anything that touches that solder tip immediately needs a toxic waste sticker in most jurisdictions I am aware of. Unless you really have a need for leaded solder, I would recommend to use lead free. \$\endgroup\$ – Attila Kinali Sep 20 '17 at 17:44
  • 3
    \$\begingroup\$ Unfortunately lead free solder makes really bad quality joints when manual soldering. It doesn't flow as well, wetting small pads can be difficult, often joints end up being weak and matte, instead of strong and shiny, etc.. Lead-free solder is best used with a reflow oven, in paste form. I always buy leaded solder wire. I think the environmental aspects are irrelevant in prototyping. Now sure, the fumes are toxic. Be careful. Not any more dangerous than working with any other toxic chemical vapor. Leaded solder consistently gives better results with less work or rework. \$\endgroup\$ – Drunken Code Monkey Sep 20 '17 at 22:22
10
\$\begingroup\$

Sn63/Pb37 is better than 60/40 because it is a eutectic alloy. That means it has the lowest melting point of any Sn/Pb alloy, and it solidifies relatively abruptly at one temperature rather than over a range. Generally both are advantages or neutral.

Combinations with small amounts of (say) gold tend to be for reducing the tendency of solder to dissolve the material (gold in this case).

Many solders these days avoid the use of lead and are often mostly tin with other materials such as copper, bismuth, silver etc. This is done to reduce the toxicity of electronics that finds its way into the waste stream. In my experience it is worse in every way compared to tin/lead solder except perhaps in applications where high melting temperature is important.

Flux another matter- there are a number of different types.

If RoHS compliance (and toxicity) are of no concern, 63/37 Sn/Pb solder with RMA rosin flux is an excellent choice, and is good for high reliability applications. Fine for hand soldering or reflow.

For production for world markets, it may be necessary to use lead-free solders with more finicky temperature profiles and inferior performance. Sometimes water soluble or no-clean fluxes are acceptable, depending on the product and how much it might affect the process (and possibly the product functionality).

\$\endgroup\$
  • 4
    \$\begingroup\$ Most important about eutectic alloys: As it solidifies, the composition of the solid portion is equal to the composition of the liquid portion. So you don't get a jumble of different-composition regions (with different physical properties like TCE) in the solidified material. \$\endgroup\$ – The Photon Sep 20 '17 at 16:51
  • \$\begingroup\$ @ThePhoton Ah, that's likely the reason the solidification is more abrupt. Thanks. \$\endgroup\$ – Spehro Pefhany Sep 20 '17 at 16:58
  • \$\begingroup\$ Yes, that too. Because the liquid composition doesn't change as the sold starts to form, the fusion temperature doesn't change. \$\endgroup\$ – The Photon Sep 20 '17 at 17:00
  • \$\begingroup\$ I like the answer of @Minho below. He points out that the eutectic 63/37 alloy solidifies quicker, which is not good for automatic alignment of SMD components. \$\endgroup\$ – K.Mulier Sep 20 '17 at 17:00
  • \$\begingroup\$ @K.Mulier Self alignment is quite complex- involving wetting and surface tension. Leaded 63/37 solder is generally better than lead free. See, for example Study of the Components Self–Alignment in Surface Mount Technology K. Dušek1), M. Novák1), A. Rudajevová2) \$\endgroup\$ – Spehro Pefhany Sep 20 '17 at 17:21
4
\$\begingroup\$

Solderability has more to do with oxidized copper surface and choice of flux to reduce oxidation and tombstoning.

I recall the Eutectic or lowest temp solder mix is 63/37.

In some cases solder contamination with antimony can result in poor joints.

But the temp differences are minor compared to; surface prep, cleanliness , choice of flux , storage temp of paste, open air time, shelf life; pad design for reflow, board preheater (frying pan or IR oven, or ?) and hot air or radiant heat profile curves with thermocouple and lack of turbulence but smooth flow.

These all affect wetting, burn rate of volatiles, surface tension, shorts, opens, tombstone, etc. and be sure pad design is optimized for reflow, and thermal profile.

Also ensure DRC has been done before design is sent out and checked by board shop.

\$\endgroup\$
  • 1
    \$\begingroup\$ Thank you very much. What do you mean by: "In some cases solder contamination with antimony can result in poor joints."? How can this happen? \$\endgroup\$ – K.Mulier Sep 20 '17 at 16:29
  • \$\begingroup\$ supplier quality.. Our company once had a chemist check solder in the 80’s routinely for antimony and other contaminants when mainframes had fine pitch board that looked like Remington shaver foil outer layers with 50 copper layers in between and huge dumpsters for shopping cart liquid gold immersion plating. I highly recommend you use some type of immersion gold or equiv plating. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Sep 20 '17 at 16:32
  • \$\begingroup\$ Gold plating contacts is actually worse than run of the mill HASL. To plate gold to copper, you need a nickel layer in between, otherwise copper migrate through the gold and ends up corroding it. Nickel is not a very good conductor, and so gold plated pads generally have higher resistance than plain copper. The only reason to plate contacts with gold is to prevent corrosion in the long term, but you can already use conformal coating in spray for this. Gold also dissolves in tin, which changes the composition of your solder, which affects joint quality. Immersion silver is best. \$\endgroup\$ – Drunken Code Monkey Sep 20 '17 at 22:35
  • \$\begingroup\$ Thanks DCM, I was actually thinking ENEPIG but said IG, yes ENIS and IS are very good. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Sep 20 '17 at 23:18
3
\$\begingroup\$

From a practical perspective, the main difference between the two is that the 63/37 is a eutectic alloy. In short, that means that it has a single melting point and not a plastic range like the 60/40 solder.

For prototyping and hobbyist use, it's really more a question of preference. If you are soldering wires together for example, 63/37 is easier to use because it will solidify faster and not cause cold solder joints. However, if you are soldering surface mount components on a pcb by hand, then the plastic range of 60/40 can help as that allows the components to "snap" into place.

\$\endgroup\$
  • \$\begingroup\$ Interesting. Can you dig deeper into this "snap into place" thing? I didn't know the two alloys would be different in this particular way. I like to know more :-) \$\endgroup\$ – K.Mulier Sep 20 '17 at 16:40
  • 1
    \$\begingroup\$ Because many SMD devices are so light, the surface tension of the molten solder will force them into proper alignment. This can only occur in when the solder is in a molten state. A non-eutectic solder alloy will have a wider range of temperature where it will be molten and this is important when you are soldering manually. \$\endgroup\$ – Minho Sep 20 '17 at 16:56
  • \$\begingroup\$ That makes sense. Thank you for pointing this out :-) . Do you have experience with Bismuth-based or Indium-based solders? \$\endgroup\$ – K.Mulier Sep 20 '17 at 16:58
  • \$\begingroup\$ I have used Chip Quick Low Melt that has Bismuth in the alloy. The main reason to use this is for desoldering heat sensitive components or connectors. \$\endgroup\$ – Minho Sep 20 '17 at 17:01
  • \$\begingroup\$ Oh, okay. But how about soldering instead of desoldering? \$\endgroup\$ – K.Mulier Sep 20 '17 at 17:02
2
\$\begingroup\$

"Very fine diameter for soldering small components (like LFCSP package, 0402 or even 0201 resistors"

This implies that you're talking about solder wire, as opposed to solder paste. Henkel Multicore products are my go-to.

My short answer, personally: Henkel Multicore Sn63/Pb37 0.38mm Crystal 400. (Digi-Key 82-117-ND).

Henkel Multicore Sn60/Pb40 0.35mm Ersin 362 (the latter name is the flux composition) is also alright but I like the 63/37.

Some of the more exotic alloys, such as Bi58/Sn42 (or Bi57.6/Sn42/Ag0.4) are obtainable as solder paste but don't really exist as solder wire.

These Bi alloys are attractive as they are a low-temperature eutectic alloy that is Pb-free. Many of the disadvantages generally associated with Pb-free alloys stem from higher working temperatures.

Personally, I do often prefer Pb-containing alloys for small-scale lab use, like you. Obviously, if you need RoHS-compliance for your product then you need Pb-free, and if you're manufacturing your PCB assemblies on any industrial PCBA line then they will all be set up with Pb-free alloys and temperature profiles anyway, so you use it anyway even if you do not strictly need RoHS.

ChipQuik make solder pastes in a range of different alloys, including the 138C bismuth alloys, in small packs compatible with small R&D or hobby users.

If you're reworking a board that already has solder on it, it's best to use the same solder originally on the board, otherwise you get an "in between" alloy which can have unknown metallurgical properties. This is particularly important with the low-temperature Bi/Sn alloys which can dramatically change their melting point and mechanical properties if a little bit of accidental Pb "doping" is present.

Sparkfun sells a Sn96.35/Ag3/Cu0.5/0.15Sb alloy which they claim is great - but it seems like a weird alloy! It's close to Ag03A, but it's different, and it does not appear in the Great Wikipedia Table of Solder and Solder-Like Alloys.

\$\endgroup\$
  • \$\begingroup\$ Very interesting answer! Thank you for including the DigiKey numbers. \$\endgroup\$ – K.Mulier Aug 30 '18 at 9:08

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.