Very good method Barkuti,
Link those small solder balls when available.
For solder blob I use large flat horse shoe tip.
Then I just sand battery surface, flux it, and when large horse shoe tip is loaded with solder i just make fast 1- 2 second down-up move.
luminarium iaculator, this is my current sale link of choice: http://www.ebay.com/itm/Roses-Rose-alloy-metal-Lead-Bismuth-Tin-low-temperature-melting-90-gramms/132292315369
There are small formulation differences, or so it seems. Melting temperature shouldn't vary much, though.
Cool stuff, I like it. :-)
Cheers 
one place i can see this stuff being useful is making a thermal fuse in a pack.
solder the interconnects in spring tension.if pack overheats alloy melts and opens circuit.
like a grasshopper fuse.
this alloy is often used in sprinkler heads.it melts and activates the valve.
What about for making the individual cell fuses—that is what i was considering to use it, but was concerned about loose hot solder dripping about in the event that one cell went thermal. Maybe a small blob to hold the cell fuse wire would not be a significant amount to worry…?
If you assemble a battery pack, you can hang a small weight from the interconnecting wires. This results in an open circuit condition if the batteries reach 93 - 96°. Excellent from a safety standpoint. ;-)
snakebite, :THUMBS-UP: .
Cheers ^:)
Barkutti, thanks for your video and explanations!
I see myself in the possibility to have to solder the top of some unprotected 18650 batteries, with not extras, apart from the solder.
I have flux (JBC), I have solder wire (Sn60, 0,8mm), but no pre-made blobs!
Is it “safe” or appropriate to just make some blobs with the wire I have, and them proceed as you did (add the flux, put the blobs on the top, and let it “melt”)?
I’ve never put solder on the cells, so I don’t know if “normal” components will do the work correctly…I didn’t want to spend money in the blobs or in other “tops” for it…
Thanks in advance!!
One rather unorthodox method I use is to melt a hanging blob of solder on the tip of the iron, put organic flux on the surface I’m going to hit with it, then hit said surface with the molten solder. It’ll sizzle away and hopefully quickly transfer enough heat to cook the surface and get the solder to stick.
The conventional method of using the iron to heat the surface and melt the solder draws away that much more heat from the tip, potentially taking longer, or even cooling down everything to the point where you’ll end up with a cold joint or one that just doesn’t stick.
Heating up the surface takes X amount of heat. Melting room-temp solder to its liquid phase (and latent heat of fusion) takes Y amount of heat. You have to hope that your iron has enough temperature and thermal mass to at least exceed X+Y by a good margin.
Take Y out of the equation, and you need only X amount of heat!
MascaratumB, solder in blob form is not a requirement, it is just the way the solder I use comes as: Rose's metal / Rose metal (Bismuth, Lead, Tin alloy) 75 g. LOW MELTING SOLDER @ zhuykoff
You'll need appropiate flux for steel (aggressive, acidic), but you can also use some fine grain ”wet” sandpaper to slightly sand down the surface to be soldered over.
With regards to using standard Sn60Pb40/Sn63Pb37 solders, the problem here is you'll need to raise the cell surface temperature nearly 100°C higher, which is a lot. What this means is if you don't have a powerful iron and a fat tip to transfer the heat fast better forget it. You can try, but if in at most ≈5s your solder hasn't properly pooled yet holding onto the steel surface, I'd say better give up. Alternatively, buy some Rose's metal, makes your life muuuch easier and safer. 
Cheers :-)
Thanks for the answer Lightbringer! My iron is 30W, and so far ir only had 1 or 2 problems with some solder. I hope I can do a similar process like you, or like Barkutti!
I’ll try make a clean job not to damage the cells! 
Gaaaah! Get a hotter one. It’s a one-time investment, so make it good. I got an AEIOU variable-temp iron, and crank it up to maximum when doing that kind of thing.
Hit it fast, hit it hard, get in and get out. You’ll never have to linger for 20min cooking everything else in creation just to make one little solder-blob.
I had a 30W iron long ago, tried to solder with Sn63Pb37 onto Ni-MH AA cells unsuccessfully: cold joints and damaged cells.
Also, while Lightbringer's X + Y tip is fair please note that, in this case, heating up the cell takes away most of the heat and, in any case, you should at least pre-moisten the iron's tip in some amount of solder in order to maximize contact surface and thus heat transfer.
Cheers :-)
The big hanging solder-blob is moisture enough.
Thanks for the answers, to both 
I guess I will try first in some old batteries (from a powerbank) and see how the solder iron behaves! I didn’t want to buy another one, budget is becoming short for some things on the hobby
I’ll experiment! Thanks again
Well, just tried it and it was not the most perfect surgery, but so far the cell is working
Further work will be needed, namely to combine the flux and the soldering time over the battery! I’ll learn!
Again, thanks to the answers!
Huh, it looks like in a local shop it costs about $7 for 100g pack.
I also used different brass buttons to add to the top
I found those on Aliexpress which might a nice option for soldering
Definately cheaper than the brass buttons from Fasttech or Kaidomain
https://www.aliexpress.com/item/50pcs-lot-4-5-6-7-8-10m-Yellow-Copper-Round-Flat-Metal-Spacer-Beads-Fit/32799436603.html
https://www.aliexpress.com/item/50pcs-lot-5x2-5mm-Vintage-Original-Brass-Loose-Spacer-Beads-Round-Flat-Beads-Nepal-Beads-fits/32807367148.html
Even gold plated ones
After reading this thread i bought some 1 lb ingots (~$15) from rotometals. i was hoping i might form it into wire such as we use for solder, but it is much too brittle and not suitable for wire. It melts at 208F and does not contract or expand between liquid and solid phases.
It melts and drips off a soldering iron quite easily, and it is easy and fun to make the “blobs”. i used a hacksaw to cut an ingot in half to give to a friend, but in hindsight it would have been easier to ‘cut’ with a hot iron. He is using a few ounces to heat up and pour into his antique rifle chambers, then press it back out with a wooden dowel. From this he can measure the diameter and wear in the bore and see the erosion of the rifling edges to determine if the barrel is okay or not.
If i hold the iron high and let it fall the drops splatter into thinner and flatter pieces than the “blobs”. These might be easy to mail by taping them to an index card, then put it in an envelope with regular postage.
That’s how you get cold joints.
x and y have to be at least as hot as the melting point of the solder to fuse both materials together.
Otherwise the solder will just flow into the surface imperfections which is more like glueing the solder to the cell.
Mechanical force can tear the solder off and the electric resistance is much higher.
And how is that technique defective?
A wet/molten solder-blob hanging off the iron is already at temp, and you’re not relying on pressing a solid to a solid with small contact area, but letting the liquid flow all over the surface and provide lots more contact area.
The other surface (eg, button-top) heats up more quickly, and unless it gets up to temperature, the solder simply won’t wet to it.
So if you have a ball of solder just stuck to it, of course it’ll be a lousy connection.
But what I’m specifically referring to is having a HOT iron with equally HOT blob of solder, quickly heating up the other surface, wetting to it, and then finally cooling down.
What Y is, is the heat necessary to initially heat up the solid solder to its melting point, actually melt the solder (LHF) and further bring it up to temp, enough to wet and stick to the button-top.
It’s the exact opposite of someone saying, “hot water freezes faster”. No, hot water has to drop in temp, to the FP of water, and then freeze. Ice-cold water already has that head-start.
You don’t press a solid against a solid the tip is always a bit wet.
The tip is used to heat up the object not the solder. After heating up the object you melt the solder directly on the object.
That’s the only way to guarantee that the object is hot enough.
To speed that process up you can bring the solder wire and tip together as soon as the solder starts to melt and melt on the object and tip simultaneously. This will than melt bigger amounts faster.
Solder will wet to an object that is not hot enough, otherwise cold joints due to cold irons weren’t possible but that’s a well observed phenomenon.
Also your flux is burning up while you are forming that blob on the iron.