Soldering (batteries, etc) … with Sn42Bi58, Rose's Metal and other low melting solders

Hello!

Major update. Received my Rose's Metal from Zhuykoff, and made a little test:

Pools fast and easy over a clean surface, of course. Used Goot Super Soldering Flux.

Über-pleased in amazingness!

Cheers ^:)

Interesting. Have you done any strength testing yet? I have to say, it’s not pretty though :smiley:

Please note that the wire/sheet joints were already done with standard Sn63Pb37. The thing is just a powerbank circuit board testing assembly, on which I previosly was using neodymium magnets and hot glue to connect the cells.

The Bi50Sn18Pb32 joints are quite strong, no measurements done but suffice to say I could tear the copper sheet trying to rip it from the cells with “ease”.

The great news is the cells barely heat up (if barely at all) when soldered with this stuff.

Cheers ^:)

well, with ANYTHING, and thats any device… most people “concentrate” on what somethng doesnt have, what it doesnt do, what isnt good about it

when i remember to, i try to concentrate on what it has, what it will do, what is (might) be good about it.

i have previously put off soldering onto batteries, simply because i have a chinese french fry, not even a chinese potato, lol… and on top of that i admit to not being a great soldering guy, there i said it.

my combination of having almost the worst tool at hand, combined witha lack of experience to even try to work-around it? makes me pass…

…now, seeing his quick success with rose metal? makes me want to try it… makes me wonder if i could solder a emitter to a thick copper DTP with regular method, then maybe use THIS stuff to solder the pill to the copper board…

i mean, for SOME people, with some tools, and their level or lack of experience? this has possibilities.

===

on a “red-yellow-green” system of ratings? I give this… yellowgreen, or, greenyellow, cant decide at first blush until i try it. Just doesnt look like a “red light”.

my ALTERNATIVE is to buy a better soldering station, and then gain the experience i need, to do somethng i rarely do and would like to dabble with…

sedstar, I presume you'd like to be successful in soldering MCPCBs to copper/brass pills. Having seen how fast a rolled-up (around an old iron's tip) copper sheet turns into a hard to wet black mess of oxidized metal, I bet oxidation may be problematic when heating a chunk of copper/brass over a pan.

So yes, this may be a solution. If you believe its fusion temp is a tad low for your purposes, it can be easily raised by mixing it with pure tin and/or standard cheap lead-free solder (Sn99.3Cu0.7). I say this because, according to the Solder article on the Wikipedia, Sn48Bi32Pb20 fuses between 140 and 160°C, an alloy with a nearly equivalent amount of bismuth and lead vs the Rose metal alloy I bought (Bi50Pb32Sn18).

Cheers ^:)

You should try this out, it's soo amazingly easy!

Cheers fellows

Hey Robinjack, I think it would be a good article if you JUST GET LOST

Hello Barkuti,

I am just beyond newbie level, and I started to slowly gather required materials for LED reflowing (not planning to make solder joints on batteries in the near future).
I looked into some LED datasheets, and I have seen, that e.g. Cree recommends to use only lead free soldering paste to make the reflow.

I ended up buying this Sn42Bi58 alloy:
https://www.aliexpress.com/item/high-quality-100g-Sn42Bi58-SD-528T-low-temperature-SMT-LED-lead-free-SMT-solder-paste-Melting/32824147086.html

My plans including reflow for Convoy L6 lights with selected (better CRI) XHP70 emitters, and for various other lights with 3.5×3.5mm format emitters (XP-L, Nichia 219B/C).
In the recent months, I have read here many cases about someone killed/crippled XHP70 emitters with too much heat and/or steep temperature deltas during reflow (especially the newer XHP70.2), so I assume, it is clearly safer to use Sn42Bi58 instead of Sn63Pb37, because of the 40+ °C lower melting point.

Should I be concerned about this, or it is just a minor factor, and the fact, that presumably the LED manufacturers only recommend lead free solder paste for most of their LEDs means, that it is strongly recommended to use them?

I just ordered some Rose’s metal from here, in the US:
http://www.ebay.com/itm/282261610946

I’m building a pack of 24 Shockli 5500mAh 26650s for an electric bike project. Soldering with typical lead/tin solder can probably be done without significant harm to the cell, but this seems like a better solution. While soldering with standard solder the active material close to the negative end of the cell probably gets close to 200C, and I would imagine this does do some damage locally.

I've never used Sn42Bi58 adam7027, should work nicely but this is just I guess. I'd personally use Sn63Pb37 reflowing over electric induction stove, which I can more or less easily monitor with an IR thermometer to keep it hovering slightly over 200ºC.

To solder with Rose's Metal over steel I use an aggressive liquid flux with helps a lot (Goot's Super Soldering Flux), EasyB.

Cheers ^:)

Based on the material composition it seems to have much worse heat conductivity as well as electrical conductivity.
200C won’t damage a battery if the terminal only gets like that for a few seconds.
There are solders that melt at about 150C too, which are meant to conduct heat and electricity properly.

How much worse is and how does it impact overall performance in our parameters of interest?

”Much worse” is a relative term. A thermal paste is also “much worse” than soldering, but its usage is way more pragmatic for lots of applications.

With an aggressive flux and Rose's Metal it is super-easy to solder copper sheet tabs or wires to cells with massive contact surfaces over their steel (a much worse electrical and heat conductor, by the way) terminals. I am pretty sure this method beats industry standard nickel tab spot welding practices in electrical and thermal properties by a lot. The key here is to determine how vibration resistant is this stuff, so a quick down to earth advice for those after this in mobile applications is to improve solder contact gripping with a file, for example, and avoid blindly relying in the solder contacts as structural support.

To finish off, Rose's Metal and close alloy variants (like my Bi50Sn18Pb32) are relatively affordable if you know where to get them, of course. I bought 75g for $10 (and got an :-) 85g bag, thanks zhuykoff).

Cheers :-D

Cheers ^:)

You stated that as if it was a fact, but it is just your speculation, correct? It’s possible the damage would be minimal, but 200C is quite high and outside the stability window of Li ion chemistry.

Well 50% of it is bismuth, which if you look it up has about 10x worse conductivity compared to tin, so…

The 200C is only at the terminal, the farther from the terminal the lower the temperature.
If it’s only there for a few seconds there is a high temperature gradient leading to not very high temperatures inside the battery.
Premade battery packs are often spot welded rather than soldered too, which gets to much higher temperatures than 200C with no damage done because it’s fast.
You can also look up specs for soldering different types of batteries which tell you the safe temps, such as “Max. 2 times reflow soldering (Max. 260℃ )” for a particular one from a random google search.

which substance to clean the glue after use Sn42Bi58 ?

I thought I would show off the battery pack I’m building for an e-bike.

24 shockli 5500mAh 26650 cells in 6s4p configuration.

The Rose’s metal has worked great. I’m not sure I fully took advantage of the lower melting temperature as I still had my iron set to around 300C. Don’t really know what setting results in the least heating of the battery active materials since you still want the tip temperature to be relatively high to heat the surface adequately and quickly.

I used regular flux that I had and the Rose’s metal wet the cell casing and copper stuff similarly to regular lead/tin solder.

Barkuti, what are your thoughts on iron temperature setting in this situation? One question is what temperature is needed for the solder to wet the cell casing quickly? If that temperature is not lower for Rose’s metal then the melting temperature is not relevant. However, after the solder has wet the cell terminal, the lower melting temperature does make it easier to solder the wire to it.

I think temperature controlled irons suck because we cannot fully take advantage of the iron's heating thermal inertia while at the same time having the temperature limit set low, but I guess this is no big deal (how about a +t°C/+t°F button?).

I also set my tip around 300°C, this is not really a problem as cells powerfully sink heat and anyway since the low melting solder liquifies almost instantly this also helps a lot in raising the cell surface temperature on par, so getting good joints is a piece of cake.

Your assembly looks pretty neat, at least compared to this little botch of :-D me but heck, I do not have that pretty good looking copper braid:

Bosch PSR 1200 battery pack overhaul.

Cheers ^:)

Is it thick copper braid or thin solder wick?

With thin strand solder wick it’s pretty easy for high current to melt it. In my experience.

Mmm, I was also going to comment on what JasonWW is pointing out: to add more conductor cross section. It may not be required but I like to play it safe, using ≈AWG18 cross section for each 10A of continuous current.

The sheets you see above in my battery pack are of 0.1mm, 2x Sn63Pb37 stack soldered.

Cheers

I am concerned about the current capability of the wire. The braid connecting the cells in series is relatively thicker than typical solder wick. It is equivalent to AWG18 based on weight of copper per unit length. Sustained output should be less than 5A per cell, but bursts could be up to 20A per cell. The braid is probably adequate but I should make it thicker.