Best practices for soldering 18650?

Hi guys,

I would like to build my own 18650 pack. I was thinking of Samsung 25R cells and 2x5S packs in series with 3 or 4 cells in parallel. This way I could balance charge them separately with my 6S charger and would not need a BMS.
However for a single battery it is not economical to buy a spot welder so soldering is my only chance.
After reading about nickel strips, or copper welding rods for the connection of cells I was wondering what the actual best method for soldering 18650 packs is. Can you simply use 10mm^2 cables, remove the insulation at the position of a cell and solder them onto the cell? Or use 2x6mm^2 speaker cables?
Also I read about the danger of the cells becoming hot - can that be solved by dipping the rest of the cell into a liquid (distilled water or isopropyl) to allow for fast heat dissipation while soldering the cable onto the other end?

If there is a tutorial on how to solder 18650 packs let me know - so far I only read about finished products and not so much about the actual process.

1 Like

Usually easier to solder with copper rod. To minimize the contact time of your soldering iron on the cell, make sure the surface is scuffed up sufficiently and you use plenty of flux to allow for fast solder flow. To keep it cool i would suggest using compressed air before and after making the soldering connection.

Also, one other thing, it is better to have a good quality high wattage iron with good thermal capacity so it can deliver the heat to the joint quickly so you don’t have to hold the iron to the battery for ages and let the heat seep into it, causing damage to the battery.

Safety wise, it really isn’t that dangerous, especially because you are buying good quality cells. I have soldered 100’s of used laptop cells without any issue.

1 Like

I once saw in a video that you should never blow on a fresh solder joint because the rapid cooling would cause wrong crystallization of the solder and decrease the quality of the joint.
I usually don’t care for my simple joints with LEDs and stuff but was wondering if this is an actual issue when soldering batteries - especially for the type of high currents we use?

1 Like

I don’t mean rapidly cool the joint. You cool the cell then work your way up to the top, and by then the joint will be cool enough for the direct cold air. It is the internals you need cold.

The tip of the cell doesn’t transfer a ton of heat to the cell due it it not being 100% connected, as you would be able to see on the top around the positive terminal.

What about freezing them beforehand or would that damage the cell?

So blowing on a joint is indeed not a good idea? I will probably be terrified to just leave the joint to cool on its own when there is a battery involved :wink:

Blowing on the joint is bad, yes.

Your goal is to have the cell stay cool, not frozen and to allow the joint to cool normally.

No need to scuff the cells, you are soldering nickel to nickel. Use a good liquid flux and pre-tin your cells with a little pool of solder hitting them with a wet sponge after removing the iron. We only scuff metals that have oxidation, remember soldering is not a mechanical bond.


And please use at least an 80 watt iron to do this. You have to get in and out quick. If you use a little smd iron you will need to let the iron dwell for much too long and the cell will get hot.

I was going to use a 85watt weller iron (450°C max. temperature).
Also I am not using nickel strips but either copper desoldering braid, thick cables or welding rods.
What is the sponge for? I thought rapid cooling is not ideal.

It not that rapid, it will be fine. Don’t scuff the cells, do scuff the bare copper. I wouldn’t use anything too rigid to join the cells. Something flexible is better, more resilient.

I use a Weller 85w WD-1 soldering station and it gets the job done WELL!! If you are using the WP-80 or WSP-80, you will most definitely want to get a nice fat chisel tip. retains temp well and transfers loads of heat quilckly.
I use the Weller LTD 4.6mm chisel tip.
Also, it should go without saying, but you gotta tin those leads up before you try to attach them to the cells. 2-3 seconds per joint max should get it done. :v:

I do my battery’s with a 200 watt soldiering iron. In and out fast, and afterward you need to test your weld by tugging on lead. I’ve made about 8 battery packs this way.


I’m also interested for the first time of making my own 18650 battery. I plan on running dual R-Spec 190Kv motors and would a 10S3P be good enough for a 170lb rider like me? I’m looking for 10-15 miles which I’m sure will be quite tiring after that distance since you’re constantly leaning on side or the other. Anyhow, I’ve been searching and what cells are good? I mean I know LG and Samsung or Panasonic/Sanyo are the best but what specs are good? Is 10A discharge enough for this setup? I will also be soldering them since I don’t have a spot welder :frowning: Thanks in advance everybody.

You’re pretty much building a raptor, so yea, you should be able to get that mileage no problem. As for cells, Enertion uses Samsung 25R’s for their space cells. I’ve also heard good things about a123’s LiFePo4 cells if you want to go that route. Legit Panasonic and LG cells are a good choice too. You’ll probably want 50A+ continuous (rating for the VESC) at least, which means a 3P pack of 25R’s would suffice (60A continuous). Make sure you get a 85w+ iron if you’re going to solder as the less time in contact with high heat for the battery, the better.

1 Like

Thank you. I was looking at the Panasonic NCR18650PF which stated that it would have 4000 charge cycles. But it does say each cell is 10A only. Would that be OK for my setup if I built it 10S3P? Perhaps not since that would be only 30A or so continuous? I’m sorry for these dumb questions but I guess it’s never too late to learn. Thanks guys.

you’d probably want a different cell that can handle more continuous discharge. Also where did you read 4000 cycle count? That’s abnormally high for a li-ion cell.

I saw it on eBay. Just exaggeration? Just wondering I was searching for the Samsung 25R and states that the cycle life is only 250? That’s very low for any Lithium Ion battery isn’t it? I mean that’s like Lithium Polymer low! Can anyone expand on that? I know I like LifePO4 since I got an eBike (Prodeco) with it. They are not as energy dense but safe all around. A little pricey but hopefully they will come down with time…as with everything else.

250 is about the right cycle count for 70%~capacity. Keep in mind this is done in a test environment discharged to 2.5v and recharged to 4.2v. This puts a lot of strain on the battery so if you treat it well it should last a good amount more than that. It doesn’t flat out die at 250 cycles, just losses a bit of capacity over time. LiFePo4’s are heavier and less dense, but provide over 2000 cycles.