soldering vs spot welding debate

[999zip999]

The art of soldering with 80watt, 100 fat tip.
1. File, clean and tin the tip.
2. Pre tin wire, buss bar. Just a small touch of flux.
3. Lightly flux cell apply iron then apply damp sponge. Can be on a third hand above cell or spring loaded. Maybe foot activated sponge on a stick.
Light on flux. Use gasket on pos end ( before) . Careful of cold solder joints.

 

[999zip999]

Tomjasz what ampage are you going to use that pack at ?

 

[tomjasz]

Tomjasz what ampage are you going to use that pack at ?

Which?

 

[Overclocker]

[youtube]0ZRwMXL-Rvs[/youtube]

here i applied the iron for LESS than 1 second. the joint is cool to the touch in 2 seconds.

spot welding only takes milliseconds. i got a video as well, shot at 400fps:

[youtube]gOUbVF2ef_o[/youtube]

spot welding definitely heats the cell a LOT LESS overall. but applies more intense heat on those 2 spots. but it's done at the bottom where the metal can is thicker so it's not a problem. try welding to the side to see what i mean

 

[Kapalek]

Soldering for DIY is the way to go in my opinion.

One very importand thing is to remember that when you want the solder to grip properly onto battery surface make sure there is sufficient heat at point of contact for the solder to grip. It there is not sufficient heat on the point of contact nothing will stick properly.

 

[999zip999]

That looks like an experiment.

 

[rojitor]

 

[Robocog]

sand both nickel strip and batt ends. Tin both WITH flux. Blow on batt end immed. Put nickel strip on batt end, apply soldering iron and feed solder into it. IMMEDIATELY it will flow. You put more flux right? Then, mash down with a metal object, to get max contact. This will immediately start to cool joint. Then, blow on it.

I did this and it was SO cool within microseconds, that it was COOLER than spot welding. The contact area is broader. I used the thick nickel, so no doubling up.

This is a fine way to build battery. B UT, you cannot make any money doing it this way.

You cannot just apply soldering iron to the metal without melting the solder into iron. THIS IS KEY. It will go WAY TOO SLOW if you don't.

 

[tomjasz]

That looks like an experiment.

A bad one...all those engineers and scientists building batteries are wrong. When will they figure it out and solder...

 

[Punx0r]

Then, mash down with a metal object, to get max contact. This will immediately start to cool joint. Then, blow on it.

I did this and it was SO cool within microseconds, that it was COOLER than spot welding.

You're just making shit up there...

 

[spinningmagnets]

I recently found a useful graphic of a common 18650 cell construction, and after pondering it, I am now convinced that the "Overclocker method" of "one second" soldering to the positive nipple is actually a very good option for the home pack-builder.

I have read in several places that the most sensitive part that is fragile as it pertains to the heat of soldering or an over-amps spot-welding is the bottom (negative) plate of the cells. this graphic also verified that. I am now convinced that the best option for the DIY pack builder is to use flat copper ribbon on the bottom negative, and put a button magnet on the outside of the ribbon so it clamps the ribbon to the broad and flat negative plate. Those button magnets are surprisingly cheap, and hold in a very strong fashion. Thanks, ES member radad...

https://endless-sphere.com/forums/viewtopic.php?f=14&t=60517#p917007

In humid climates, it would be beneficial to nickel-plate or zinc-plate the copper ribbon, however...I now embrace this as my preferred method for attaching a bus-plate to the negative.

Getting back to the small positive cathode nipple, that is where I would attach a fuse-wire, and solder is a very adequate method if done right (IMHO). Tin the tip of the fuse wire with solder, tin the tip of the positive nipple, and then press them together with a hot fat soldering iron tip for just one second...

 

[fechter]

If the can surface is really clean and you use some flux, you can get a good joint without tinning first. I prefer to tin anyway. I'm not sure if a single application of heat is less damaging than two slightly shorter ones.

 

[Overclocker]

so i sacrificed an NCR-B for science. just like the sanyo i opened up earlier it also has a plastic insulator disc at the bottom. so the active materials in the jelly roll should be insulated quite well from the heat from soldering* or spot welding


*assuming you do my 1-second soldering technique. soldering for a longer time would still spread heat to side of the metal can and into the active materials

 

[fechter]

The resistance heating method does have a nice advantage of holding the strip in place while cooling. With a soldering iron, I have to use another tool to hold the strip down after I remove the iron. Looks like a MOT with a timer might work well for that technique. Make sure that paste solder is not acid flux.

so i sacrificed an NCR-B for science. just like the sanyo i opened up earlier it also has a plastic insulator disc at the bottom. so the active materials in the jelly roll should be insulated quite well from the heat from soldering* or spot welding

Thank for sacrificing one in the name of science. Good to know how they are constructed. As long as the plastic doesn't melt, placing the joint centered on the bottom of the can looks like the best place.

 

[liveforphysics]

Neither solder or spot welds are used in automotive applications for good reasons.

Both can be possible to work, neither end up with a very appealing end result.

 

[jonescg]

The use of copper strip looks ideal - far better conductor than nickel and the process of resistance welding it to the can looks more benign than a soldering iron.

 

[agniusm]

Better non of these, better pressure contact. Snath proved that it could be reliable over three years of use in ebike environment. Why bother.
I remember couple of years ago there was not many using bms/PCM to protect their packs. Today is all changed and more people use protection.
I think this might change too

 

[fechter]

Neither solder or spot welds are used in automotive applications for good reasons.

Both can be possible to work, neither end up with a very appealing end result.

Those ultrasonic wire bonding machines are a bit spendy for DIY guys. Guess I could build my own. Do you know if they use crystals or magnetostrictive transducers?

 

[liveforphysics]

If you don't have an ultrasonic wire bonder (which is understandable as they are $$$), I would make a Snath style pack with pressure contacts for electrical. Solder conducts poorly, spot welds make stress risers that tear cans in vibe or break the welds, and require welding some thin Nickel strips which also conduct quite poorly.

 

[spinningmagnets]

Tesla needs to solve problems that not one e-bike builder cares about

Most ebikers are happy playing the odds on generic battery packs. 4 out of 5 never give the owners any problems, but...20% early failure rate is horrible. If you stick with a well-regarded battery supplier using authentic name-brand cells and a consistent manufacturing process, then the odds shoot up to 99% satisfaction and safety.

For the DIY battery builder around here, I actually care about having the information about what is better, and in what way the new concept is better. ES DIY pack builders can pick and choose which ideas they like, but...ES needs to be the place where all the ideas are posted, so we have a choice.

 

[fechter]

While solder is a poor conductor, if done right, the path length of solder for the current flow will be very short so won't add much to the total resistance. I can see why Tesla or any EV manufacturer doesn't want to solder cells, but if the soldering doesn't damage the cells in any way, it seems like a good alternative for building a one-off bike pack.

To me the real question is whether the heat from soldering actually damages the cells. I know from experience you can melt the insulator if you use too much heat. Also from experience, if you do it correctly, the cells seem to perform exactly as expected. Long term results are needed.

 

[Matador]

Neither solder or spot welds are used in automotive applications for good reasons.

Both can be possible to work, neither end up with a very appealing end result.

Can ultrasonic "welding" device be accessible to consumer for any reasonable amount of money ?
I never understood why they did not make the 18650 cells with screw on terminals like the headway cells. That would solve many headaches... Weight concern maybe ?

 

[Matador]

The use of copper strip looks ideal - far better conductor than nickel and the process of resistance welding it to the can looks more benign than a soldering iron.

I wonder and have been wondering for a while now... What is the maximum current that the 18mm diameter surface of the 18650 cells can can handle anyways ?
What about that positive terminal (which is even smaller in diameter).

 

[Matador]

If you don't have an ultrasonic wire bonder (which is understandable as they are $$$), I would make a Snath style pack with pressure contacts for electrical. Solder conducts poorly, spot welds make stress risers that tear cans in vibe or break the welds, and require welding some thin Nickel strips which also conduct quite poorly.

The question that I always wonder about.... Considering that all metal get oxidised with time (gold being the exception), some more than other.... Will pressure connection ever be able to have less resistance that physical connections like welding, ultrasonic welding, soldering. I want your input on this.

 

[Punx0r]

I never understood why they did not make the 18650 cells with screw on terminals like the headway cells. That would solve many headaches... Weight concern maybe ?

It would add significant cost and size (length) to the cells