LiIon as source for a spot welder? (1000A for a few miliseconds)

So I’m considering building a spot welder (for battery tabs mostly)

There are plenty of DIY versions, typically a power source (rewired microwave transformer or ultracapacior or car battery).

since ultracap or car battery would be a one time investment for which I have no other purpose, I was thinking to use for example VTC5A as the power source.

What do you think about that?

The power requirement would be that they are able to provide some 100-200A for up to 20ms each cell, to total up to cca 1000A for up to roughly 20ms

I’m not sure if they would need to be all connected to parallel (for example 10p) or up to 3 in series and then in parallel

FYI the solution roughly consists of

- attiny to control the pulse, pulse goes to gate driver, gate driver fires controls MOSFET

problems I see:

- ringing when disconnecting mosfet

  • MOSFET dying and short circuiting the power source (need some safety option)

That’s subjecting the cells to a direct short, which is just about the most unsafe thing you can do to a LiIon. Buy or build an actual spot welder.

Stick with the ultracap. It’s well-suited for high discharge current.

Thanks.

@both of you. Is dead short for 20ms still unsafe, provided that for safety I fuse cells individually? I’m not able to find anything online on this matter.

My reasoning is “if VTC5A is rated for some 30A continuous / 85*C threshold, and some 60A burst. what’s the problem of driving it at 100-200A for 20 miliseconds?

I accidentaly shorted a VTC5A for a second or so, and it literally evaporated some 4 square millimetres of the nickel tab. So I think it does have enough high discharge.

For spot welding often a car starter battery is used, some 45Ah capacity.

so, still a no-no or you would give it a chance?

Getting away with it once doesn’t qualify as a basis for designing a repetitive use item. You probably won’t get a reccomedation to do this but no one can stop you if you want to try it anyway.

Yes they can deliver very high current when shorted. No it’s not safe to let them do it. The End.

Too bad mythbusters is not still around. This would be a good one.

fully charged their dead short current is around 200A, so that’s some 700-800W heat dissipated, If one could keep them shorted long enough they would probably melt their internal thermal fuse before that, if not then we would have thermal runaway. This would be an interesting experiment.

It’s a bad idea, but as I see it, it would work fine as long as you remember the internal temperature of the battery is rising fast. Even faster than you know by feeling them. If you have multiple spot welds to do you could push them too far. Oh- and incase you need to be told, use long wires so you are nowhere near the cells if you try this.

But ‘working’ is not an excuse. Just buy a spot welder. Or build it right.

Mythbusters Need Not Apply

The inventors of the technology tried that. They published their results for all to read and for peers to review.

According to Sony’s Lithium Ion Technical Manual, a fully-charged 18650 in a dead-short rig merely discharges, gets pretty hot, and then stops at ~0v.

3-3-2 Abuse Tests
table(table#posts).
|Test Item|Battery
State|Temperature|Test Method|Test Results|
|Short-circuit
test|Fully
charged|20 ±5°C|The battery terminals
are terminated until
voltage is 0.1 V
or less with a resistance of
50 mΩ or less.|No ruputure or
ignition; 150°C
or less of battery
temperature 55±2°C|

Belief is always based on someone else’s opinion. I prefer to believe the people with “skin in the game”, especially the inventors.

So far, I can find no one else who’s been able to accurately, honestly replicate Sony’s testing, so people unwilling to find out on their own will have to believe the scary BS on Youtube.

Here’s an idea: You Be The One. Set it up, honestly without an agenda to push, and video the results. Our finest and most-highly-trusted Reviewers here basically all started that exact way.

Just sayin…

Having said that, no way in H311 would I do what you’re trying!

Any decent car battery will cost less, deliver more, and provide more collateral utility than any pack of 18650s I could imagine.

Try it, some time ago people didn’t travel too far in ships because everyone else told them they would fall off the edge of the world :wink:

What’s the worst that could happen……

One guy built spot welder with A123 26650 cells,which are safer,have much longer cycle life,and they are rated for 30C (70Amps),but they were dead after very short amount of time.

I’m thinking how to provide good connection that will not break circuit while the battery is melting and catching fire :slight_smile: I’m mostly worried about Swedish regulations (yes there is a rule for everything).

I found some videos where a trustfire went thermal runaway after a brief short circuit which apparenty caused internal damage. So it is not impossible that a brief short circuit leads to undesired chain of events.

@led4power - can you provide a link? It’s difficult to google out because I’m only getting welders that weld lion batteries, can’t filter for welders which are powerd by liion batteries.

Yep, if you damage it internally you will have a time bomb ticking. Exactly why we are saying this is a bad idea, and not to be anywhere near the cells if you entertain bad ideas.

I’m an engineer. I find bad ideas irresistibly intriguing :slight_smile:

I found VTC5 specification

it states 100A is allowed for 3 seconds.

Ha. Ok. Don’t coin any new phrases like ‘curiosity maimed the engineer’.

hehe I’ve been electrocuted more than once :slight_smile: Still can’t get enough of it.

I find this idea very intriguing - the problem is that I would use this welder a lot, and then I feel I would have a 10pack of VTC5A cells sitting in my room, being short circuited for hundreds if not thousands of times over a year or two. And only one internal short circuit could be potentially catastrophic, like having a ticking bomb.

Still I’m intrigued to somehow test this premise. I could build it and then torture test it, set it in an infinite loop with pulsed short circuits to test if something bad would happen. I live in a flat so I don’t have a safe way of doing it.

Just buy some D-cell-size Maxwell supercaps. They’re damn impressive. And they won’t explode and kill you.

You don’t need to get all fancy with the control side, either. Just a momentary foot switch that applies power to a FET’s gate and there’s your spot welds.

yeah I will feel better with caps and not shorting VTC’s. When I was a kid I had tons of fun with some 0.2F caps, Now I can afford much more capacity :slight_smile:

I already ordered FET’s, I plan to control them with attiny85, but I think I need a gate driver to avoid ringing/spikes. I don’t yet know how to make gate driver, or which integrated gate driver to use.

This guy did all the math:

http://www.turtlesarehere.com/html/cd_welder.html

This is not a test result for an actual battery, but a test specification with criteria for “test passed”.

Note that “does not vent” or “does not spew electrolyte” or “cell performance still meets specifications after test” are NOT listed as pass criteria.
That is, a cell passes this test even if it vents and spews out electrolyte and is dead afterwards. It must just not ignite and not explode or pop. Controlled venting and terminal damage is allowed.

A high-current cell without PTC (like the VTC5A) will usually vent on an extended low-ohmic short (if the shorting wire can take the current and doesn’t melt after a second). At the very least, the cell is damaged afterwards.

Ohm’s law: 200A current requires max. 20mOhms total resistance for a fully charged cell. That is, internal resistance plus contact and wiring resistance plus welding point resistance. So it could be possible to get the required current out of a VTC5A for a fraction of a second, but maybe not very wise to do this repeatedly. And if the welder fails to end the current pulse (FET shorted or the like), you can only hope that some other component will melt faster than it takes your cell array to start venting.

Edit:

Actually, this is a requirement specification for an “over current protective circuit” (7.2.4.3), not a performance specification of the cell.