kWeld - "Next level" DIY battery spot welder

It originally started as a one-off, but through continuous support from

https://endless-sphere.com/forums/viewtopic.php?f=14&t=89039 and http://www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here/ , I am now able to offer my solution as a kit for purchase.

https://www.kicksurfer.de/index.php/product-category/kspot-welder-kit/





Here are some key specs of my system:

Hardware:

• Input voltage for welding: 12V 30V
• Extremely rugged MOSFET-based power switch and mechanical current bus design
• Maximum switch current: 2000A
• Power switch internal resistance: 120Ohm
• Logic supplied directly from battery no need for a separate auxiliary power supply

Software:

• Central weld control algorithm uses a Joule metering approach instead of a simple timer, eliminating the need for two-pulse firing and providing more consistent welds the amount of energy that is deposited into the weld spot is always kept constant
• Ability to detect a failed weld, and acoustic feedback to the user
• Manual mode, triggered from external switch
• Automatic mode, accompanied with a warning sound and triggered with a 0.5sec delay, once the system detects that both electrodes are in steady contact with the weld material
• Audible feedback of weld process completion
• Numeric feedback from an executed weld, assisting the user to achieve best results: deposited energy amount, required pulse time for this energy, measured current flow, measured ohmic resistance of weld spot

It is not the cheapest solution on the market, but it has shown bullet proof stability. It is more than capable of welding 0.3mm nickel strips.

These videos show the system in action:

Cheers
Frank

Nice design. Impressive. Especially the sparky parts

Couldn't you make a holder that holds both electrodes? That would minimize electrode contact. The holder could be adjustable in width the electrodes are positioned.

Nice work, though. A lot better than what I could of come up with

My experience with soldering is dirty and error prone and not good for my iron lung so I may be looking for something.


Other than wanting to try one, I don't know anything about them.If you send one to me I'll give it an honest review, leveraging my extensive experience picking peaches.

It is neat that it runs from a battery though I think I'd want to run it from mains.





Yourring looks like it was zapped with 2000A but I don't judge.


Oh yea, we would get along just fine... safety f... whatever! Stuff happens.

Seems simple.I like it. Best of luck!

A battery spot welder this well engineered... This could probably sell for 500$, easily more.

I like it alot. very nice. I also recomend a holder for the electrodes like the commercial versions have. Could be 3d printed and then spring tensioned so you get even pressure.

hmmm, the quoting button doesn't seem to work on my mobile, so I'll respond this way.

At first hanks for your best wishes all!

@Korishan: if this business grows far enough then I plan to resell one of the professional probes. But there is also a drawback to fixed probes: the autimatic mode that my systrm has triggers on probe contact. it is easy to push down one probe, keeping the nickel strip in place, and then push down the other one. This way welding is very fast with my solution. Maybe 1.5secs per weld. I don't know how well the handling can be with a fixed probe armature.

@kazbach: please feel free to order one, I will give a discount to everyone making a review: 50% off the labor cost.

You can run my system from a Lipo (they don't like this treatment and lifetime will probably be short), a lead acid battery (car batteries do the job), or from AC. This requires a large capacitor though (several F, with very low ESR), which is quite costly - hard to get below 100 only for that. This will be one of ny next projects.

The ring was like that, it's not an accident

@daromer: I thought of that, but 3d printed parts typically have low heat resistance and will probably melt when this is used on high duty.

Maybe you could do a combination of manual and automatic mode. In the sense that you could have a holder for the two electrodes with a small button on it. When the button is pressed, that's when sensing goes through the contacts and the auto-mode kicks in.

I was thinking about keeping the welds evenly spaced and minimizing 'ol sparky.

I find it interesting that you can hold onto the electrodes bare handedly and not get shocked when the thing fires off. That just baffles me. I know it's voltage, but still. It can still hurt if it goes the wrong way.

Korishan said:

Maybe you could do a combination of manual and automatic mode. In the sense that you could have a holder for the two electrodes with a small button on it. When the button is pressed, that's when sensing goes through the contacts and the auto-mode kicks in.

Click to expand...

I've also seen this on other spot welding probes, that would definitely make a 3D printed part interesting. I have put making this task to the others, including making a electronics housing design. Using the system without protection would be very dangerous.

I plan to make another video, just letting the camera film while me making a complete battery pack. I am sure that the automatic mode can be used reliably as it is now, I think the audible "contact! contact!" feedback will help a lot, and an additional switch will not be necessary. I'm curious how this will work out.

Maybe also worth to mention: there are generally no sparks when the contact to the nickel strip is good, and when it is kept steadily. Whenever you see a spark in my video, I made something wrong. For example, hesitating right in the moment of auto-firing. Or welding through the battery, damaging it. Or destroying cookie cans. (Although that was fun, not wrong )

Korishan said:

I find it interesting that you can hold onto the electrodes bare handedly and not get shocked when the thing fires off. That just baffles me. I know it's voltage, but still. It can still hurt if it goes the wrong way.

Click to expand...

Why should you get shocked by any system that works with 12V? Current is impressive, but it is only harmful when directed through your body. In this case, it is 1000A through the weld spot, and a few hundred microamps through my hands.

The highest voltage is developed during the turn off transition at the end of the pulse. That is in the range of 50V, but only for a few microseconds.

The SELV law limit (safety extra low voltage) her in Europe is 60V DC, and that was chosen because this the maximum voltage that you don't feel yet when touching it directly with whatever part of your body. (Except your tongue)

Korishan said:

I was thinking about keeping the welds evenly spaced and minimizing 'ol sparky.

Click to expand...

I have no experience with monolithic probe holders, but that is exactly what my biggest concern is with them: that you do not recognize it when you don't distribute the pressure on the electrodes evenly - that's when you get sparks. That is probably much simpler when having one of them in each hand. And I had a simple idea regarding how to get even spacing, it will also be in the video.

On the hand held probe i got there is springs on both probes. so you press the tool down and then press the foot pedal.

daromer said:

On the hand held probe i got there is springs on both probes. so you press the tool down and then press the foot pedal.

Click to expand...

Can you show me a picture please? Is that one available somewhere as a replacement part?

Yeah, I was kind of thinking of the same thing, but in relation to the twin holder. Might even make it so that the electrodes dont make contact until you put sufficient force for the brains to recognize there's contact. So the spring wouldn't make contact with the probe and wire, just the probe and housing. Pressing firmly in would make the electrode make contact with the wire (possibly through another heavy duty spring to make sure stays in contact)

Just a thought. Throwing out ideas of possibilities. It's quite possible all my ideas could be bogus in practical application even though they sound good in theory

I'm still interested in the workings of it. (Though, I'm not going to be using it myself, at least not with the powerwall builds)

Korishan said:

so that the electrodes dont make contact until you put sufficient force for the brains to recognize there's contact. So the spring wouldn't make contact with the probe and wire, just the probe and housing. Pressing firmly in would make the electrode make contact with the wire (possibly through another heavy duty spring to make sure stays in contact)

Just a thought. Throwing out ideas of possibilities. It's quite possible all my ideas could be bogus in practical application even though they sound good in theory

I'm still interested in the workings of it. (Though, I'm not going to be using it myself, at least not with the powerwall builds)

Click to expand...

Much appreciated!

When thinking in this direction, you have to keep in mind that you are dealing with at least 1000A of current. That really is a lot. Despite me being an experience EE with knowledge in power electronics, my first two attempts directly went up in smoke.

For a spring loaded switch this means: either is has to have very good contact, or none at all. If it is somewhere inbetween, you get a weld at a location where you don't want it. I don't think that a reliable construction can be made. Take for example the cable connections. They need to be crimped to the cable lugs with a tool that delevops several hundres of kilograms of force to obtain a reliable connection. (In fact, this results in a cold weld when done properly.)

Thinking of this, another possibilty comes into my mind: a spring loaded electrode, but with a permanent and flexible power connection. Above a given force a sensing switch is actuated. This way you don't need your foot nor a third hand, and at the same time make sure that the pressure is high enough. Make it better and do this on both electrodes.

another possibilty comes into my mind: a spring loaded electrode, but with a permanent and flexible power connection. Above a given force a sensing switch is actuated. This way you don't need your foot nor a third hand, and at the same time make sure that the pressure is high enough

Click to expand...

I was thinkin of this too, just didn't know how to word it I think this one would probably be the best option above all.

Will it spot weld aluminium to a 18650 battery?

Aluminum is softer than Nickel, so I dont see why it wouldn't. Just turn down the power rating for it. On thing with aluminum is it oxidizes over time, whereas the Nickel doesn't. But, I suppose you could cover the Aluminum with something to keep it air sealed.

Have you spot welded a fuse wire yet ?
One electrode on the wire and one electrode on the battery.

1958 greyhound said:

Have you spot welded a fuse wire yet ?
One electrode on the wire and one electrode on the battery.

Click to expand...

Two customers of mine are using kWeld for this, and after having lowered the minimum energy setting to 0.1 joules it works very well. There is a picture showing this on the product page: https://www.keenlab.de/index.php/portfolio-item/kweld/

I have also tried to weld rotor windings to commutators of small brushed motors, this also works reasonably, but you have to be careful there not to melt the commutator's connecting tabs as well.

But it works a bit different to your suggestion (you would vaporize the wire that way): one electrode goes to the battery, and the other electrode pushes the wire against the battery. Don't slip ;-)

1958 greyhound: There has been a lot of experimentation over at the electricbike forum (endless sphere). Aluminum and copper both conduct electricity and heat very well. Because they have very low resistance, they do not get hot when passing a spot-weld current through them, Some resistance is needed to get the metal soft enough to fuse together.

Nickel is conductive enough to be useful as a bus material at lower system currents (5A per cell, etc). DIY powerwalls typically have a lot of room in which to build a system, but electric bikes have the opposite problem. They need high range and high power in a small package, so it can fit on a bicycle frame.

For a DIY powerwall, it makes sense to find lightly used 2200-mAh cells for $1, which is only $0.45 per 1000-mAh, especially when you are using hundreds of them. A popular cell for high performance ebikes is the 30Q cell. A new one is expensive. If you find them for $5, that is $1.66 per 1000-mAh. Ebikers gladly pay that because this cell can output 15A, allowing for high amps in a small package.

However, at the high currents that ebikers want, the nickel bus strips become more of a resistor wire instead of a conductor. At high currents, nickel converts much of the battery watts to heat. Doubling the layers of nickel strip is helpful for performance, but it is more expensive, twice the work to install, and still has inefficient performance.

The conductivity of metals is rated against copper (IACS). Copper is 100/100, aluminum is 61/100, but...nickel is 22/100. The shell of an 18650 cell is nickel-plated steel (although to be fair, the current-path is thin). Brass is much cheaper than nickel, it is much more corrosion resistant than raw copper or aluminum. Its conductivity is 28/100 (6 points / 27% better than nickel), but still has enough resistance to be able to spot weld it.

One downside to brass bus strips is that 0.20mm thick nickel is very stiff, but 0.20mm brass is very flexible. You would need cell-holders and a housing to take the strain off of the bus strips if they are brass.

Don't jump into using brass until the experiments are finished, brass buses may require an energy level where carbon electrodes are needed (Don't know yet, I ordered some 1/4-inch carbon gouging rods from Zoro). Also, if anyone wants to experiment with brass bus strips, make two 3mm wide strips and tape them together side-by-side (imagine a 7mm wide strip with a split all the way down the middle). If you use a solid 7mm wide brass strip, the high conductivity will allow the spot-welding current to pass through just the brass, with hardly any flowing through the cell-tip.

https://www.bluesea.com/resources/108/Electrical_Conductivity_of_Materials
http://eddy-current.com/conductivity-of-metals-sorted-by-resistivity/

Concerning spot-welding of fuse wire, fuse wire has two requirements. It must conduct electricity with low resistance at normal amp levels, but when a surge of amps hit the fuse-wire, it must melt quickly. There is a temptation to use a thin strand of copper wire, which is free from a scrap bin. This works, but...copper has a much higher melt temperature compared to fuse wire (1100C vs 300C)

Typical fuse wire for continuous currents below 15A is similar to tin/lead solder ("large current" systems sometimes use copper strand as the fuse, so the spray volume of melted material would be smaller, due to a thinner diameter fuse-wire being adequate a the rated flow). The tiny glass-bulb fuses that youtube "average joe" uses are dirt cheap, and they are they best option if you want individual cell-fuses.

---

Here is a useful design of electrode holder if you want to use one hand to hold them. The original design had the two aluminum bars solidly clamped together (with electrical insulation between them), but due to the difficulty in getting the two electrode tips to apply the same amount of pressure every time, the two sides can now flex a little in relation to each other.

https://endless-sphere.com/forums/viewtopic.php?f=14&t=89950#p1336769

Ordered, very keen to get my head around using it with fuse wire.

Dont want to solder as my partner wants to help and our Rugrat will be around as well.

I believe that fuse-wire can be safely soldered onto the positive electrode quite easily with consistent results (cleaning surfaces, proper flux,pre-tinning the cell and wire, 100W soldering iron, etc), if using the proper tools and techniques. However, the negative electrode is quite sensitive to heat damage.

For the negative end, I personally believe that compression, magnets, or spot-welding are the only acceptable methods.

http://www.electricbike.com/inside-18650-cell/