kWeld - "Next level" DIY battery spot welder

[tatus1969]

Hi,
What’s the wait time now? I’ve just ordered (#6151) and am wondering when I’ll get to play with this beauty thanks...

Already shipped by the time that I'm writing this 8)

 

[Taswegian]

I saw in another thread someone making very neat welds of copper sheet to batteries using tungsten electrodes (perhaps repurposed from TIG?)

Any thoughts on offering something similar? Obviously much less conductive than copper so presumably the idea would be to clamp them in copper and only have a minimal length of tungsten protruding.

Extremely hard and with a melting point of about 3500deg they should last forever with much less trouble of probes welding to the job.

 

[spinningmagnets]

It is true that Tungsten has a very high melting temperature. As a result, it will not get soft at the tips and begin sticking to the work when using the higher power settings. Be aware it has a higher resistance than copper electrode tips, so Tungsten will get very hot. I would recommend also investigating the spot-welding tips that are an alloy of half Tungsten with copper. They do not get quite as hot as pure Tungsten, and they sometimes cost a little less than pure Tungsten.

 

[Taswegian]

Good point on W/Cu blend. Tungsten is a bit over 3x resistance of copper, not quite as bad as nickel but not far off. 50/50 about halves it while still retaining most of the desirable properties.

 

[ossivirt]

My biggest problem with welding copper with kweld is that it seems so hard to make good weld at negative end of the cell. Positive is easy and as strong as 0.1mm copper can be I think. Negative maybe needs some extra joules but I am little afraid of it because I already made one hole and smelled the sweet sent of electrolyte

Electrodes sticking to welds have not been issue so far. I was thinking next to try weld nickel on negative end and copper on top of it so there is no danger of welding holes through :?

And I have tried nickel plating and slotting but I really don't see any difference in 0.1mm

 

[spinningmagnets]

I am working on some experiments where copper is bonded to the nickel strips in a separate operation. If your pack design can be satisfied with a 25% increase in conductivity, you might consider trying strips made from yellow brass. And of course I would experiment on old cells that would no longer be used.

 

[tatus1969]

My biggest problem with welding copper with kweld is that it seems so hard to make good weld at negative end of the cell.

That seems to be related to the fact that kWeld uses DC current, although I have on my list to do further experiments to see if it isn't actually related only to arcing.

I already made one hole and smelled the sweet sent of electrolyte

Do you have the latest firmware rev2.9 that implements anti-arcing? That should not be possible anymore with that firmware release (unless you crank up the energy setting to make that hole).

 

[ossivirt]

My biggest problem with welding copper with kweld is that it seems so hard to make good weld at negative end of the cell.

That seems to be related to the fact that kWeld uses DC current, although I have on my list to do further experiments to see if it isn't actually related only to arcing.

I already made one hole and smelled the sweet sent of electrolyte

Do you have the latest firmware rev2.9 that implements anti-arcing? That should not be possible anymore with that firmware release (unless you crank up the energy setting to make that hole).

I don't have latest firmware but so far arcing has not been an issue. I just had little too much energy involved. Maybe I should update it because if my remember correctly later firmwares have fix for the selector thing bouncing and skipping :?

 

[Taswegian]

My biggest problem with welding copper with kweld is that it seems so hard to make good weld at negative end of the cell. Positive is easy and as strong as 0.1mm copper can be I think. Negative maybe needs some extra joules but I am little afraid of it because I already made one hole and smelled the sweet sent of electrolyte

Electrodes sticking to welds have not been issue so far. I was thinking next to try weld nickel on negative end and copper on top of it so there is no danger of welding holes through :?

And I have tried nickel plating and slotting but I really don't see any difference in 0.1mm

Could you share your settings & results please? I am only starting with this welder so I'd like to learn from others experience. Probably should be collated in the OP actually. I'm starting with pretty vanilla stuff, 0.2 nickel but have designs on copper sheet later.

 

[ossivirt]

My biggest problem with welding copper with kweld is that it seems so hard to make good weld at negative end of the cell. Positive is easy and as strong as 0.1mm copper can be I think. Negative maybe needs some extra joules but I am little afraid of it because I already made one hole and smelled the sweet sent of electrolyte

Electrodes sticking to welds have not been issue so far. I was thinking next to try weld nickel on negative end and copper on top of it so there is no danger of welding holes through :?

And I have tried nickel plating and slotting but I really don't see any difference in 0.1mm

Could you share your settings & results please? I am only starting with this welder so I'd like to learn from others experience. Probably should be collated in the OP actually. I'm starting with pretty vanilla stuff, 0.2 nickel but have designs on copper sheet later.

For reference my battery gives 1500A. That probably is relevant when going for copper..

For 0.1mm copper 45-50 joules is OK for my setup and positive end of the cell. I'm still working how to make it reliably at the negative end.

For nickel maybe around 20J. In my experience it welds nice at almost any setting.

 

[DogDipstick]

I got mine, works wonderfully. Best spot ever, and I have rewound transformers on 110v spot welders before. Nicest and consistent welds I think I may have ever made. I am using a NMC Lipo I designed with Chevy Volt cells. Beginning @ 24.9v, it seem that the initial 100 test welds has not discharged the battery much. After 100 welds, still 24.7-8v. I am wondering how many welds I can do with this pack. I am guessing a few thousand in a row.

The only thing getting warm is the silicone 8Ga. Does this welding machine have a duty cycle not to be exceeded? I dont wnat to abuse it whatsoever, and I will be patient if need be.

So is the calibration basically a measure of the short circuit current? I got a metered 1750 A using my lipo. Does that mean that my Lipo has a Maximum burst rating of 100C or ( 113.333C is where i calculated it out to be... )?

If you use an undersized lipo, are you not in danger of melting the cell level connections? Or will it just "undercurrent warning"? Most Turnigy, Graphene Pulse Gens and other Hobby batteries are not very robust: I have de soldered connectors in flight ( 5mm+) with continued pulls with a 16s 300A max rated motor, pulling about 30C ( actual) on a 5Ah 12s pack. These batteries are rated for 65C and more... such as the Graphite ones. They still failed. The tabs are small, soldered, and weak metals. They burn easily with compromised connection. Even the smallest resistances can build into arcs at this level... burning things.

I think this would mean that if I continue to weld, using this battery, I will discharge the whole thing in under a min of weld time with 50ms pulse width? I am trying to see if it is feasible to use my battery. I do not want to go below 3.0V.

Time, 50ms avg. pulse, Set @ 35J, .3 mm Nickel, average current, 1675A, battery size, 15.25Ah, battery voltage, 6s, 24.9v charged....
Thankyou for making this possible.

Spinning Magnets: you mentioned Tungsten electrodes, I am in a position to use them, I have all vairity of Thorated, and non, here, from tiny to large. I once had a 3ph 400v battlship welder, water cooled tig, but scrapped it eventually. 500lbs of copper is 500lbs.. lol.. but nayway...

Tungsten will oxidize red hot, and above. The metal will become brittle and chips of tungsten will break off in the weld. Tungsten is great for a shielded arc, but add oxygen and the metal will falter and burn when colored red or greater. This happens rather quickly when you are welding and run out of gas from the regulator, the tip will burn. I do use a piece of tungsten for my resistance soldering iron, though, when it is not red it can transfer appropriate heat for a good time. Not much current though because of the resistance value, it turns red quickly. I will be experimenting with all metals.

 

[Taswegian]

The only thing getting warm is the silicone 8Ga. Does this welding machine have a duty cycle not to be exceeded? I dont wnat to abuse it whatsoever, and I will be patient if need be.

After the weld pulse keep the the trigger switch down. The display will cycle through some info, one of these is temperature of the MOSFETS. The welder will self protect and show an Over Temp error when that gets to 70. I get the impression that is set at a level to comfortably avoid damage and if tripped you just need to let it cool before resetting the power and resuming however it seems wise to keep a decent bit below that. Either weld at a pace that allows that, or consider putting a fan on it.

I'd strongly advise putting it in an enclosure, however crude. With these energy levels just a bit of stray copper etc finding it's way onto the board could result in a spectacular pop.

Making sure all the connections are clean and tight so you're not generating unnecessary heat somewhere at the board seems advisable too.

 

[krlenjuska]

When i add kit to my cart i have mwst (i guess its VAT) 19%.
When i buy something online i always pay VAT in my country.
Why this is different?

EDIT: My mistake - i added adress in Slovenia (my friends adress) so thats why VAT is added to total price.

 

[tatus1969]

When i add kit to my cart i have mwst (i guess its VAT) 19%.
When i buy something online i always pay VAT in my country.
Why this is different?

EDIT: My mistake - i added adress in Slovenia (my friends adress) so thats why VAT is added to total price.

When selling to private buyers in Europe, then I have to deduct VAT (called MwSt in Germany) and forward amount that to our goverment.

 

[tatus1969]

I got mine, works wonderfully. Best spot ever, and I have rewound transformers on 110v spot welders before. Nicest and consistent welds I think I may have ever made.

Thanks 8)

Does this welding machine have a duty cycle not to be exceeded?

It will protect itself from overheating. But to my experience, the electrodes are the electrodes are the weakest spot. As they are in direct contact with the melting material, they inevitably receive some of that heat. That's why professional machines use water cooling, but I considered that being inappropriate for a hobby level tool.

So is the calibration basically a measure of the short circuit current?

No, it measures the resistance of the path between module and electrode tips, which is required for pulse energy metering at the weld spot. That's what is unique to kWeld and gives it its repeatability (and also the new anti arcing feature 8)).

I got a metered 1750 A using my lipo. Does that mean that my Lipo has a Maximum burst rating of 100C or ( 113.333C is where i calculated it out to be... )?

No, this is basically the short-circuit current of your battery. The burst rating is somewhat related with that, but cannot be derived by this.

If you use an undersized lipo, are you not in danger of melting the cell level connections?

Yes you are, which is why I had tested the models that I currently recommend. But as you say, HK isn't known for its consistency, this still isn't a guarantee that a bad production lot causes problems like melting cell interconnects.

I think this would mean that if I continue to weld, using this battery, I will discharge the whole thing in under a min of weld time with 50ms pulse width? I am trying to see if it is feasible to use my battery. I do not want to go below 3.0V.

Exactly :wink:

I'd strongly advise putting it in an enclosure, however crude. With these energy levels just a bit of stray copper etc finding it's way onto the board could result in a spectacular pop.

Making sure all the connections are clean and tight so you're not generating unnecessary heat somewhere at the board seems advisable too.

I'd like to add a few important things that haven't got their way into the manual yet:
- make sure not to apply external voltage to the electrodes - do not let them touch different voltage potentials of your battery under construction.
- use the lowest possible energy setting to avoid damaging cell internals from welding - make tearing tests with increasing levels.
- when calibrating, make sure that all bolts are tight (also set screws of electrode holders). The SHORT step should not create any sparks, otherwise clean tips with sandpaper.
- ensure full contact of bus bars, avoid any dirt or metal chips. Bad contact will cause false current readings.
- even when using an enclosure for the welder, never touch the LCD frame with the electrodes as that is connected to the negative battery terminal.

 

[Taswegian]

I'd like to add a few important things that haven't got their way into the manual yet:
- make sure not to apply external voltage to the electrodes - do not let them touch different voltage potentials of your battery under construction.
...
- even when using an enclosure for the welder, never touch the LCD frame with the electrodes as that is connected to the negative battery terminal.

I was thinking of doing a modified version of the enclosure to include mounting for firmware update board with access to the usb port at the side.

These comments have me thinking it would also be nice to add some holders to park the probes on the sides (kinda like a desk pen set). A ridge that sits a little proud of the display to reduce chance of accidental contact with the frame might be good too.

 

[DogDipstick]

The only thing getting warm is the silicone 8Ga.

Either weld at a pace that allows that, or consider putting a fan on it.

I'd strongly advise putting it in an enclosure, however crude. With these energy levels just a bit of stray copper etc finding it's way onto the board could result in a spectacular pop.

Making sure all the connections are clean and tight so you're not generating unnecessary heat somewhere at the board seems advisable too.

Yes the machine is in its tidy enclosure, nicely fitted and very neatly cut. I was just excited to get to weld practice.

- when calibrating, make sure that all bolts are tight (also set screws of electrode holders). The SHORT step should not create any sparks, otherwise clean tips with sandpaper.

I cleaned the electrodes, do I need another calibration for the accurate response from the welder? My initial calibration did create a sparking action ( or arcing? ) at the electrodes. I cleaned the tips, and welding current increased to 1850-1920A. Small welds, all under 50ms, many much less when welding with lower energy, such as 10J-15j. Most welds are completed in under 20ms. Some thinner tin I did spot with >10j and >10ms.

I played (rather UN-scientifically) with pure nickel, brass, bronze, berrillium, copper, and the like I had laying around. I have experimented with the DeWalt style connectors too. I beleive they are thick nickel plated copper. I have read about certain alloys, I will be trying to repeat and reproduce results with consistency. I have read about the fluxes vs. the clean metal, ect. Some have recommended using a brazing flux between the metals to be joined, for conductivity at the joind with the melting of the metals. I have been traditionally taught to only weld CLEAN metal my whole life, so using fluxes seems to be faux pas for me.

Whenever I try to weld anything purely conductive, I only get an over current. Consistently, time after time. I have read this thread extensively, and it seems the answer to my problem is to lower the input voltage. I will shortly build another (4s?) NMC Lipo for this application.

Tatas, if you had your perfect battery configuration for this welding machine, @ what voltage would you ( the weldor) choose? I am in position to build the battery of choice for my application.

If I was to switch electrodes to a Tungsten electrode, do I need to re-calibrate the machine? I have in my hand some 1/8 in Tungsten, pure and annealed, ready to experiment. The supplied electrode is .2 in. My Tungsten is .125in. I have ( copper) collets in the size needed to adapt the current electrode holder to my Tungsten... is there any danger in using collets in this configuration?
Also, I wonder about tip shape: Many weldors recommend "channeling " the current with a specific grind, or ball, on the tip oi the Tungsten. Would this be advantageous to welding highly conductive metal?

Thankyou for all your efforts, and knowledge. I will report more results as soon as I overcome my voltage and current situation. Consistency to this point has been flawless.

 

[tatus1969]

I cleaned the electrodes, do I need another calibration for the accurate response from the welder?

I'd suggest doing that. You can repeat this as often as you like, and watch if the measured resistance isn't jumping too much (should be accurate to a few hundred microOhms). Just make sure that you don't get sparks, press hard.

My initial calibration did create a sparking action ( or arcing? ) at the electrodes. I cleaned the tips, and welding current increased to 1850-1920A. [...] Whenever I try to weld anything purely conductive, I only get an over current. Consistently, time after time. I have read this thread extensively, and it seems the answer to my problem is to lower the input voltage.

That current is too close to the overcurrent limit of 2000A. Going that close to the limit of course reduces pulse times, which is very good, but it also causes more stress to the components than necessary. The "sweet spot" is a current of 1100 to 1500A during calibration.

if you had your perfect battery configuration for this welding machine, @ what voltage would you ( the weldor) choose? I am in position to build the battery of choice for my application.

It's always a compromise. A low voltage increases the variation of welding current with different situations, and a high voltage causes too much wasted energy. The voltage across the spot is only a few volts, the cables and module eat up another few volts, and all the rest needs to be "eaten up" by the battery's internal resistance. The higher its voltage, the higher that resistance needs to be to get the same current. I have only experimented with 3S batteries, but a 2S battery with more massive paralleling hence lower ESR will further lower losses but still keep variation in current low enough.

If I was to switch electrodes to a Tungsten electrode, do I need to re-calibrate the machine?

Yes. The purpose of calibrating is to measure the output resistance that the control module sees, so that it can cancel it out when welding. That way it can then do accurate weld spot energy metering.

 

[Kajman]

I have your welder still in a box. But battery is a question. You say it shouldn't be too high voltage, I have many batteries, so what voltage is best? Current is not a problem,have very strong lipos, lifepo and liion, so what is yourideal voltage and supply current?

 

[flippy]

how does the welder deal with constant use?

with constant use i mean several hours of making about 2 dozen welds per minute with only breaks between mounting nickel strips.

i am insterested in how it deals with being heat soaked and the diodes getting flogged constantly. what is the first or weakest component in this chain?

 

[Merlin]

weakest is your hands with gloves who cant touch the electrodes if you hammering as fast as you can.
after that the battery + wire will heat up.

the welder itself you can touch without problems. i wonder why its that "cold" while rest is hot as hell.

recently i build with 0.3 nickel and ~100J
5-6 Spots and cool down is needed for electrodes.
50 spots and my battery reached my (own) temp limit of 50°c (6000mAh 65c 3s)

 

[flippy]

weakest is your hands with gloves who cant touch the electrodes if you hammering as fast as you can.
after that the battery + wire will heat up.
the welder itself you can touch without problems. i wonder why its that "cold" while rest is hot as hell.
recently i build with 0.3 nickel and ~100J
5-6 Spots and cool down is needed for electrodes.
50 spots and my battery reached my (own) temp limit of 50°c (6000mAh 65c 3s)

the reason the welder is cold is because the reistance is much lower. resistance creates heat.

i plan on modifiying the electrodes with water cooling if they last decently long. right now i am using pure copper nails and they are annoying because they blunt down real fast and need constant sharpening and wear out pretty fast.
the low resistance is great as they dont heat up very fast but when they blunt down and the tip oxidizes you need more pulse time and that increases heat.
my current battery is a 60Ah lifepo4 from thundersky, its capable of dealing 1500+A easy and i keep it floating at 14V with a 5A charger. it does not heat up in any meaningful way.
but if this thing can weld .3 comfortably then i will probably want to get the capacitor upgrade.

the packs i make are VERY big for this forums standards, some are dozens of Kwh's, so i am looking for a more durable option instead of keeping 2~3 cheap welders on the shelf as backups if one blows up. i'd rather have one i can rely on...

 

[tatus1969]

the reason the welder is cold is because the reistance is much lower. resistance creates heat.

Exactly, that's also the main reason why kWeld is more expensive as other seemingly comparable systems. The MOSFETs that I have chosen are among the strongest that are available on the market (in this physical size). I have made long term stress tests with several 10,000 pulses with high repetition rates and a water cooled test load, and never saw power switch temperatures above ~50°C. (The fuse gets quite hot during heavy use though.) The current system also has thermal protection and will refuse pulsing at overtemp. And after having sold more than 1000 units, I only have a single warranty case that is related to lifetime stress. Most failures happen shortly after purchase, and in many cases they are clear user faults. Still less than 1% in total, which is a number that I had never expected considering this self-assembly DIY tool operating with ridiculous levels of electrical current.

the packs i make are VERY big for this forums standards, some are dozens of Kwh's, so i am looking for a more durable option instead of keeping 2~3 cheap welders on the shelf as backups if one blows up. i'd rather have one i can rely on...

I'd suggest running two kCap in parallel then, both of them equipped with fans. The kCap power supply that I have been working is also on its way, the first production batch is ordered and I'm about to put it in the webshop for preorder shortly.

You may also want to beef up all wires to 6AWG or even thicker, this reduces heat and gives you more welding current.

 

[flippy]

Exactly, that's also the main reason why kWeld is more expensive as other seemingly comparable systems. The MOSFETs that I have chosen are among the strongest that are available on the market (in this physical size). I have made long term stress tests with several 10,000 pulses with high repetition rates and a water cooled test load, and never saw power switch temperatures above ~50°C. (The fuse gets quite hot during heavy use though.) The current system also has thermal protection and will refuse pulsing at overtemp. And after having sold more than 1000 units, I only have a single warranty case that is related to lifetime stress. Most failures happen shortly after purchase, and in many cases they are clear user faults. Still less than 1% in total, which is a number that I had never expected considering this self-assembly DIY tool operating with ridiculous levels of electrical current.

I'd suggest running two kCap in parallel then, both of them equipped with fans. The kCap power supply that I have been working is also on its way, the first production batch is ordered and I'm about to put it in the webshop for preorder shortly.
You may also want to beef up all wires to 6AWG or even thicker, this reduces heat and gives you more welding current.

thank you for your answers.

cost is more secondary to me, having a bunch of spare cheap welders on the shelf is just as expensive or even more expensive then 1 good one.
i already am using 4 and 6AWG in my current welding setup and the cables are cold to the touch.
is the welder capable if dealing with that amount of power 2 sets of those pretty badass caps can deliver?

i see the voltage limit is capable of 30V, is using a lifepo4 bank also a option? i have a lot of 40Ah lifepo4 cells on the shelf. and a 30V lifepo4 bank could deliver much more then even those caps can push out at 8V. that bank would also only require a light charger to keep it topped off. as my setup is completly stationary i can have a much bigger bank behind the welder then most people would do in a DIY solution.
or would the caps give some kind of other advantage?

 

[spinningmagnets]

One of the benefits of using supercapacitors in the kCap is that the battery (or power supply) does not necessarily need to be very high amps and very low resistance, like the nanotech LiPo that was the initial power battery for this design.

Many builders wanted to use a common 12V car battery, but if they did that, it needed to be fairly new so it had low resistance. This is also why the V4 has a power test function, so the builder can tell ahead of time if the battery they plan to use is adequate.

If using only one kCap, I believe the max voltage is 8.1V, but I would have to check that.