kWeld - "Next level" DIY battery spot welder

[veer]

Hello guys!
I bought a kweld last year, but didn't get a chance to use it properly. I was using a cheapo lipo battery with insufficient power output. I was getting 750a max output from it and the weld were decent, but nothing good or promising. So I bought a Kcap and Ksupply to power the kweld. I received it day before yesterday and yesterday I was assembling it. I encountered a major problem.

There are 2 adjuster knob type screws on the back of the Ksupply. 1 for voltage, 1 for current. Voltage is to be set at 8.1v and current to its maximum point. I did do the voltage accurately to 8.10v, but when I tried to move the current adjuster, it just kept rotating and rotating. It's a bad adjuster I think. So I set it to the middle-ish position and connected my kweld and Kcap to it. But now, the Kcap power LEDs, only the left of the 2 turn on. The manual indicates that insufficient voltage or current. Voltage is fine, it's the current for sure! Kweld at the same time shows 'insufficient battery" and shows voltage, 8.00v to 8.09v when plugged in, and falls slowly when disconnected.

I have contacted Frank also, I'll update with what he suggests. If anyone has had a similar problem and can help me, that would be greatly appreciated as well!
Cheers

 

[Eraser-1]

Hello

The newer kCap have different caps and need a higher input voltage. Increase the voltage and the kcap will function.

I have written something about the new kcap a few posts before. I think on my kcap I have increased the voltage to around 9 to 10V.

 

[veer]

Thanks for your reply!
I increased the voltage to 8.7v and the kcaps right led came on, meaning "module is fully charged". I haven't kept the system going for 10min total though!

The kweld has a problem now, I clicked the CAL for calibration, and since then the screen has been stuck on "offs= 63, Std= 3.4, Ok" this is going on in a loop! And with no disconnect plug, I can't even restart it.

Should I remove the wires and rewire it?

Edit: I did that, first time i'm going to do is put the low battery voltage to 7.5v. Is something else recommended? I tried to search but couldnt find relevant to this.

Also, I'm considering putting a connector between the kcap and the kweld, something like xt150. It can do sustained 100Amps so might be able to do the pulses for kweld. Is this a good idea? Or am I just going to increase the resistance of the system and put it off its calculations?

 

[veer]

Update: Kweld is still able to weld. BUT, it is stuck in auto mode. Whenever I try to click on the MODE in the menu, it shows "Manual" and then goes onto show some values that used to be shown after a weld was done. like "Spot R, =000004, R=7825.4, t= 0.40m, I= 0A, T= 36.3C," in cycle. And there is no way to go back to the menu. Only shutting down is the option. That is VERY shady to do as right now I am removing the input wire from over the fuse!
Welds at just 20J on 0.01mm nickel to nickel were VERY strong though. Has my kweld gone bad? Or it is something else?

Update: Solved all the issues discussed and encountered new ones! :lol: . The connections for the footpedal had corroded and were shorting, i think. That was why it was happening. Thought about deleting the post, but might help anyone in the future going through something similar.

New Issue: Stuck in the menu for some reason, I think the knob has had it. clicking on the EXIT menu option leads me back to CAL option. Noticed some weird movement from the knob before also. While rotating the knob to decrease the energy, made it increase the energy to double the original!! But started working normally after that. Till I got stuck in the menu. I have restarted the module by disconnecting the Kcap power cables, but that is still leading me to the menu itself. I am seriously wondering what to do at this point!

 

[veer]

Update: The ksupply current adjust trimmer seems to be broken. And it either giving out 0.198A @ 8.7V or 0a if I turn it too much. There is no option or way right now to get out of the menu settings on the kweld. Talking with Frank and figuring out a solution, he's really helpful and answers all the questions, however stupid and naive.

 

[fixvid]

Hi,

I bought a kWeld last year with 4 of those big super caps which Frank was selling separately. I got around 1150A which for steel is enough but i have problems with 0.2 nickel and most of the time only one spot of the two is sticking. I make 3 pairs of spots weld but i am using it for commercial application and i can't afford to have half of the spots not cooked properly.





I can rip one of the spots easily and i'm relying on the other one that "hopefully" it will stick(50J pulse). I often stack 4 layers of 0.2 nickel(one after the another) and i have to be absolutely sure that is sticking. I changed the battery cables to 6 awg which gave me a total of 1300A, i tried cleaning and maintaing a perfect shape of the tips , putting enough pressure but nothing seems to help.

I am thinking that 0.2 nickel just needs more current to cook properly, i would go to the max 2000A or even higher by shortening the probe cables and rotating the spot welder 90degree for shorter reach. It would be also nice to change the probe cables to at least 6awg, but i dont know how to change them and still use the original probe holders.

If i knew from the beggining i would have bought 8 supercaps for 4s2p configuration. I dont know if Frank still has some of those caps. Another power supply ideea would be a 4s battery built with A123 lifepo4. Those can give like 120A for 10sec, i suppose 8 of those in parralel can give 2000A in a short.

I would be glad to hear some of the forum members about their experience with kWeld and what power supply they use for properly cooked welds and also opinion.

 

[leffex]

Just got my welder working after having a former battery being used tested at 750-850A that didn't give any welds. I'm now using a battery that gives out 1860A and maybe I can add some cells to get the last Amps as well if I want.

Tried some welding and checking nickel holding onto the cells and 50J was best as well as 55J and 60J using 0,2mm pure nickel. The difference I think is that the weld quality also depends on how hard you press the copper rods. So I'm going to make a holder for the rods so I can press harder and also have the same length between the rods and so to say get more consistent welds.

thanks inventor!

 

[TrotterBob]

Hi,

I bought a kWeld last year with 4 of those big super caps which Frank was selling separately. I got around 1150A which for steel is enough but i have problems with 0.2 nickel and most of the time only one spot of the two is sticking. I make 3 pairs of spots weld but i am using it for commercial application and i can't afford to have half of the spots not cooked properly.

I can rip one of the spots easily and i'm relying on the other one that "hopefully" it will stick(50J pulse). I often stack 4 layers of 0.2 nickel(one after the another) and i have to be absolutely sure that is sticking. I changed the battery cables to 6 awg which gave me a total of 1300A, i tried cleaning and maintaing a perfect shape of the tips , putting enough pressure but nothing seems to help.

I am thinking that 0.2 nickel just needs more current to cook properly, i would go to the max 2000A or even higher by shortening the probe cables and rotating the spot welder 90degree for shorter reach. It would be also nice to change the probe cables to at least 6awg, but i dont know how to change them and still use the original probe holders.

If i knew from the beggining i would have bought 8 supercaps for 4s2p configuration. I dont know if Frank still has some of those caps. Another power supply ideea would be a 4s battery built with A123 lifepo4. Those can give like 120A for 10sec, i suppose 8 of those in parralel can give 2000A in a short.

I would be glad to hear some of the forum members about their experience with kWeld and what power supply they use for properly cooked welds and also opinion.

I've seen 3,000F Nesscaps on feebay for 1/2 what they go for new. Not sure if they're still around but there are deals to be found on slightly used caps. What size wire gauge are you using on those caps? Looks a bit small to me.

 

[I'M TechMech]

And how could I possibly connect 2 Kcap modules in parallel ?

Just to be sure, Is it by wiring the outputs of the Kcaps together (+ve to +ve and -ve to -ve) at the Kweld input screw terminals AND

Leaving the input slots of one of the Kcaps not connected to anything AND

Connecting the input slots of the other Kcap to the output slots of the Ksupply and PSU

?

Thanks!

 

[MagicToolbox]

I've been reading most of this thread - thanks for all the information. I received my Kweld recently and have assembled it, I'm printing the last part for my custom case now, and Saturday I received 8 used Headway 32120 cells for a supply. I can set these up as 2S4P, 3S2P, or 4S2P. Is there a good reason to use one over another? My plan is to set up the Headways as 2S4P and run the calibration to see what results I get, but I'm willing to learn from others!

I've got a box of K2 26650 3800mah Lifepo4 Cells from Battery Hookup that I want to make packs with, but I have a bunch of salvage 18650 cells that I will practice with first.

 

[TrotterBob]

From what I've seen on youtube those K2 26650 3800mah Lifepo4 Cells need more joules to weld vs your standard issue 18650s. Havent tried it myself. Just a heads up.

 

[nuxland]

Finally could weld 0,2mm nikkel to 0,5mm copper with my kweld Welded so that one electrod was in copper side and other in nikkel side.
Now can start to design my new battery that has 0,5mm copper and 0,2mm nikkel welded to it so I could then weld the nikkel to the cells. 14p28s Samsung 40T cells in 7p rows, it will be then 35mm2 between series that is enough for me

 

[NetPro]

Received my kSupply and kCap a few days ago and I am amazed how well they work: Super fast charging (using a power supply for a Dell server with very close specs to the HP HSTNS-PR01) and nearly silent.
These are superb products Thank you, Frank!

I do have a question regarding the voltage listed in the manual of 8.1v.
Turns out the SkelCaps capacitors used, pack quite a punch and if I set the voltage to 8.1, cannot calibrate the welder because it reports "overcurrent", probably because I mounted the welder right next to the kCap and there is minimal current loss there.

I found two ways to fix this: by replacing the AWG 4 cables from the welder to the welding head with thinner ones, OR by lowering the voltage on the kSupply to 7.1 volts.
At this voltage the calibration reports just a few amps short of 2000. Consistently

Is it OK to run my setup at 7.1 volts?
No warning LEDs come on and I am getting wonderful welds even using 0.2mm pure nickel strips, which I was not able to do well before I got the kCap.

Not sure if there is any negative long term effect by running it 1 volt lower.

 

[Eraser-1]

On my kCap with the SkelCaps I need tu use more than 8,1V, otherwise the indication LED's on the PCB of the kCap would indicate that it is not fully charged. I increased the input voltage to around 9-10V and that is fine.
As written a few posts before, it looks like that the new SkelCaps can handle a higher voltage and so the cut-off-voltage of the input was now also increased. Unfortuantely this information is not listed anywhere and on the homepage of Keenlab there are still the old caps listed and also in the manual.

The measured current on the kWeld I get from the kCap is around 1650A with the original cables used for the connection between kCap and kWeld.
Original cables are AWG8 I think.

So I always thought that the caps are not able to generate a higher current because of the internal resistance of them, but it looks like (like you said) that the caps can generate more than 2000 amps with a thicker connection cable and so that if I also increase the wire diameter my current will also rise. Very interesting...

Do you have some pictures of your setup?

 

[NetPro]

I didn't bother to take pictures prior to assembling (I doubt anyone would want to put as many hours on this project as I did) but took some today to give you an idea.
It is hard to get the idea by looking at the pictures once assembled because what I did was to connect the welder and kCap with solid copper bars and spacers, in a three-dimensional fashion: you can see a little of the bars between the two devices and they look like a short, but they are far from it.

I made 2 spacers about half an inch and connected the positive side, leaving enough space between the bottom of the PCB board and the bar.
Then, made two spacers --taller-- so I could connect the two negative connectors with plenty of clearance between it and the positive one.

Call me crazy but I retired recently and have a lot of time to play with the toys and that includes machining parts in my lathe and milling machine.
Since I enjoy working with copper, brass and aluminum, this was a good opportunity to use left over copper I had bought a long time ago to build a welder designed by a Fritz G., when he was selling the bare PCB on the internet.
Never got it to work well for thicker nickel tabs but it did fine with the thinner ones.

In any case, I wanted to have the three Keenlab components and the power supply in one solid unit and solid it is: That phenolic resin I used for the base is rock-solid and a dream to work with with the right cutting tools.
I threw that acrylic cover around it to prevent any accidents with metal objects touching the wrong spot. Not my finest creation but it works.

 

[spinningmagnets]

Very nice welding station!

Take a quick read of the copper/nickel sandwich method. It might be a good option for you.

 

[Ianhill]

Looks nice, how repeatable is the trigger mechanism for setting up the pulse at the same point of interaction between cell and tip each weld ?

Id been tempting to set the switch up with some sort of optical distance meausrement or preasure based trigger to set the pulse up.

The optical method is painful theres non ideal surfaces of reflection to get a reliable data point continuously so the best method in my mind is an arduino measuing a load cell then set the pulse up with engagement.

Benefit it will trigger for an 18650 and a 21700 cell or even spot some tabs together.

 

[NetPro]

It is very consistent. I once tested it with a scale under the electrodes and it was firing always within a kilo (or two at most).
It is important to note that the smoothness of the two rods sliding up when you weld is critical for this accuracy.

Before I installed the linear radial sliding bearing (two on each side) it was not accurate and the welds were not consistent.
Those bearings make the rods slide like silk, even when you increase the pressure to avoid sparks or to weld thicker material.

I have a pneumatic welding head but I like this one more since I don't have to deal with air compressors and the noise that comes with them and the welding nugget I am getting after installing the kCaps would pass a microscopic audit, I am sure

I like the Arduino idea: Sounds accurate! If you come across a sketch and some how-to instructions, please post them here or PM them to me as I would love to give it a world.

 

[Hummina Shadeeba]

Doesn’t the pole get in the way?

 

[NetPro]

It is an optical illusion due to focal length magic.
This is a side view picture of it and for my needs (so far just power tools pack rebuilds) there is plenty of headroom.

For making large packs, I am sure it can get in the way but I haven't had that come up yet.
Replacing the pole with a galvanized pipe with a couple of 90 degree elbows would certainly allow for larger packs to be welded.
It should be an easy bridge to cross if I ever get there.

 

[leffex]

Hello again

I just updated my a123 life battery of 2s2p with another 2s2p of a123 26650 s (to use it instead of my other battery 4s40AH lipo 12C rated 1750A in CAL)

From having about 800A of power at the CAL it now outputs between 950-1077A, charged and sitting at 6,58v so not fully charged but still.

What I tried welding was 0,15 to a 0,2mm piece of pure nickel and it just don't grab hold but gives very weak welds. I use 10-15 J, maybe that's why.
Before undercurrent kicks in.

I need to upgrade the battery even more

...

 

[skeetab5780]

I would just get the Ultra Cap module and use a power supply, its easier than having to monitor batteries at high current and charge

obviously it cost a little bit more but it will last longer

 

[Hummina Shadeeba]

For 100$ I’ll give u 100 A123 cells. All 95-99% capacity. The capacitor bank will last longer of course but even 50 cells are more than enough. Leave it connected to the power supply and easy and likely best bang for buck.

 

[Rafalkut]

Can you help me to investigate demage ?
Spot V NE error dont go calibration show overcurent 20 000A from battery 700A rated

https://youtube.com/shorts/rGBgtH2PVdI?feature=share

 

[Joshlotz]

4 of the capacitors on my kcap blew up on connection. Everything wired up as described in the manual. Positive and negative correct.
Hooked caps up to a 15v 40ah lab supply from powertech. Set volts to 8.2v and aps to max (40). Hooked up leads then turned on power supply. Light for overtemp was on and kcap had a lot of heat coming off it. Also 2 balance lights on. Two steady lights for topping up were on also.
After about a minute one cap blew up, shortly followed by 3 more.
No idea why this happened other than potentially the kcap being faulty, or maybe affecting the board with static electricity during assembly. (I did ground myself repeatedly) [