kWeld - "Next level" DIY battery spot welder

spinningmagnets said:
I would highly recommend investigating the copper/nickel sandwich method for welded buses. It's a proven method that provides more than three times the series current, while also running cooler and generating less voltage sag under load.

Nickel is a poor conductor, so increasing thickness from 0.15mm to 0.20mm is fighting to desperately increase conductivity 33%...using copper for series is a 300% improvement.

If corrosion from bare copper ribbon is a concern, Keith at 18650 heat shrink and cell holders has nickel-plated copper ribbon.

Yes, excellent suggestion that I will research and test.
Only downside is, since cooper's resistance is so much lower, it is harder to weld and you would probably need more power to get it to work. We'll see.
The thought of using copper (not sandwiched with nickel) occurred to me but I have read on the forums how much more difficult working with copper is when compared to pure nickel.
At that time I even looked at the ultrasonic welders available online but the price for a good one was a show stopper.

Anyhow, thanks for the suggestion! The first step is to buy good copper. . .
 
https://endless-sphere.com/forums/viewtopic.php?f=14&t=108006

Molicel 21700 P42A cells, 0.10mm copper sheet, plus 0.15 nickel squares only to facilitate welding. 20S / 6P

kWeld used, setting is 50-Joule

file.php
 
Hi everyone,
Does a cheap crimp tool like this one is enough to assemble the kWeld ?
https://www.amazon.com/IWISS-Battery-Cable-Crimping-Cutter/dp/B00MVE48Z6/ref=sr_1_3?dchild=1&keywords=hx-50b&qid=1604947739&sr=8-3
 
This crimper should be more than enough, avoid soldering after crimping, it adds resistance.

Spinningmagnets, like the idea of little pieces of nickel to weld copper...a little effort to have bigger gains.
 
Hey

I have a question about power source. Made myself a 12v power source from 6pcs of 3000F 2,7v ultracaps.
And somehow blowed off my malectrics v2.2 spot welder with 9ms pulse (two mosfets died)
So my question is, can kweld handle that or I need less powerfull powersource?
 
For the last couple years I've been using a DIY spotwelder using a microwave transformer and a controller from avdweb.nl in The Netherlands to build E-Bike packs. Results were not that spectacular especially with thicker tab material and 26650 batteries. So I recently purchased the Kweld kit and used an Optima Spiral AGM 750CCA car battery as the power supply. It worked better but lugging that HEAVY battery around was a PITA. So, I have several A123 36volt modules that use 48 ANR26650M1A LiFePo4 cells in a 12s4p configuration. Each of these cells are rated to provide a 120amps load for 10 seconds and 50amps continuous. I purchased them from Battery Hookup, and normally use them in E-Bike packs. Battery Hookup gets them in regularly but sells out quick because they are only $49.99. https://batteryhookup.com/products/...9&_sid=8cef8175d&_ss=r&variant=31780910137476


I reconfigured the module into a 14.4 volt 4s12p pack and this thing kicks ass as a power supply for the Kweld. Way better then the 750cca car battery and super light. When I ran Kweld calibration it showed 1436 Amps during the short test. I just wanted to share this info with others.

I don't have a video on the battery build, just photos. So here goes.
1.jpg

2.jpg

3.jpg

4.jpg

5.jpg

6.jpg
With cover removed. There are several #10 torx screws holding the cover on, one under the label. Also remove 6 more screws to lift battery pack from the case. Removing the BMS requires cutting or unsoldering the connections along the edge of the board and unscrewing several torx screws. I cut the balance connectors with a knife and small snips. Be careful, don't cut yourself!
7.jpg


7b.jpg
I removed the upper and lower cell holder plates to access the batteries. The lower plate is a pain because the 4 cells on each end are joined together with an external buss. I had to cut those sheet metal busses between each of those end cells (8 in total) and fold the metal up so the lower black cell holder plate could be removed. After that I had to separate cells to reconfigure the pack. I did not want to spot-weld new tabs to the batteries. Sometimes pulling the old tabs from these cells can perforate the cell. So I cut the cell tabs apart along the red lines shown creating 12 sets of 4 cell packs. The cells sets are highlighted yellow. I then repositioned them into the areas that are highlighted blue. I've only highlighted 4 areas, but they repeat throughout the pack. I used small pieces of nickel plated copper strips as jumpers to rejoin the cells but in a 4s12p configuration. I also doubled-up the strip to create buss bars across the 12 batteries on both ends of the reconfigured pack. They were soldered to the tab material running between the cells. The 0.3mm thick nickel plated copper strip came from https://18650shrinkandcellholders.com/
8.jpg
End view showing the two female output terminals. I disassembled the terminal assembly and soldered 8awg cables into them directly. Be careful, it's pretty easy to unsolder it from the board also. Oops, I did, but just had to wire wick the solder from board and resolder the terminal block back in.
9.jpg

10.jpg
I cut and reconfigured the heavy original copper connectors, soldered them to the battery and bolted them to the original output terminals at the end of the case. Used a short piece of 8awg wire for the positive connection.
11.jpg

12.jpg
8awg cables soldered into the output terminals. They go to the Kweld. Later on I may add a 4s BMS.
 
nuxland said:
Hey

I have a question about power source. Made myself a 12v power source from 6pcs of 3000F 2,7v ultracaps.
And somehow blowed off my malectrics v2.2 spot welder with 9ms pulse (two mosfets died)
So my question is, can kweld handle that or I need less powerfull powersource?
Hello.

Someone can probably calculate how your capasitor bank behaves with kweld if you have datasheet for your caps. Or atleast ir value I think..

But I can only say that probably yes it can handle it. However if I had supercapasitorbank of unknown specs I would use adjustable cc/cv powersuply or dc converter to charge it to lowest possible voltage that kweld can operate (refer to manual). Then do calibration and check what the current was. Then raise voltage little and repeat untill you max your capasitor voltage or welding current goes close to 2000A overcurrent limit.

But if you don't have kweld yet then you need to get one :lol:
 
ossivirt said:
nuxland said:
Hey

I have a question about power source. Made myself a 12v power source from 6pcs of 3000F 2,7v ultracaps.
And somehow blowed off my malectrics v2.2 spot welder with 9ms pulse (two mosfets died)
So my question is, can kweld handle that or I need less powerfull powersource?
Hello.

Someone can probably calculate how your capasitor bank behaves with kweld if you have datasheet for your caps. Or atleast ir value I think..

But I can only say that probably yes it can handle it. However if I had supercapasitorbank of unknown specs I would use adjustable cc/cv powersuply or dc converter to charge it to lowest possible voltage that kweld can operate (refer to manual). Then do calibration and check what the current was. Then raise voltage little and repeat untill you max your capasitor voltage or welding current goes close to 2000A overcurrent limit.

But if you don't have kweld yet then you need to get one :lol:
Already bought one yesterday :)
And read like 2h the thread and figured out the same that need to start from 8V and calibrate and see test result.
https://www.elfadistrelec.ee/Web/Downloads/_t/ds/SPSCAP_SCP3000C0-0002R7STB_eng_tds.pdf
from the specs 220 micro ohm 0r 0,22 mohm. So 6 in series is 6x0,22=1,32mohm and what I read I need to add some 8awg cables to use it charged to 12v :) But as I have laboratory power supply I can start with 8V.
 
Finished my capacitor pack enclosure and added awg10 with xt90 for charging and awg6 with QS8S connector (its larger than xt150) for discharging. If awg6 is to large then I can change it to awg8 and xt150 connectors (will be making that output wire also).
Also printing out bottom enclosure with my ancient 3d printer that still works after 10 years :)
Now waiting for the kWeld, that should arrive on monday.
 

Attachments

  • Capacitors.jpg
    Capacitors.jpg
    93.3 KB · Views: 5,119
Got my kWeld and with 11V got these CAL numbers: 2,5mohmi and 1263A
Output cables are both 50cm so together 1m and intput cables are 45cm each, so together 90cm (but these are 6AWG).
So I entered 1,5m in cable lenght, because 90cm of 6AWG should be ~50cm 8AWG :)
But it seems I need to shorten these 6AWG cables (I was affraid that there would be to much current).

Tested these 4 cells with 20J and measured
13,5ms
1132A
1,17mR
28,6degrees
 

Attachments

  • kweld.jpg
    kweld.jpg
    106.1 KB · Views: 4,999
Shortened the wires and got 1700A CAL current with 12V.
Also told to the system that I have now 1,2m cables so overcurrent max should be 1800A in system now.
Resistance was also a little bit lower, at 2,48 mR
 

Attachments

  • kweld_12V_1700A.jpg
    kweld_12V_1700A.jpg
    155.1 KB · Views: 4,997
get rid of the connector, bolt the welder to the capacitor block and make the leads longer. then you can probably also weld at a lower votage or use the voltage to better regulate the available current.
longer leads are more practical for making bigger batteries.
 
Hummina Shadeeba said:
What settings u use for .1mm copper?
Around 50 joules with my 1500A battery. With copper amps matter a lot more because weld time should be as short as possible to make sure energy is used to weld and not just heating copper around the weld..

With nickel or steel on top method I have not tried yet but it should be better way to go..
 
The 0.10mm thick copper may sound thin, but since copper is four times more conductive, with 1/4th the resistance (compared to nickel), the 0.10mm copper is equal in current-carrying ability to 0.40mm thick nickel.

It's also cheaper and more readily available locally than nickel, plus easier to cut.
 
ossivirt said:
Hummina Shadeeba said:
What settings u use for .1mm copper?
Around 50 joules with my 1500A battery. With copper amps matter a lot more because weld time should be as short as possible to make sure energy is used to weld and not just heating copper around the weld..

With nickel or steel on top method I have not tried yet but it should be better way to go..



it seems easier than the boss welder in that it has just one knob for joules but it expects you to figure your battery's internal resistance (how?)

https://www.keenlab.de/wp-content/uploads/2018/07/kWeld-operation-manual-r3.0.pdf


i'd like to use it with my 5s litium iron battery at like 16 volts and need to maybe figure the wire length for that or i just keep firing from low joules and work my way up till it sticks? my wiring isnt longer than a meter.
 
spinningmagnets said:
The 0.10mm thick copper may sound thin, but since copper is four times more conductive, with 1/4th the resistance (compared to nickel), the 0.10mm copper is equal in current-carrying ability to 0.40mm thick nickel.
It's also cheaper and more readily available locally than nickel, plus easier to cut.

and it utterly destroys your welding tips.
 
Wanted to test what is the max I can use with kWeld.
So shortened my welding cables to 40cm each and input cables are 15cm each but 6AWG so together should be 1m total lenght.
Had to lower voltage to 11,8V to be able to calibrate to 1971A (12,0V gave me overcurrent).
kweld_CAL_1971A.jpg
And then tested to weld with 150J using 0.1mm nikkel to 0.5 copper (that I was using in my 25kwh A123 pack).
Got these values:
I=1834A
R=0.47mR
E=150.0J
t=136,61ms
Copper went rather hot and weld was so/so one I could tear apart and in another one nikkel teared.
It seems 0.5 copper is too much :)
kweld_0.5copperTo0.1nikkel_1834A.jpg
 
flippy said:
spinningmagnets said:
The 0.10mm thick copper may sound thin, but since copper is four times more conductive, with 1/4th the resistance (compared to nickel), the 0.10mm copper is equal in current-carrying ability to 0.40mm thick nickel.
It's also cheaper and more readily available locally than nickel, plus easier to cut.

and it utterly destroys your welding tips.
i think depends on your tips. ive been doing fine after like 1000 welds
 
I think depends on your tips. I've been doing fine after like 1000 welds

That is like <20% of a single battery for me.
 
I’m finding the battery amp ability and voltage makes a huge difference


++++ need like 30 joules to weld .1mm copper as said above and not 10 joules as i originally thought. the welds sucked+++
 
I've been having fun welding bits of metal together with my new Kweld but it's not beeping at me as described in the manual. Is this something I can fix?

Thanks!
 
pretty sure the constant beeping is a sign it's working ready to weld. dont know if any way to stop the noise. maybe rip off a chip!

how many welds is appropriate? anyone tested resistance of welds? ive been doing 4 blasts making 8 welds. doing .1mm copper.
 
DIY battery dudes I garnered a nugget of what I think is fun fascinating potential with possible pitfalls with potting in polyurethane n I don’t know what elements/chemicals are produced with zapping of copper or nickel but my cure turned bubbly instead of water clear n something has inhibited the cure.


I love sinking it deep in resin. With flesh like silicone molds. The challenges. The obstacles. A father bedside during gestation. A beauty or the danger.
 
Back
Top