kWeld - "Next level" DIY battery spot welder

bigbore said:
I'm almost ready to turn my Kweld on for the first time. These are the massive 4AWG cables with 10mm diameter brass electrode holders. The electrodes I'm going to try for the first time are made with 3mm copper rode.

20180210_003327_1.jpg

20180212_231542_1.jpg

Hi, where I can buy the same electrode holder which support 4 awg wire?

Thanks
 
It seems that using kWeld with 200J (over 1600A and 100ms) is not ment to be. I have already blown 3pcs of 300A fuses.
I need these settings to weld 0,2mm copper to cells (and can do it only when I split the copper).
And yes, I'm using 0,1mm nickel on top of copper.

With 20J or even 60J there is no problem when welding regular 0,1mm or 0,2mm nickel. But over 150J I can see how the fuse slowly melts with every weld.
 

Attachments

  • InterConnections.jpg
    InterConnections.jpg
    132.7 KB · Views: 1,653
  • InterConnectionsCloseup.jpg
    InterConnectionsCloseup.jpg
    195.6 KB · Views: 1,653
Try 0.15mm copper, and thin steel sheetmetal for parallel. Also put one probe on the cell, and one probe on the welding point. That will remove half the current from shorting across the nickel, forcing all of it to go through a single weld point.

As posted earlier, in the conventional technique (two probes on the nickel strip when welding) one probe is hotter, and after the current passes through two weld points, the current coming back through the cooler probe is weaker. It is the heat that makes the world, so put the hot probe on the top of the bus welding spot, and the cooler probe on the cell-tip.

This way, you will get the hottest weld with the lowest setting. If you want four welding spots per cell-end, then you must make four separate welds with this technique.

0.10mm copper for series has been proven to work, and it has the conductivity of 0.40mm thick nickel.

0.15 copper would be an improvement for high-amp cells, but 0.20mm copper (equal to 0.80mm nickel) is problematic. How many pack amps and amps per cell do you want?
 
spinningmagnets said:
0.15 copper would be an improvement for high-amp cells, but 0.20mm copper (equal to 0.80mm nickel) is problematic. How many pack amps and amps per cell do you want?
The pack will be around 100A-150A continious (~8A per cell). But I have 3 separate banks and I need to make interconnections between these. I intend to make these with 12AWG wires 4 pcs-s. And also output from the negative/positive cell the same and then to BMS that has 2x 8AWG input/output wires and I will use 6AWG wires coming out of the battery. So 4x 12AWG -> 2x8AWG -> 1x6AWG for negative and 4x12AWG->1x6AWG for the positive.
So this 0,2 copper (0,2x10=2mm2) must carry the same amperage as 12AWG wire (4mm2).Copper is only few cm in length but wire 7cm then this will equal out.
 
It looks like 26P, yes? That adds up when you say 8A per cell, and 150A peaks for the pack. I apologize for not looking at the pictures more closely before.

Since the pack(s) is/are already assembled, your only option is to overlay copper onto the current bottlenecks.
 
spinningmagnets said:
It looks like 26P, yes? That adds up when you say 8A per cell, and 150A peaks for the pack. I apologize for not looking at the pictures more closely before.
No its 18p20s pack using 2000mAh cells that i have tested with 10A continious discharge (and concluded that 8A is ok for continious).
spinningmagnets said:
Since the pack(s) is/are already assembled, your only option is to overlay copper onto the current bottlenecks.
Yes it is assembled and nickel for series is fine but as i said I need interconnections between these three parts.
And for that I'm using copper out to the side (because they will be against each other like in picture) and then 12AWG wire.
 

Attachments

  • WeldingDone.jpg
    WeldingDone.jpg
    241.4 KB · Views: 1,636
nuxland said:
200J (over 1600A and 100ms)
If you can get that 1600A up closer to 2000A you can probably use a lot less joules and get pulse time also down and not melt fuse. With copper pulse that long just heats the copper all around the weld and not just the spot weld.
 
ossivirt said:
nuxland said:
200J (over 1600A and 100ms)
If you can get that 1600A up closer to 2000A you can probably use a lot less joules and get pulse time also down and not melt fuse. With copper pulse that long just heats the copper all around the weld and not just the spot weld.
Thats no problem with my supercapacitor pack. Right now I have it at 11,8V because more than that CAL will give me overcurrent.
But yeah, I could cheat: Calibrate at 11,8V and then turn it up to lets say 12,5V, then there should be around 2000A. But then I have a chance to blow something in the welder also I think.
Actually with some welds current is around 1800A and 99ms. And you are right with 2000A and 80ms it would weld better.
 
Hello :). I am a new user of kweld. I have a problem. I got it yesterday I connected to the Varta battery during calibration, it reports 1550A. The problem is that once it welds, I think that after a while it burns everything even at the minimum settings, for example 2J. Reports then exceeded time It is random.I'm sorry for my English :)
 
Are there any Kweld kits available in the US? The 18650shrinkandcellholders.com site seems to be out of commission.
 
Just heard back from Keith at 18650shrinkandcellholders. If anyone is interested, check here:

https://docs.google.com/spreadsheets/d/1QjoAae693dKpFUWA9E86qQfWChKewNNyRHhEMz0IoLI/edit#gid=0

Or contact thru the ebay store:

https://www.ebay.com/str/18650shrinkandcellholders

Also thanks to SpinningMagnets for the info. :thumb:
 
Pawlo2525 said:
Hello :). I am a new user of kweld. I have a problem. I got it yesterday I connected to the Varta battery during calibration, it reports 1550A. The problem is that once it welds, I think that after a while it burns everything even at the minimum settings, for example 2J. Reports then exceeded time It is random.I'm sorry for my English :)

Sounds like you have bad connection or dirty tips or something. Usually when I burn things my connection was bad and sparks fly. However kweld should have protection against that too. Cant remember the correct term for that but its adjustable. What kind of readings you get after good and bad weld?

To me varta sounds like AA battery but that must be car battery this time :lol:
 
Daruben said:
I don't think they need to be twisted since its DC.
No the inductance of a twisted pair is lower than a parallel pair, but the difference is small. Most important is to reduce loop area if you consider the current loop to be a single winding coil.
 
Bought kWeld with stock cables
Getting 1200A welding current with this battery https://hobbyking.com/en_us/turnigy...tery-pack-w-xt60-connector-roar-approved.html (140C 3S LiPo) @ 12.6V
Is this result good enough? Thinking of getting second one in parallel for the nickel plated steel & copper sandwich method.
I assume 1200A won't cut it for welding any copper. I'd like to keep the current cables like they are.

As for weld quality, I'm happy with the results. 0.2mm nickel is very easy to weld. Seems worth to me
 
silentbike said:
Bought kWeld with stock cables
Getting 1200A welding current with this battery https://hobbyking.com/en_us/turnigy...tery-pack-w-xt60-connector-roar-approved.html (140C 3S LiPo) @ 12.6V
Is this result good enough? Thinking of getting second one in parallel for the nickel plated steel & copper sandwich method.
I assume 1200A won't cut it for welding any copper. I'd like to keep the current cables like they are.

As for weld quality, I'm happy with the results. 0.2mm nickel is very easy to weld. Seems worth to me
As you said. 1200A is plenty for nickel and nothing is enough for copper basicly :lol:. With nickel copper sandwich you probably get better results with more current. With 2nd battery paralel you may or may not go over current limit. Really no easy way to be sure before you try with your batteries and connectors.
 
Why people say not copper. I’ve done thousands of welds of .1mm copper and some even .2mm copper if w a slit in it and that’s on the a123 cells which have a very thick can.

This is thinner .05mm copper folded for better flex without breaking.

https://imgur.com/a/zgOZjfx

All welds here are the 2x .05mm strip. Strangely can’t get it to weld to the thick can though and give up and don’t want to try higher joules. How high is ok to do? But I can weld .1mm copper on the can ironically. Dont know why this won’t.
 
I’m getting repeated timeouts when calibrating and doing the “short” part. Luckily it still will weld just can’t calibrate again. Should I do Higher voltage? Pretty sure my battery isn’t to blame for a lack of power delivery, which I think is what the timeout means..not getting some amount of current out within a time. Maybe my battery connections could be to blame and should beef it up?
 
I'm in the same boat Hummina. But I'm pretty sure its my anemic 3s 10ah lifepo4 pack. Had it at 70% SOC and it only delivered 300 amps. By default the Kweld looks for an 800a or better pulse.

Whats weird is I tried a very new and healthy turnigy panther graphene 6s 4ah 75c pack. Didnt even try going thru the calibration as it made the kweld emit some static crackling noises. That pack was also at storage charge so I guess the kweld just didnt like 22.2v. Doesnt look like the kweld was damaged in any way. Just weird to hear that static and it did subside after a few seconds. Not going to attempt powering it with 6s again.
 
So what are the pack specs to get top output power?

Voltage range?

Actual peak vs continuous maximum amps drawn?

What are the specs that might get to be "too much"?

Would a pack tested capable of true 25C continuous be better or worse than one that can only handle 15C?
 
john61ct said:
Would a pack tested capable of true 25C continuous be better or worse than one that can only handle 15C?

what metric is used to qualifty what is better or worse?
 
Apparently the welding circuitry does not limit current drawn, I've read about batteries delivering "too much" power?

I was hoping (if that is the case) there might be guidelines for getting to "as much as possible without going over".

Or maybe going to big lead units would be better, can get huge capacity but with more resistance not releasing such a high C-rate?
 
I think to get max power out of the welder requires getting a battery powerful enough, which isn’t hard, and cutting its wires so the resistance is just enough that you don’t get an “over current” and the wiring decides the current. Or you could trick the calibration process and go beyond its amp limit by calibrating and then shortening the wires. And then after you calibrate to have the peak amp draw possible crank the joules to their peak which will close the circuit longer. Is it 200 joules possible?

I don’t know if the welder takes the increasing resistance from the heating wires into account and I imagine that could change the output a lot.
 
Yes better to rig a current-limiting circuit that does not vary so much as voltage / SoC drops.

Of course a PSU / charger powerful enough to "just barely almost" keep up with the welder's Ah/minute consumption

would allow for a smaller battery, just using it as a short-term buffer, maintain a steady state, near constant SoC.
 
Back
Top