kWeld - "Next level" DIY battery spot welder

I agree that they don't owe me anything. That is why I tried to describe the problem as broadly as possible. I have even added resistance symbols for better clarity. I even believe that we are somewhere close to the truth, but this requires an outside opinion.
And I ask you to forgive me for my wording as I'm not a native speaker.
 
Thanks for designing a great welder! Such an improvement to my earlier car battery-solenoid-timer relay DIY welder.

LTO seems to work great as a power source as well. 4S Yinlong 40Ah cells produce 1800 amps at calibration on the stock cabling lengths. 16301798343011564373918584375593.jpg

Cells are not new, with new cells 3 would probably be fine. Also might manage with 3 if I charged these up to full voltage.
 
Here are my colleague's welding results.

Due to the lack of complete confidence in the correctness of our calculations, I cannot provide you with the amount of consumed energy in Ws. However, I can say that welding took place at a voltage of 4.0V. Two welding pulses of 1.5ms and 5.0ms duration were used.

There is an assumption that the total energy including losses was 50Ws of which only about 19-24Ws was used to obtain welded joints.

Resistance of an open welding circuit (from electrode to electrode) with open mosfets (Ri + Rs +Rc) is 1.28mΩ.
On the photo is original of Sony Murata VTC6 cells and pure nickel 200μm.



IMHO, this is how welding spots should look like. This is already close to the quality obtained with industrial spot welding machines.
 
john61ct said:
So what input voltage range do you use with the 4S?

How low would you go with 3S?

4S was at 9V when I tested it. Fully charged would be about 10,4V. 4S voltage range is between 10,4V-8V

3S would work between 7,8-6V.

I tried 2S first, it gave calibration current of slightly over 1000A, so would also work if I lowered the 800A min current value on the welder. Not optimal though.

These are also the worst cells from 30 I ordered last December, with the other 26 being in a pack already.
 
Looking for some advice on power source configuration.

I have 2 6V 225ah on hand that I'd like to use with my kWelder. While they're not CCA rated, they are pretty hefty batteries. Is it better to configure them in series or parallel for the welder? Or should I not bother and just use 1?

I also have a standard 6V/12V SLA 6 amp battery charger I can run concurrently while welding which seems like a good idea.

If it helps I'm planning to use the sandwich method with .2mm copper "H"-strips and either .15mm nickel "H"-strips or .15mm nickel-plated steel "H"-strips if the pure nickel doesn't work out. All of these are slotted at the weld areas.
 
vanturion said:
If it helps I'm planning to use the sandwich method with .2mm copper "H"-strips and either .15mm nickel "H"-strips or .15mm nickel-plated steel "H"-strips if the pure nickel doesn't work out. All of these are slotted at the weld areas.
Doubt this will help you. Frank himself wrote that he could not weld even 100μm of copper.
 
vanturion said:
Looking for some advice on power source configuration.

I have 2 6V 225ah on hand that I'd like to use with my kWelder. While they're not CCA rated, they are pretty hefty batteries. Is it better to configure them in series or parallel for the welder? Or should I not bother and just use 1?

I also have a standard 6V/12V SLA 6 amp battery charger I can run concurrently while welding which seems like a good idea.

If it helps I'm planning to use the sandwich method with .2mm copper "H"-strips and either .15mm nickel "H"-strips or .15mm nickel-plated steel "H"-strips if the pure nickel doesn't work out. All of these are slotted at the weld areas.
Series unless you get overcurrent error. Then paralel and if thats no good then just 1. You also might have to get thinner copper or thinner steel strips if you are not able to get good welds. 0.2 copper is just so hard to weld. You need weld current close to kweld max current just to have chance..
 
Silvaticus said:
Doubt this will help you. Frank himself wrote that he could not weld even 100μm of copper.

srapy's post and spinningmagnet's last link on the Copper/nickel sandwich thread gives me some hope it'll work out. That and it doesn't look like they used slotted copper either. Crossing my fingers..

ossivirt said:
Series unless you get overcurrent error...

OK - it won't be hard to reconfigure either way, but I'll start like you say unless I hear otherwise.
 
vanturion said:
srapy's post and spinningmagnet's last link on the Copper/nickel sandwich thread gives me some hope it'll work out. That and it doesn't look like they used slotted copper either. Crossing my fingers..
Quite interesting results. I'll have to buy copper and try it. Moreover, we have the opportunity to use almost twice the current of kWeld average with a pulse duration of about 10 ms.
Thanks a lot, I haven't seen this post before.
 
Designed a new case more friendly for printing:

kWelder Case.jpg

Very minimal support material required only at 4X captured nuts features. Just needs a small sight window for fuse.
 
Has anybody made a kind of base plate to connect the kCap and the kWeld physically, so that it is one part?
 
Hey folks, I'm a little embarrassed to post this here, but I was hoping someone could provide some guidance.

I got my kweld recently and was incredibly pleased how well it worked straight out of the box. The plan was to build a case for it to protect the electronics, but I was too distracted by actually using it to build my battery. Then one day while carelessly adjusting a nickel strip, I accidentally touched one of the electrodes to the display unit and zapped it.

The welder doesn't work now. I ordered a new display unit hoping that I only fried the electronics on that, but after installing the new display, the welder still shows no sign of life. When connecting it to my power supply there is no beeping, nothing on the display and the foot pedal does nothing either. I can measure power flowing in various parts of the welder, but I'm not sure how to diagnose which component has failed. Aside from the one pin on the display that has visible damage (the electrode touched the leftmost pin on the display board, see photo), the rest of the components look fine. The bottom of the board did start to feel warm after attaching the power supply.

Any clues as to how to diagnose the issue would be greatly appreciated!

Edit: Did some troubleshooting with a friend. Our best guess is that the MCU is likely cooked. Will either need to get a new one and reprogram it, or order a whole new board.

IMG_20210913_091423.jpg
 
Hello everyone, I just finished setting up my unit and everything seems to be working but I have some questions about the calibration that the manual does not appear to address. For calibration the manual mentions the resistance rating and also says it should be somewhere between 2.5 and 3. While it does state that a higher number than 3 is not a good thing it does not say if there's anything to be concerned about with a lower number than 2.5. I am currently getting a reading of 2.11. I'm also not entirely certain what the rest of the readings given are for. I assume that "I" is the measured amps which i show as 1816. Which as I understand it is more or less exactly where I want to be since I want to be close to 2,000 but not over. I have included photos of my calibration results could somebody please let me know if everything looks good? I just want to confirm before I go zapping things :)
kYJxtkm.jpg

s25jM0A.jpg

dik0dVK.jpg

VghkZYx.jpg


Don't worry about the bare board I did get the laser cut housing kit but I wasn't smart enough to get glue ahead of time so I'm waiting for the glue to help protect my welder. Any insight you all can provide would be greatly appreciated!
 
am283721 said:
The welder doesn't work now.
This forces thousand amps through the logic board, which usually breaks the ground connection (look underneath the board for burn marks). If that has happened, then the board is beyond economic repair as this causes most chips to fail. Also, the MCU is preprogrammed with a custom encrypted bootloader, which I don't have tools for to prepare.

LordKitsuna said:
For calibration the manual mentions the resistance rating and also says it should be somewhere between 2.5 and 3. While it does state that a higher number than 3 is not a good thing it does not say if there's anything to be concerned about with a lower number than 2.5. I am currently getting a reading of 2.11.
The new heavy duty probe system has a lower resistance, which I haven't updated in the manual I'm afraid. Your readings are perfect.
 
tatus1969 said:
LordKitsuna said:
For calibration the manual mentions the resistance rating and also says it should be somewhere between 2.5 and 3. While it does state that a higher number than 3 is not a good thing it does not say if there's anything to be concerned about with a lower number than 2.5. I am currently getting a reading of 2.11.
The new heavy duty probe system has a lower resistance, which I haven't updated in the manual I'm afraid. Your readings are perfect.

Thanks for the confirmation, I ended up going with a heavy duty starter battery it was rated for 900 CCA so I figured that at room temperature it would likely be pretty close to 2k. That might be a lazy way of looking at it but it appears to have worked out for me I'm very close. I have been reading a lot of posts about people trying to weld copper strips I just placed an order for some 0.1mm copper and 0.15 pure nickel looking forward to testing both separate and sandwich.

Although I will need to redo my calibration I did extend the length of the battery cables using some 4AWG wire (just lazily by braiding them into then soldering together with the existing wire) but I assume that using that different size cable screws with the math so I'm going to need to just go get some proper cable at the length I want so I can calibrate it properly
 
vanturion said:

Very clean in general, just a few ideas I had:
- add a 4,5mm lip around the display to the lid
- add parts to the lid, that compensate for the slots and close the holes/gaps
- fuse window could be as small, as the grid-slots (just for design)
- try to reuse the original 4 stand-offs and screws for installing the lid to the body
- add a version for JST plug for the foot-switch
- add countersinks to make the screw sit flush with the lid

 
Some good ideas in there for sure, and I like your JST mod.

Because I like ya, my discounted design rate is $40/hr, min 3 hours. LMK :wink:
 
I thought i posted this but I don't see it so i guess not. Wanted to ask, if i am understanding the math on calibration in the manual correctly it seems like wire size matters. I was going to use 4AWG for the connection to the battery but is the calibration expecting 8AWG in particular? If so is there any kind of conversion calculation i can easily do or is it better to just stick with 8AWG
 
Back
Top