kWeld - "Next level" DIY battery spot welder

[ossivirt]

even the offical bussman fuses are well under a tenner,

https://www.amazon.com/Cooper-Bussmann-ANL-300-Current-Limiter/dp/B000WSWXDC

I’m doing 16 volts. I guess could charge up to18 for 5s iron cells. But why you two don’t think there’s a problem w the welder? Set to 8 joules at 16v w a good battery it blows and the time on seems extra long I’ve also had “timeout” happen when the battery drops close to 14v but not before. The timeout warning telling if the battery isn’t strong enough to put out the joules within the time period. When it does do a timeout there’s a lot more current going through snd the wires get much hotter, but no fuse blown

your battery cant get the amps fast enough so it takes longer wich results in a popped fuse.

Kweld has also software overcurrent and overtime limits. I have no idea what is wrong with his welder but something seems off.

Hummina Shadeeba if possible you should post your calibration results and all the info that is shown after weld (time, current, joules, resistance). Also have you done any modifications to welder or electrodes? That may tell us if something is wrong. You can also send email to developer so he can help you with this.

 

[flippy]

But the welder “timeout” should happen if the amount of time that the welder uses isn’t enough to get out what joules it’s set to. 31joules in my case. I’ve had that happen lots of times in the past when the battery was going down to 14v. But this happens on the 16v full charge every time and when down to even 8joules.

wich means the battery is gone.

its just like a worn out car battery, it can keep the lights on but cant start the car.

a IR test and a capacity test would show the battery is degraded.

 

[Hummina Shadeeba]

These are good cells in that they’re used from batterhookup and capacity tested full capacity or within 95% by me and I only get timeout when they drop to 14v. At 16v and 31joules I had no timeouts. 5s15p a123 26650

even the offical bussman fuses are well under a tenner,

https://www.amazon.com/Cooper-Bussmann-ANL-300-Current-Limiter/dp/B000WSWXDC

I’m doing 16 volts. I guess could charge up to18 for 5s iron cells. But why you two don’t think there’s a problem w the welder? Set to 8 joules at 16v w a good battery it blows and the time on seems extra long I’ve also had “timeout” happen when the battery drops close to 14v but not before. The timeout warning telling if the battery isn’t strong enough to put out the joules within the time period. When it does do a timeout there’s a lot more current going through snd the wires get much hotter, but no fuse blown

your battery cant get the amps fast enough so it takes longer wich results in a popped fuse.

Kweld has also software overcurrent and overtime limits. I have no idea what is wrong with his welder but something seems off.

Hummina Shadeeba if possible you should post your calibration results and all the info that is shown after weld (time, current, joules, resistance). Also have you done any modifications to welder or electrodes? That may tell us if something is wrong. You can also send email to developer so he can help you with this.

I’ll see if I need to get some fuses and shoot again or maybe it has the record of last time and I’ll check for that and post all.

And i found a monthly tool rental space in San Francisco and was going to come in snd use theirs today but he called back snd said his kweld is broken.

The kweld I was just using is also borrowed waiting for them to be in stock.

My boss welder broke before.


I’m suspecting if misfiring is a danger, such as when u don’t get the tab down hard against the cell face, andit blows, (which I rarely have happen with these electrodes n kweld as apposed to what the boss comes with), maybe that could be the cause and these recycled cells have bumps on them. Maybe I need to avoid them better

Other than that I’d been at 31 joules and 16-15v happy for hundreds of welds with no timeouts doing .2mm copper.

 

[Addy]

If you have a crappy fuse installed, the resistance of the fuse itself could be limiting your weld current. Worth trying a proper fuse IMO.

 

[flippy]

These are good cells in that they’re used from batterhookup and capacity tested full capacity or within 95% by me and I only get timeout when they drop to 14v. At 16v and 31joules I had no timeouts. 5s15p a123 26650

Other than that I’d been at 31 joules and 16-15v happy for hundreds of welds with no timeouts doing .2mm copper.

*my car started fine all these years and now it suddely wont start so it cant be the battery*

i said you need to test capacity AND internal resistance. capacity is not the only metric to check a batteries health.

 

[john61ct]

https://www.mouser.com/Circuit-Protection/Fuses/_/N-ba85y?P=1z0x661Z1z0s1tm

Some of these are pretty reasonable.

Apparently slow blow costs more

and the time-delay ones wow.

Note, on eBay/Amazon/Chinese sources

Cooper Bussman is possibly counterfeit.

Best to buy from mouser, digikey, allied, newark, arrow, element14, avnet. . .

https://www.eevblog.com/wiki/index.php?title=Component_Search_Engines

 

[Hummina Shadeeba]

Thanks. I already ordered this one:
https://www.amazon.com/gp/aw/d/B000WSWXDC/ref=ya_aw_od_pi?ie=UTF8&psc=1
And this one: https://www.amazon.com/gp/aw/d/B009BA1ZWC/ref=ppx_yo_mob_b_track_package_o0_img?ie=UTF8&psc=1
Maybe I can see how legit the fuse is measuring it’s resistance.

They are quick to come so will see what happens

I’ll post the complete battery resistance but in my looking capacity is closely aligned with resistance and the tests are often interchangeable. 75.5 amp hour battery shouldn’t have a problem all of a sudden and it’s been good so far but I’ll do the test. Just have to find a giant resistance I think. Don’t think I’ll find one and need to buy one. Guess could do each p group test separately and need a smaller resistance. I’ve done a charge test only, measuring how much energy the batter holds from two voltages when charging at 5 amps
But as I said I think the “timeout” would be happening instead of blowing a fuse repeatedly especially at only 8 joules

 

[Hummina Shadeeba]

i tried to calibrate with some wire for a fuse. i dont think should be doing nearly this current and i get strange results. no data showed for the "open" part which isnt right, and the "short" part i get a huge current and other weird data. the welder has no burn marks though and all the bits look good. it was just my make-do fuse 12awg wire that smoked a bit at its connection

i have fuses coming but im not going to bother trying with them and for sure something is wrong with the welder. or i could use one to get one shot at results i could post, but i think the calibrate data should reveal a lot as is.

https://l.facebook.com/l.php?u=https%3A%2F%2Fimgur.com%2Fa%2FkBfmGkc%3Ffbclid%3DIwAR1N9fKrgDOJuU81X2dmjluxKuQU1eRpWr6CSasUX0-H_yLKmDJwww7-PMw&h=AT1ynfr8wcmaURSclty6fZNj3cE6gwq-QGx2I_pTNgCSwss8e7fzHZtT_VqT7dpUu4qI80XDRDFCPoJnvnBgXb_Mw-nzNbVpmD87kwnHxi1HY9vIQ2eCBfdl59MMeMKSXFA2UCzgqMjSwg&__tn__=R]-R&c[0]=AT2bD6Yo1rSMphYfKhiY8YvvLhwu5n2UUAnXdwQNvVE7Cx9NqXQfSmp-MUhqsasBXOYiUiH5HIFHGZDM6xhkQ4JElixNSk1jcBo8RGB0VfuEpe-EQri5m_P595GlC8ZIfseCLGnG3pAOoZ8ckFbiEWWuRjqjAHoIe4MAqJuy8akqXWzQTN3qxTy7tgmyDq2UEm9j6uR3moA


and ive been doing .1mm copper sorry not .2mm as i wrote earlier

 

[flippy]

capacity testing is done with discharge, not charge. the IR also changes with SoC so you need to measure that at intervals during the charge/discharge as well.

fuse son welder pop because you give them enough time to heat up and melt it. that is why you can push 2000A with just a 300A megafuse. the pulse is simply too short to melt it. but if the current output drops (you can test this with the welder with the callibration feature) of the battery the welder needs to stay on longer to reach the desired joule setting wich gives the fuse enough time to melt.

 

[Hummina Shadeeba]

Makes sense that u could measure capacity while charging too no? Maybe not as accurate but that energy going in..is energy going in. But after doing the calibration and hitting 1800 amps I think that’s a sign the battery is solid. It’s huge at 5s15p. 37amphours of a123. More the isssue I think is having been using 16 or 15 or even 14 something volts and that’s when I’d hit timeout. Too slow to get the current across. But then again I only had timeouts a couple times when the battery was in the 14 volt range. But that’s not even my problem, I keep blowing fuses. Will try in an hour when fully charged

Got more fuses. Rockford this time. Looking on the kweld site at the fuse for sale or the two instructional PDFs nothing about a slow fuse. They sell a fuse for 3$.

So going to 18 volts and will do a calibration w the fuse n see what data I get. From what the guide says my data is normal at least in the shorting part. The open part I have to do again

But I was talking with the designer and he said do a diode test on everything. I’m doing it with the fets and while the resistances are consistent for all of them I only get the OL reading when touching the center pin that doesn’t come back to the board if u know what I mean. Otherwise I’ll get a resistance going either way(less one way) on all the fets. Seems I’m doing it wrong snd surely I didn’t blow all 6 fets. https://www.fluke.com/en-us/learn/blog/digital-multimeters/how-to-test-diodes

 

[flippy]

Makes sense that u could measure capacity while charging too no? Maybe not as accurate but that energy going in..is energy going in.

you stop charging at intervals so you can measure properly.

and the energy going in is the least important metric. the only thing that matters is energy going back out again. but the relation between how much goes in vs how much came out does tell you a lot about the wear level of the battery.

 

[Hummina Shadeeba]

Makes sense that u could measure capacity while charging too no? Maybe not as accurate but that energy going in..is energy going in.

you stop charging at intervals so you can measure properly.

and the energy going in is the least important metric. the only thing that matters is energy going back out again. but the relation between how much goes in vs how much came out does tell you a lot about the wear level of the battery.

what you think about diode testing on the fets? am i doing it wrong? the designer said to test everything with a diode short test.

 

[john61ct]

Wut 14V for 5S is 2.8Vpc

Not a good sign, at all! Especially if your "big" Ah pack means you're keeping C-rates reasonable.

I mean, you're not "damaging" other than losing lifecycles

you bought them used right? likely past EoL by industry standard 80% cap?

If not, definitely a sign your discharge cycles should be shorter, recharge more often

maybe consider putting a big amp charge source on concurrently to help support the voltage.

What do they recover to, isolated an hour before recharging?

 

[Hummina Shadeeba]

it was a mistake going down to 14v. it was in the high 14s though at least. i was surprised how much juice the welder uses. i was going between 16 and 15. not the wider range of ion or lipo chemistry. its pretty new to me.

connect it to the power supply while welding? i like the sound of that.

the capacity of all the ones i tested were at min 95% capacity. doing a make-do test just going from voltages while charging. batteryhookup claimed them at 95 percent. if you havent checked batteryhookup theyve got great deals. dont have these cells anymore though.


++but i assumed too much and pretty much kept skipping a page related to wire length. i skimmed through and assumed shorter wires the better. duu. i just had added new wires and theyre fat and very short. doing calibration it shows 2.55milliohms and the baseline is 4.7. i have to lengthen my battery wires to build resistance.

lengthened battery wires and all is firing fine. have to actually read the instructions.

 

[Hummina Shadeeba]

Makes sense that u could measure capacity while charging too no? Maybe not as accurate but that energy going in..is energy going in.

you stop charging at intervals so you can measure properly.

and the energy going in is the least important metric. the only thing that matters is energy going back out again. but the relation between how much goes in vs how much came out does tell you a lot about the wear level of the battery.

whats the easiest way to measure the battery resistance? i wonder how conductive my somewhat weak welds are.

 

[flippy]

whats the easiest way to measure the battery resistance? i wonder how conductive my somewhat weak welds are.

search the "oriental supermarkets" for a SM8124A or something to that effect if you want fairly accurate and cheap for batteries under 100V.

or do it the old fashioned way by measuring voltage drop under load. but that tends to be inaccurate with the stuff most people have on hand.

 

[ossivirt]

Why even try to measure ir if you use the battery for kweld? Just use kweld to measure amps during weld pulse and that should be enough in my opinion. 1500-1999A if you want to try copper and if just nickel up to 0.2mm little less is enough. I would like to have over 1000A anyway.. If battery can do that then ir reading is just a number without context

 

[Hummina Shadeeba]

Why even try to measure ir if you use the battery for kweld? Just use kweld to measure amps during weld pulse and that should be enough in my opinion. 1500-1999A if you want to try copper and if just nickel up to 0.2mm little less is enough. I would like to have over 1000A anyway.. If battery can do that then ir reading is just a number without context

I want to figure the resistance of the completed battery i make with the kweld and .1mm copper. the a123 cells are easy to weld to on the negative side but the positive welds are weak and i wonder how conductive they are. im already at 1600amps and the battery wire length are over what they should be and creating dangerous inductance as is and dont want to ahve to go higher current

I cut my wires down and get over current, so make them longer snd then risk the inductance. I go back and forth with different lengths and there’s no solution to get beyond 1500 amps without wires that are too long.

 

[ossivirt]

Why even try to measure ir if you use the battery for kweld? Just use kweld to measure amps during weld pulse and that should be enough in my opinion. 1500-1999A if you want to try copper and if just nickel up to 0.2mm little less is enough. I would like to have over 1000A anyway.. If battery can do that then ir reading is just a number without context

I want to figure the resistance of the completed battery i make with the kweld and .1mm copper. the a123 cells are easy to weld to on the negative side but the positive welds are weak and i wonder how conductive they are. im already at 1600amps and the battery wire length are over what they should be and creating dangerous inductance as is and dont want to ahve to go higher current

I cut my wires down and get over current, so make them longer snd then risk the inductance. I go back and forth with different lengths and there’s no solution to get beyond 1500 amps without wires that are too long.

Ok. I assumed that you were talking about the battery you use for the welding.

Even weak welds may be ok when measurin ir or current but real issue can be how well it can handle bumps from riding.

You really should figure out how to keep your wirelenght inside the spec. That likely can cause problems for the welder..

 

[Hummina Shadeeba]

Why even try to measure ir if you use the battery for kweld? Just use kweld to measure amps during weld pulse and that should be enough in my opinion. 1500-1999A if you want to try copper and if just nickel up to 0.2mm little less is enough. I would like to have over 1000A anyway.. If battery can do that then ir reading is just a number without context

I want to figure the resistance of the completed battery i make with the kweld and .1mm copper. the a123 cells are easy to weld to on the negative side but the positive welds are weak and i wonder how conductive they are. im already at 1600amps and the battery wire length are over what they should be and creating dangerous inductance as is and dont want to ahve to go higher current

I cut my wires down and get over current, so make them longer snd then risk the inductance. I go back and forth with different lengths and there’s no solution to get beyond 1500 amps without wires that are too long.

Ok. I assumed that you were talking about the battery you use for the welding.

Even weak welds may be ok when measurin ir or current but real issue can be how well it can handle bumps from riding.

You really should figure out how to keep your wirelenght inside the spec. That likely can cause problems for the welder..

if i shorten my battery wires i get an "overcurrent" when calibrating. i need to add resistance but not inductance somehow. ive gone back and forth adding and cutting trying to find a sweet spot but if i go shorter than maybe a meter and a half ill get the overcurrent and cant weld. from what i read the welder alters the output time period and not the current and the current is fixed when you do the calibration based on the resistance of the whole system.

 

[chilau2000]

Hi

Does anyone have a KWELD kit for sale and that can ship to UK ?

KWELD aren't currently shipping to UK

Thanks

 

[ossivirt]

if i shorten my battery wires i get an "overcurrent" when calibrating. i need to add resistance but not inductance somehow. ive gone back and forth adding and cutting trying to find a sweet spot but if i go shorter than maybe a meter and a half ill get the overcurrent and cant weld. from what i read the welder alters the output time period and not the current and the current is fixed when you do the calibration based on the resistance of the whole system.

Smaller diameter wire would add resistance but that would get hot. Less cells on paralel or series will reduce current. Or just different battery if you have. Car starter or rc Lipo maybe

 

[Hummina Shadeeba]

I’m saying if i shorten my battery wires too much I get “over current”, but when I lengthen them, to like 2meters including probes now, I have too much inductance to do more than maybe 1200 amps as shows on their posted chart. I want to do more. How?
https://imgur.com/a/qu13QBy

 

[flippy]

lower the SoC of the battery or take of p cells

 

[Hummina Shadeeba]

thanks and i lowered battery soc and can now weld with half a meter of wiring including the probes and dont get an overcurrent. I have no problem getting good welds with .1mm copper using only 25 joules on the negative and 50 on the positive A123 green cells. i imagine .2mm copper wouldnt be too hard depending on the cell and i can weld .2mm copper to the negative of these cells at 75j but not the positive. Is there any other limit or am i free to go to 100 joules?

the temp limit is 70c but I stay to maybe 40c max taking extra time when doing 50j.