Off grid power for my AC spot welder

rg12

100 kW
Joined
Jul 26, 2014
Messages
1,591
Hey guys,

I have a very strong spot welder that trips a 16A breaker all the time (220v grid) and I can't put a bigger breaker since the wires in the wall are not thick enough.

I was thinking, why don't I just take a solar power inverter (12v to 220v) that is capable for about 7-10kw and connect the inverter to a car battery (at home of course) and that way I can pull even 40A as the inverter below is rated for 10kw (probably peak but a pulse welder is pulse, not constant amp draw so...)

Like this one:
https://www.ebay.com/itm/10000W-Car...10000w+220v+inverter&LH_BIN=1&_from=R40&rt=nc

What do you guys think?
Would work as planned?
 
rg12 said:
Like this one:
https://www.ebay.com/itm/3000W-1000...00w&LH_BIN=1&_from=R40&rt=nc&LH_TitleDesc=0|0
This listing (**********) has been removed, or this item is not available.

rg12 said:
What do you guys think?
Isn't it easier to buy a battery-powered spot welder instead?
 
Fixed the link.

I have one like you say and it's pretty simple and not professional for thousands of welds everyday and without pneumatic air pressure for even welds.
 
It would be interesting to see if the inverter works. My experience of them is they suck at handling loads with high startup surges (like a welder), but then I've not played with one claiming anything like 10kW! Though I wonder how true that rating is considering how low the price is...

I take it you've tried a slow response MCB (type C or maybe even D)?

Another idea might to adapt the welder with a zero-crossing detector to reduce the inrush?
 
Replace the breaker with a D class.

That inverter is never going to work.

And just buy a battery based spot welder. The chinese transormer based ones will blow up.


Edit: just buy this and a cheap car battery.

https://malectrics.eu/product/diy-arduino-battery-spot-welder-kit-v3-2-2-full-bundle-lipo-battery/
 
Punx0r said:
It would be interesting to see if the inverter works. My experience of them is they suck at handling loads with high startup surges (like a welder), but then I've not played with one claiming anything like 10kW! Though I wonder how true that rating is considering how low the price is...

I take it you've tried a slow response MCB (type C or maybe even D)?

Another idea might to adapt the welder with a zero-crossing detector to reduce the inrush?

I don't think a slow response MCB will work since the new welder trips the switch at even half power so it's a matter of power that is missing from the grid, alot of it.

What is a zero crossing detector? will it enable me to work fast with the machine without waiting between welds?

Those inverters are usually rated 5kw and 10kw peak which means that it wouldn't shut down if I pull 10kw for a fraction of a second.
Even 5kw is about 23A at 220V which is way over my 16A at home, so why wouldn't it work?

and about that arduino welder, I have this exact one, it does a good job but you can't really work fast with it, it gets hot after a while with and it doesn't have pneumatic electrode heads that spread even pressure for each weld.
I build packs for a living, I can't work with that thing all day.
 
rg12 said:
and about that arduino welder, I have this exact one, it does a good job but you can't really work fast with it, it gets hot after a while with and it doesn't have pneumatic electrode heads that spread even pressure for each weld.
I build packs for a living, I can't work with that thing all day.

i understand you issues perfectly. i have the exact same issues. but due to pack size i cant use a pnumatic arm so manual electrodes is the only way for me.


inverters simply cant supply this burst of power. you need inverters 10x mores powerful to be able to deliver that pulse without exploding.

prehaps if you have a inverter powerful enough it might just do it but it will not survive. you will blow up the coils and transformers inside the inverter. i know this for a fact because i already tried this. i still have a real industrial class D RCD in my breaker box to prove it and a inverter that is really black on the inside.

if i were you i would focus on getting the pnumatic arm and treat the electronics behind it as a separate project. if the malectrics welder gives up i would recommend getting the K-weld with the supercap expansion and hook that thing up to the pnumatic arm.

as a fellow professional battery builder please trust me and dont trow away your money into fooking about with inverters, its never going to work with stock components. i got almost a 1000 euro's into that dead end in buying them and repairing them to prove it.

a air cooled malectrics welder and some REALLY heavy duty electrodes is the way to go until you get your pnumatic arm.

you know you can build such a thing youself? its not that hard with stock stuff from aliexpress. the wleder electronics are completly separate.
 
Thanks for the info man, you really saved me from buying an inverter!

I really don't have time to build a unit myself so I guess I will try to somehow get an electrition to come and install thick a** wires in the wall and a strong breaker while calling the electricity company to raise the amps for the apartment, I WANNA WELD COPPER DAMN IT! 0.1mm is all I'm asking.
 
right now i am using a 60Ah thundersky/winston lifpo4 battery on a CCCV ebay charger keeping it topped up and forces air cooling over the mosfet block of the malectics welder with a small 12v fan and 16mm2 1m long cables to the tungsten electrodes. it works a treat and i usually have sub 10ms welding times for 0.2 nickel. with the tungsten it should do copper without melting the electrodes to the copper.

having a desk fan blowing from above cools the electrodes (until i get water cooling to work) so i can make long welding runs.
usually i split time between cutting strips with a very small bench shear wich allows the electrodes to cool after welding 50 cells or so. also prevents getting numb hands by switching jobs every 15 min.

i got a second welder as a spare "just in case"so i can keep welding if the mosfets blow or somthing.

if the current one dies i will try the k-weld. it seems a bit more durable with the IGBT's. still, the previous malectrics one i broke lasted 56000 welds.
 
That sounds impressive although I hate working with electrodes, I need to have my hands free or at least one of them.
I will extend the pneumatic welder's head (CNC some copper) so I can go like 9-10 cells deep, upgrade the grid and that would be perfect for me.
 
You don't need to upgrade the supply to your home or the wires in the wall to your socket! The average current your welder draws is very low, it's just the momentary inrsuh current that is the problem. Lots of amps but only fpr milliseconds. The normal type B breaker is intended for normal domestic appliances, the inrush to your welder looks like a short circuit to this breaker and it trips. Type C and D are designed exactly for difficult loads, like welders and will permit short-term current overload without tripping.

The breaker will cost about 10 dollars/euros/pounds and take 10 mins to fit, so worth a try!
 
i used a C and D type breaker. even the D popped at least 3 times a week.

after spending about 2000 euro's in total on AC type welders i came to the conclusion battery/capacitor based welders are WAY superior in almost every way.
 
999zip999 said:
Go find some A123" s with enough P and. S or LTO and get started. Yea LTO. Lipo No.

4 lifep4 blocks are pretty enough and will live forever and are mutch tougher then lots of separate A123 cells. 40Ah blocks is way enough and can deliver 2kA shots. my 60Ah blocks can do almost 3kA shots wich is WAY too much for that poor malectrics mosfet board but the newer version has much better diode protection and active cooling so it should last a while before the diodes start to degrade.

right now i have to choose between real hard but very short pulses that dont heat the electrodes up or use a old lead acid with way longer pulse times so the diodes live longer but the electrodes get very hot very fast.
 
A123 20ah pouches on my bike for 5 years and 20000 miles they're heavy and I bought a balancer for it but never needed to balance the cells Just Bounce them because I was interested remember 5 years 20000. miles and 1290 Cycles
. Heavy as two bags of s*** but working at 80 amp so after use you have to build a big battery and use half of it as of your demand.
 
Punx0r said:
You don't need to upgrade the supply to your home or the wires in the wall to your socket! The average current your welder draws is very low, it's just the momentary inrsuh current that is the problem. Lots of amps but only fpr milliseconds. The normal type B breaker is intended for normal domestic appliances, the inrush to your welder looks like a short circuit to this breaker and it trips. Type C and D are designed exactly for difficult loads, like welders and will permit short-term current overload without tripping.

The breaker will cost about 10 dollars/euros/pounds and take 10 mins to fit, so worth a try!

Thanks alot! will sure try that!
 
Based on Flippy's experience you probably want to go straight for the D-type, especially if you're paying an electrician to install it.

This is the difference in current each type permits over nominal current without immediately tripping:

B Curve: Trips above 3 times In up to 5 times In
C Curve: Trips above 5 times In up to 10 times In
D Curve: Trips above 10 times In up to 20 times In

http://www.quantumautomation.com/blog/c-vs-d-curve-circuit-breakers-whats-the-difference


It should just be a simple swap (B-type for D-type) without any wiring upgrades, but local electrical regulations vary around the world and so best to discuss with your electrician to be sure.
 
Wire length is also a big issur. I got about 15m from the utillity box to the welder and even the D would trip, mostly because of the inrush current but it would still trip when it got hot, becayse it will get hot under constant welding pulses. Once it was saturated with heat it would trip all the time.
 
Put a fan on it. :wink:
 
Honestly, if the breaker is getting hot then I would actually start worrying about the cable feeding the wall socket - depending on what size cable you have. If the cable rating is close to the breaker rating of 16A then I'd maybe slow down the welding a bit - get the average power consumed down.

Unfortunately conventions do vary: in the UK 2.5mm^2 cable is normal to feed a wall socket. That's good for 18A worst-case (27A clipped in free air). But we have the unusual practice of wiring sockets in a ring circuit, so the same 2.5mm^2 cable will suddenly be conservatively rated for 32A. Good for overloading a wall socket :lol:
 
the breaker gets hot because of the coil inside and the pulse load nature of welders. the cables dont get hot. because of the heating of the coil will trigger the bimetallic strip and pop the breaker constantly when it gets saturated.
 
Back
Top