DIY spot welder Transformer inrush current?

[cloudy]

I'm trying to build a spot welder using a transformer and the following "100A LCD" timing circuit (not purchased yet):
https://www.ebay.com.au/itm/3850659...15&osub=-1~1&crd=20220824073210&segname=13017

The transformer I have is not from a microwave, I think it came from the second hand store and was kindly donated to me by a friend. I think it's larger than a MOT, and only has one winding with a centre tap (called an Autotransformer).

I connected it to 240V ac mains supply and it seemed to work fine, the centre tap was exactly half way through the winding, and measured about 125V ac.


I added six turns of very light gauge wire as a test, and this produced 4.48V ac open circuit. From this, I think the transformer had about 321 turns on the primary to begin with. I've since removed about half the turns to make room for some large gauge wire for the welder. Now when I plug it in (no secondary yet), it trips my 16A circuit breaker "MCB" some of the time. I think there is randomness because inrush current varies depending on when in the cycle the transformer is energised.

Can anyone tell me some strategies to reduce the inrush current so I can energise the transformer reliably through a 16A MCB? I'm unsure if adding a secondary and shorting this through some nickel (welding) will further increase inrush current and tripping isssues. I'm guessing I'll only need <100ms welding time once the timing circuit installed.

Has removing some turns from the primary increased the inrush current? If that is the case I could add a few turns back on.

Another option is to do controlled switching and delay closing to minimise inrush, but this needs some programming and electronics I think.

Any help or suggestions are appreciated.

 

[amberwolf]

Less turns of wire means less resistance and less inductance, so it would increase the rate of change of current. (increase inrush current); this would be proportional to the decrease in wire length / number of turns, so half the turns means twice the current.


Depending on the continuous (and peak) current the system will draw, you may be able to use an ICL (inrush current limiter) which is basically a kind of resistor that starts out higher and then drops in resistance as it heats up from current flowing thru it. You'd need one that supports the worst-case current draw the system would use; it is installed in series with the AC input to the primary coil, and the fuse on that.

 

[cloudy]

Thanks AW! I'll try add turns back onto the primary winding in that case. What you say makes sense, just wasn't obvious to me at first. Thanks

 

[m3vuv]

throw that transfo in the bin,you need wire as thick as your fingers for a spot welder,you will never fit that on that as a secondary as there isnt room,i built one with an mot and a dual pulse thimer runneng a solid state relay,it never did work ok,in the end i brought a cheap chineese fet car battery welder,blew the fets on the first weld,modded it with a 47uf cap on the feed to the fet driver,also use a million microfarad 50v cap in paralell with the car battery,it works flawless,dont waste your time on a mot transformer welder,they either cook the cell or dont weld,been there done that!,mine also welds nickel strip fine to the stainless steel tank of my homemade ultrasonic cleaner for ground connections to solder onto,hope that helps paul m3vuv 73

 

[cloudy]

Too late! I've already removed a bunch of turns and added 18 strands of 1.5mm2 wire forming a 1.5 turn secondary winding.

2.47V open circuit and enough amps to melt pretty much anything. Ive ordered a control circuit that will limit pulse duration, I'm hoping to be able to control switching to reduce the inrush current and make the energy per weld less random. Even if it fails, it was still a fun project and has only cost me $AUD 30 so far.

 

[cloudy]

https://www.ebay.com.au/itm/373660950123

That's the controller I've ordered. Hard to know exactly what it can do from the description.

 

[m3vuv]

your living on cloud lala land if you think thats gonna work anything like as you hope,your wastint your time with that setup trust me!,been there done that but if you wont take good advice thats your issue,let us know how it goes or doesnt!,good luck.

 

[cloudy]

Finished the welder, seems to work for my purposes, I've mounted it in a metal toolbox. Total build cost: $35AUD
The timer circuit I used is:
https://www.ebay.com.au/itm/373660950123
This seems to work great, and allows controlled switching of the big MOSFET. It seems to reliably pick the zero point when energizing to minimise inrush current. Weld timing can be adjusted in 20 ms increments (1 cycle at 50Hz).

The Transformer I used was large and heavy, I think possibly bigger than a MOT. The secondary is as described in my earlier posts.

Keen to know how to use dual pulse welding correctly, does anyone have any suggestions for pulse duration and delay between pulses? I'm planning to do copper nickel sandwich using 0.1mm copper.




Copper electrode and some 1mm steel sheet metal welds. The button on the electrode triggers the timing circuit. The welds are strong, but there is copper spattering. I think I'll need to sharpen the electrode to weld cells.



Two pulses of 80 ms each (4 cycles) with 80 ms intermission. Does anyone know how to decide timing of dual pulses, including the gap?



The nickel strip I used is ripping when I pull it off, can anyone comment on weld quality?





Yes, it does the nail trick. Weld time for this was 1.88 seconds. I'll be wearing UV protection for my eyes and gloves if I do this again. Quite a lot of energy, and the galv coating caught fire.

 

[harrisonpatm]

I used the same controller for mine, doesn't work perfect, works just as well as I need it to, though. Made over 2000 welds for my emoto build, no sweat.

To find the setting that works, with your transformer and your voltage, you're just gonna need to spend an hour with scrap nickel, copper and dead batteries that you don't mind ruining. You'll get there. Just keep the weld times as small as possible. One possible recommendation online stated that you want your first weld to warm up the material and clear the surface of imperfections, and the second to push the amps through and do the actual weld. Don't take that as hard and fast spot welding advice though, I'm just one guy.

I used 05 setting for the time period in-between pulses.

 

[cloudy]

Can you say what your settings are for pulse duration Harrisonpatm? Just trying to establish a starting point before going at for trial and error. Reading about other posts, overdoing the weld time/energy seems to eventually punch holes in the metal strips and cell cladding, and underdone welds are weak and break in use from vibration. I think another issue may be consistency, in that minute change in electrode pressure or contamination on the strips may lead to large variation in weld quality. Is dual pulse welding meant to improve weld consistency? Hard to know how the three timing parameters we are discussing affect outcomes, and I don't think trial and error by itself will lead to the optimal answer without first understanding a bit more of what's going on.

Have you worked out what the C-1 and C-2 "power" setting does with this weld controller? The instructions say this is a power setting parameter. I don't really understand how switching of the mosfet could be used to modulate power, I thought they were either on or off. Above 50%, I'm getting some nuisance tripping of the 16A breaker in my 240V board, so it might just be affecting the closing time and inrush current. I don't think this does much for weld "power", but could be wrong here.

 

[harrisonpatm]

The C1 and C2 settings do indeed adjust output current, don't ask me how, I just know that they do. I think I currently have it at C1:80, C2:60, T1 and 2 both less than 10, interval @05. Welding .15mm nickel to 18650s and 32650s. But that's just my variables; yours uses a different transformer, and 240v vs my 120v. settings will be different. If high current trips your breaker, try decreasing weld time before you try decreasing current. Spot welding is about high current, high heat, really really short time.