Capacitive discharge spot welder help

[veloman]

I got mine to work, following these plans: http://ledhacks.com/power/battery_tab_welder.htm

The problem is, the welds are tiny and pull off way too easily. Could it be that my capacitor is not enough (real) farads? Maybe my power supply needs to be more like 15v instead of 12.36v?

Or perhaps my thyristor needs a bit more voltage to open the gate properly? I was using a tiny AAA battery, which had voltage dropped to .88v by the time it gets to the thyristor. I broke my first thyristor with too much voltage on the gate, so I'm trying to be careful with this one. This one is 70amps and a lot bigger (physically) than my first one which was listed as 100amps.

I'm using 8g wire for the probe leads, that shouldn't be an issue. I do see the voltage on the capacitor drop to 1v immediate upon activation.

 

[knighty]

sounds to me like you need more capacitors

as long as it's a good sharp/quick weld when you're using it... then the voltage/thyristor should be fine

 

[bigmoose]

It's about energy of the weld. 1/2 C V^2 A little increase in voltage, or more in capacitance. Be sure to read texaspyro's thread(s) as many of the 1 farad caps on the market are actually much smaller with concrete in the case to seem heavy. The rockford caps are up to par. I have two and have measured them.

 

[veloman]

I bought a used Rockford Fosgate 50 punch capacitor yesterday. $15. 1/2 farad. Swapping out the Power Acoustik 1.5f, it did much better spot welds. Putting both together in parallel gave me great spot welds - pulling the tab off, the welds themselves stayed. That's using 16.6v

Then I started to put my project battery together. Blew two holes through the tabbing (tab on tab). Disconnected the power supply, and blew two more holes through the tabbing on my practice cells.

Now, I'm letting the voltage drop to around 12v. But I think it will still be too much power. I don't know what changed. But it's got way too much power now.

Will use the Rockford cap alone next, at 12v.

I'm glad it's working, but damn.

 

[bigmoose]

What is likely is that the Fosgate's real capacitance is near what is on the can... the other is bogus.

 

[veloman]

But I shouldn't be blowing holes through my tabs with 1/2 farad. The guy who did the tutorial used a 2.5f.

But, then again, there is no other explanation. I'll find the sweet spot somewhere, maybe it's 8v with the 1/2 farad.

 

[bigmoose]

Right, what "is" well "is" ... as in reality. Perhaps the other guys 2.5 farad cap was better?

I would accept your behavior and tailor the voltage to what suits your needs. Welding is about Watt-seconds (joules). So accept your farads and dither the voltage to get the number of joules that makes your welds satisfactory.

For info I bought 3 fosgate 1 farad caps and they measured 1 farad +- about 8% Close enough for me. I cut another big can (size of an Ausie Fosters beer can) apart and ruined my bandsaw blade because it was a hunk of cement with a cap inside the size of two 35mm film rolls.

 

[speedmd]

Try a slightly larger area of welding. Slightly wider point. Contact pressure and time important also. If it arcs a bit too much it will burn away tab material and not make good contact between the two parts. A small piece of rubber tubing extending past the probe end will force the parts snugly together before arc hits. Popsicle stick if nothing else is avail. Short fat well connected leads are key for welding consistency. Looks like your real close.

best wishes

 

[veloman]

With more welds today, I am beginning to learn how it reacts. Voltage and farads don't actually matter as much, compared to the tabbing thickness and distance between probes. If I weld two tabs together, so that the probes are pointing *at each other* at the same location - this usually leads to blowing a hole through the tabbing.

I just wish it was more consistent and not as varied. Surface cleanliness is also affecting the connection too. These used cells with sticky-ness on the tabs from foam insulation I pulled off - is hurting the connection. With used tabs, they don't always lay on each other (or the cell) perfectly flat, affecting connection.

I also am finding that with the Rockford cap and 16.6v, and the thick tab, putting one probe on the cell itself, gives a better connection for the other probe to weld through that tab.

Thanks for the advice.

It seems the duty cycle for the thyristor is 10% or so without getting it too hot.