Bridge rectifier with capacitor and choke filter

You could accept a bit longer on the drain down time- say 100 ohms, at about 50 watts. That would drain it in about 3-5 seconds
and wouldn't get as hot (it will still get pretty warm)
mark
 
I'd recommend adding a resistor across the capacitor, to drain the charge away when the welder is turned off. Otherwise the capacitor will remain charged, which could cause an unpleasant surprise for the unwary. Voltage on the capacitor will decay at a rate determined by the time constant, R*C. For a 50,000uf capacitor and a 1-second time constant you will need a resistance of 1/.05 (50,000uf = .05 farad), or 20 ohms. It will have to have a high enough wattage rating so it doesn't burn up. I went with a 50 ohm resistor when I modded my HF welder for DC -- high power 50 ohm resistors are fairly common. Stores selling stuff to ham radio enthusiasts would be a good place to look.

Homebrew, good suggestion, as I have ruined a few screwdrivers over the years shorting CAPs. Let me pose a question on this: Since this is TIG welding, if I use the solenoid and just let off the momentary switch, wouldn't the arc drain the CAPs until it went out? This would eliminate the need to "LIFT" the torch to extinguish the arc...…...
 
You could accept a bit longer on the drain down time- say 100 ohms, at about 50 watts. That would drain it in about 3-5 seconds
and wouldn't get as hot (it will still get pretty warm)
mark

Yep. The exact time to drain the cap isn't critical, as long as it happens in a few seconds. If you wanted to get fancy you could connect an LED (with the proper series R) to indicate when the cap is discharged.
 
The power dissapated by the resistor is V^2/R so the smaller the resistor, the higher the power dissipation. On the other hand, the voltage after time t is equal to Vo * 2.73^(-t/R*C) or .37*Vo after t = R*C and .14*Vo after t =2*R*C. If the voltage is removed from the torch on completion of the weld, there should be no exposed high voltage present. If you are doing repairs or immanence and cracking open the case, you presumably are aware the high voltage may be present (if not, put a warning label) and you can either wait for the capacitors to discharge or discharge them yourself., (Whenever I worked on high voltage power supplies, I took the precaution of discharging any caps before sticking my fingers or other sensitive items inside. After getting bit by a few charged CRT's, even an old dog can learn.)
If the assumption of no unintended exposed high voltage is valid, the discharge can be much longer; say on the order of a minute or more, the design intent being not to leave a charged capacitor for some unsuspecting individual to play with. The plus of using a higher resistance is a much lower power resistor is required. A 1K resistor coulld get by with a 7 watt power rating for an OCV of 75 volts. You are also not dumping welder power into heating a resistor.
 
Couple things I see right off.
Biggest is a full wave rectifier off a 80 or so volt AC current transformer.
Current transformers attepmt to maintain a specific current output by allowing the voltage to change.
So in short the OCV (open circuit voltage) could get up beyond 80 volts RMS from the transformer.
Second issue, running 80 volts OCV into a rectifier and a capacitor bank is NOT going to have 80 volts DC output.
Transformers and most other AC voltage numbers are in RMS (root mean squared) or average output. THe other number to be aware of is the Peak to Peak voltage. this is going to be the RMS voltage devided by .707 roughly. That works out to about 113 volts. Now that is going to be a actual DC output once the AC off the transformer is fed to the rectifier and the resulting DC is filtered with the capacitors.
Be aware that the OCV will exceed 100 volts if you have 80 volts RMS going into a rectifier. That level of voltage is MORE likely to shock the hell out of you.

Selecting capacitors is critical if you are going to run this in a scratch start configuration.
Reason being is surge current off the capacitors. They will dump tons of current into the electrode as you try to scratch start to weld. Make sure that you put in the correct inductor AFTER both capacitors to help keep this in check.

Now on a side note, as far as power output regulation. You can build this with 4 SCR's in place of the diodes in the rectifier and vary the gate voltage to them causing them to conduct for less than the full 180 degrees of the sine wave input. This is similar to how a light dimmer works.
 
Thanks to all who have posted. The info and suggestions are great. I will either swap out the 50,000 mf 70VDC CAPs for 35,000 mf 120VDC CAPs to handle the OCV voltage OR I will put the 2 50,000 MF CAPs in series. Either way, it is a drop in capacitance, but hopefully won't make that big of a difference in the arc. As I have previously stated, this is more of a fun prototype project than anything else.
 
Last edited:
I never met a capacitor that I didn't enjoy blowing up on the bench. I say go for it and see what happens. Its a win-win situation. :)

Takes me back to my days at TAFE charge a LARGE cap and toss it to someone.

I've been repairing electronics for nearly 40 years and never heard of reforming caps, head in sand maybe, or just replaced with OEM as military and aviation we were not allowed to buy anything but from the manufacturer.
 
Back
Top