Dangers of filtering cap (rectified mains)

[superwolf]

I have a question regarding taking mains AC voltage, straight through a bridge rectifier to a large capacitor. Is there any inherent danger or disadvantage to this over using a NTC resistor in series to prevent the startup current spike?

The reason I ask, is I have a cheap chinese TIG welder that does just that... It creates a huge spark when I first plug it in. I took it apart and it has a couple 220uF caps with no current limiter.

Do I have to worry about the caps exploding? Shortened life? Just curious... Caps are cheap but removing shrapnel from my face isn't...

 

[Mikebits]

It is pretty common to have caps after the bridge. Are you plugging the thing in with ON/OFF switch in the ON position?

 

[superwolf]

The on/off switch controls the low level logic, but the power supply portion is always connected to the plug.

 

[Mikebits]

Really! Well there's your problem... Maybe you could add a ON/OFF switch for the AC. The surge will be the same but the switch might not spark so much.

 

[Mikebits]

Have you heard any reports of exploding caps with your model TIG welder? If not I would not worry too much.
Maybe google it and see.

 

[tonigau]

I think the power level or cost$? is to much for NTC inrush current limiting.

Small inverter welders I have been inside have a relay that is contacts open initially with small resistance in parallel after the bridge rectifier. A signal from the control board energises the relay ready for weld current.
This works well because the relay contacts are not switching any current, just the contact resistance during normal operation.

 

[Noggin]

Really! Well there's your problem... Maybe you could add a ON/OFF switch for the AC. The surge will be the same but the switch might not spark so much.

If you go this route, make sure you use a switch rated for the current. I've seen people use just any switch in situations like this and weld the contact closed.

 

[tonigau]

I dug up one of the cuircuits of a welder (that I have yet to get going) & attached an extract of the inrush current protection circuit.

Notice the relay type is a HC4 which is only 5A rated (switched load) & the welder could draw in excess of 10A & 240V.

Normally paralleling relay contacts can be problematic due to the minute time difference in contact opening & the breaking/making current would be left to the last contact to open/ first contact to close.
In this application though this relay can easily handle the unswitched primary current of this 150A Inverter welder.

In this circuit the 100R charges up the filter CAP bank & when the control board is awake it closes the relay contacts.

This probably wouldnt be too hard to add as a mod with the relay coil energised from the control board DC supply.

 

[superwolf]

Thanks everyone. But so far nobody has answered my question. Is it dangerous to charge capacitors so fast that they're just limited by their internal resistance? (0.6 or 0.2 ohm 20hz/20khz ESR my best guess given similar cap specs)

Since an uncharged cap is basically a short, and mains is 170VDC, you're looking at a peak of at least 100 amps, maybe upwards of 800 amps but I doubt that high..

Granted, the power company and FCC probably don't like it, but from a reliability or safety standpoint of the caps themselves...?

 

[Ubergeek63]

It is hard on the caps... If you unplug it every time you could put a power resistor across the switch. Quality HP stuff often puts a resistor in line and then shorts it out after the filter caps are charged.

 

[tonigau]

The inrush current will also be limited by any other resistances like: line filter choke, power cord, plug contacts, house wiring, circuit breaker...
The brief peak current will be less due to small inductance of wiring & mains filter, so all this adds up to probably ok.
& there is also the timing of switch on relative to the AC waveform & switch contact bounce.

My guess is if you did a bench test, had a low impedance 350V supply(not the mains supplied AC) & 4" of wiring I think something would go bang.
Under the right conditions It could be dangerous for the capacitors health in the long run, Electrolytic capacitiors should have a vent to expell any pressure, but if this happens the cap won't be too healthy.

Manufacturers implement inrush current limiting/protection for a good reason: To minimise unwanted stress on switches, connector contacts & of course the capacitors & to avoid blowing fast fuses & who knows if there is a fault in the Vent system of the capacitor it could blow up worst case.

Having said that If one was to try the same thing eg: 800uF & a higher voltage 415 --- 600V rectified mains then I think it could be quite likely.
Maybe you could ask bill the arc starter, he likes to blow things up

I blew the whole tip off my scope probe once, I think it was 40,000UF @ 60V, the biggest danger was me falling off my stool, it scared the sht outa me.


Tonig.au