Question on AC Capacitors used in place of DC Capacitors

[gel_fh]

Hi
I'm new here and I just would like to ask for help regarding our surge suppression system. Our surge supression system is designed to eliminate transient voltage spikes on the 3 phase AC lines to protect the SCR pucks on the scr system.

The surge supression system consist of a 3 phase bridge rectifier connected across the 600 Vac line, and on the output of the bridge rectifier is the 2 each DC capacitors, rated at 800 Vdc, 10 MFD, and a Resistor Bank.
As explained in the user manual, when a voltage transient occurs, that would normally appear as a spike on the ac waveform, it is absorbed by the capacitors as the spike attempts to raise the capacitor voltage. The small voltage increase over the initial voltage stored in the capacitor will then be 'leaked away' by the discharge Resistors.

The situation is this; The 2 DC capacitors have been replaced with AC capacitors rated 1,100 Vac, 10 MFD. and we still face the problems of replacing scr pucks as well as MOV's across them several times in a year.

My question is this: Do the AC capacitors used in place of DC capacitors serve their purpose the same? If they are, could you please suggest why we are still having pucks and movs failure?

 

[vk6kro]

It is important to know what the source of your mains spikes is.

A 10uF capacitor is a very small capacitance, really. At 60 Hz it has a reactance of 265 ohms (which is huge ) and a decent mains pulse wouldn't be bothered by that at all. Putting the capacitor in a circuit with a bridge rectifier would only raise the total impedance. You would have to absorb several hundred amps to make even a dent in a proper mains pulse.

Not that you would do this, but to take a current of 100 amps at 100 volts you would need a capacitor of 2652 uF at 60 Hz. This has a reactance of about 1 ohm. Just gives you an idea of where your 10 uF capacitor fits into the scale of things.

If the mains pulses are coming from your electrical supplier, you should be entitled to ask them to stop doing it. It may be happening when they switch lines. Talk to them. If you are getting sustained mains surges of 50% above normal (which it sounds like you might be) , nothing will filter this out except a high speed cutout.

 

[gel_fh]

Dear vk6kro,
Thanks a lot for this professional advice. you are right to say that the 10uF capacitors are very small. There are other 3 each 1,500 uF connected in series and in parallel to the these 10 uF capacitors. My apology for failing to mention this in my inquiry. The mains pulses come from the ac source whenever they open a circuit breakers of fuses.
Someone answered the question as to whether the ac capacitor can be used on the dc source and said that it serve the same purpose. The ac capacitor takes on the dc voltage and get charged. And that the surge supression will still do its work of eliminating the spikes. There are other reasons for scr pucks and movs getting damaged.

 

[vk6kro]

These AC capacitor setups are a bit of a worry.

Suppose your 800 v AC supply suddenly doubles to 1600 V.
The peak value will go from 1131 V to 2262 V and the capacitor will charge up to this as well.
There will be a surge in current which may pull the voltage down for a part of the cycle, but eventually, the capacitors will charge up to the peak voltage of the input.

But what then?

The diodes in the bridge rectifier are now reverse biased, so shortly after the peak voltage has passed there is now no load on the supply and it rises to something just under 2262 volts, maybe 2000 V.
This is still a lot more than the 1131 it usually gets and you would start getting SCRs failing.

I know you said there are bleeder resistors across the capacitors, but they would work over seconds, not milliseconds.

So, this scheme might work for extremely short pulses (provided you are using Schottky diodes for the bridge rectifier) but any pulse lasting more than a few cycles will pass unaffected.