[N4NV]

I know how to figure the value for capacitors in series and parallel, but my question is: When you put capacitors in series, does the voltage rating for the circuit increase? i.e. if you have three 100V rated capacitors in series, can they safely run at 300V?

Thanks

Vince

[lvittori]

Yes but a good design will use current balancing resistors to compensate for unequal DC leakage currents.

[kreutz]

I know how to figure the value for capacitors in series and parallel, but my question is: When you put capacitors in series, does the voltage rating for the circuit increase? i.e. if you have three 100V rated capacitors in series, can they safely run at 300V?

Thanks

Vince

Yes and No.

Yes, theoretically the voltage rating increases. In the practice, the capacitors will not have exactly the same capacitance and leakage, so the voltage distribution will not be the same. In order to equally distribute the voltage across the series connected chain of capacitors, a resistor is placed across (in parallel with) each capacitor (it is called a voltage equalizer network).

Resistors used on the equalizing network must have fairly equal resistance, and their resistance should not change appreciably during aging. The best resistors are Metallic film.

Additionally: A 100V rated capacitor is not supposed to run at 100V. It is supposed to run at a maximum 75-80 volts (it is called De-rating)

[epineh]

While you guys are talking cap's, I have wondered if it would be possible to use the big cap's the car audio guys use (1 farad and up) in series, and use the balancing resistors as mentioned ?

I am sure I have seen some rated at 24V, my thinking was to use two in series running 24V servo's.

Two reasons for this, they are cheap for the capacitance and they would look pretty cool in an electronics enclosure

Russell.

[kreutz]

While you guys are talking cap's, I have wondered if it would be possible to use the big cap's the car audio guys use (1 farad and up) in series, and use the balancing resistors as mentioned ?

I am sure I have seen some rated at 24V, my thinking was to use two in series running 24V servo's.

Two reasons for this, they are cheap for the capacitance and they would look pretty cool in an electronics enclosure

Russell.

You can probably usen them in series, the problem would be the initial capacitor charge, so that the sudden power up of 0.5F capacitor does not blow your transformer and bridge rectifier. You will require to limit the charge current for some time during power up. Also the resistors will have to be power resistors in the low Kohm range with high power dissipation.

[epineh]

I was thinking of using bleed up resistors and timer of some sort, and using the paralleling resistors as the bleed down resistors.

Apart from that, would they be usable and not introduce any other nasty side effects into the system ?

Russell.

[kreutz]

I was thinking of using bleed up resistors and timer of some sort, and using the paralleling resistors as the bleed down resistors.

Apart from that, would they be usable and not introduce any other nasty side effects into the system ?

Russell.

I suggest to control the capacitor power on charge time by monitoring the capacitor voltage instead of using a fixed time. Charging a 0.5F capacitor with limited current will take a while. You can use the equalization capacitors as permanent bleeders too. No side effects should be noticeable. Make sure you add a couple of 1 uF 200v film capacitors in parallel in order to filter high frequency ripple. You might also need to add high ripple frequency rated capacitors in parallel depending on the ESR of your 1F Caps at 20 to 50Khz.

Kreutz.

[epineh]

Sounds like a plan, thanks Kreutz !

Russell.