I have a simple circuit that takes 26 volts down to about 8 using 2 voltage regulators (7818 and 7808). Both are attached to (2) 10uf 35 volt electrolytic capacitors. Can I replace them with 10 uf 50 volt capacitors--both electrolytic. The voltage is the only different thing. Just simply put, would the higher voltage capacitor allow higher voltages, or is there other things I that require higher voltages. Hope that makes sense

The reason I am asking is one place I read that your capacitor should be 50% more than the circuit voltage (so the rated voltage should be about 39). Then another place says that the circuit voltage should be about 66% of the rated capacitor voltage (So the rated capacitor voltage should be about 43 volts).

 

That should be fine; the "35V" or "50V" rating is just the maximum voltage the cap can withstand without damage.

 

Thanks Tom66.

I know this is a vague question but I am just barely grasping it myself. How does Ripple current fit into this.

Simply put, I understand ripple current as the "unwanted junk waves"--my definition when AC is converted to DC and the capacitor "smooths" out these waves with a constant supply of "good" waves--again my definition. Does this matter when trying to pick the right capacitor?

Thanks again for the help.

 

Thanks Tom66.

I know this is a vague question but I am just barely grasping it myself. How does Ripple current fit into this.

Simply put, I understand ripple current as the "unwanted junk waves"--my definition when AC is converted to DC and the capacitor "smooths" out these waves with a constant supply of "good" waves--again my definition. Does this matter when trying to pick the right capacitor?

Thanks again for the help.

Ripple current is a function of voltage ripple and ESR in the low frequency domain (think ohms law), but it's a bit more tricky at higher frequencies.

For your uses, you will not need to worry about this. Low ESR caps or even general purpose cheap caps will be more than capable of doing the job.

 

Based on having done some MIL STD reliability calculations I always make my cap voltage spec at least twice what it sees. You get good reliability numbers that way. I seem to remember something about running caps well below their rating the effective capacitance changes, but this isn't an issue with filter caps.

Ripple current is the current rippling thru your capacitor. Each cycle (or half cycle) of the AC in causes the cap to charge, and then it discharges into your load. That means real current is going into and out of your cap.

Cap current is always given by:

I = C dV/dT which isn't the easiest thing to solve for the weird waveform on a rectifier cap. You can get a (low) approximation of it by pretending the ripple voltage is a sine and just compute Vc/Zc (V cap over cap AC impedance).

ESR (Equivalent Series Resistance) plays into how much power the cap is dissipating which plays into determining the max ripple current, but that's the manufacturer's responsibility.

It's very hard to drive a cap hard enough to violate the max ripple current spec.

 

I saw on a capacitor manufacturers site that making the filter capacitor higher voltage than necessary provides benefits up to 2x the applied voltage. (confirming Ernie)

I ran into the dC/Dv problem with oil filled run capacitors. Was trying to use motor caps for some metrology experiment and found them useless. Now ya' know which caps suffer miserably from voltage effects.

 

I saw on a capacitor manufacturers site that making the filter capacitor higher voltage than necessary provides benefits up to 2x the applied voltage. (confirming Ernie)

I ran into the dC/Dv problem with oil filled run capacitors. Was trying to use motor caps for some metrology experiment and found them useless. Now ya' know which caps suffer miserably from voltage effects.

Some ceramic caps are also bad with voltage. Y5V can lose up to 80% at applied voltage. X5R is better, only 30%, but it still needs to be considered.