I read that but ac passes through the capacitor.

Alternating current passes through the capacitor, but you asked about voltage ratings.

 

I am watching this thread with lots of satisfaction: 81 posts and all have been helpful and positive.
Additionally, discrete circuits are being discussed.

 

Thanks to all!
When considering the voltages for caps do we look at the dc volts only or the total volts (ac + dc)?
When considering the wattage rating for resistors do we look at total volts and total current or just dc in order to determine wattage?
RS

Hi again,

To answer this in the most general way, when you decide what the voltage rating for a capacitor will be it is based on the highest voltage the capacitor will EVER see. That may or may not include an AC signal riding on a constant DC level. You have to figure that out from the schematic or measurements in the circuit.
The voltage should be around 50 percent higher than the highest voltage as a rule of thumb. Caps that are rated higher than the highest voltage will last longer overall. Caps rated too close to the max voltage will tend to die sooner. To give you an example, in a fuel pump relay controller board I worked on not too long ago I found a 16 volt electrolytic capacitor on the PC board that was blown. It was a 12 volt system, but the voltage goes up higher when the battery is charging, such as 14v or maybe a little higher. Using a 16 volt cap there is not a good idea and those caps are known to fail. If it was a 25 volt cap it would have lasted longer.

The wattage of a resistor should be based on the average power being dissipated by the resistor. In a DC circuit, the average power is easy to calculate: the voltage across it times the current through it. If the DC voltage fluctuates slowly you can take the max voltage times the max current.
If the resistor sees AC or AC and DC at the same time, then you need to find the average power using the integral form:
Pavg=integrate(v*i,t,0,Tp)/Tp
where 'v' and 'i' are functions of time 't' as v(t) and i(t), and Tp is the duration of the period where the waveform is periodic.
This is easy to calculate if you know v(t) and i(t) in advance, but if you don't then you would have to measure that. You can usually figure out what those two are from a good circuit analysis method, or maybe use a free simulator like LT Spice which many people use here and so can help you with this also. You will also notice that in that integral "v*i" can be calculated in other ways such as R*i^2 or v^2/R, which would mean only having to measure one value either 'i' or 'v'.
Once you know the wattage, at least double it so you don't end up with a frying pan in the middle of your circuit somewhere. If you do know the wattage and you have a prospective resistor in mind, you can test it with a DC power supply by applying a voltage that results in the same power as it will see in the circuit, then see how hot it gets. If it gets too hot, increase the power rating and get another resistor.
There is a trick you can use if you have a lot of the same value resistors and the power rating is too low using just one. If you put two in parallel and then use two of those sets in series, you end up with the same value resistor but 4 times the power rating. It may take up more space however than a single resistor with the right power rating. It will distribute the heating power better though.