Variation in usable capacitor ratings?

I would personally replace all of them as a kit especially if they are being used as timing or wave shaping. The higher capacitance rating when being used as a filter is going to remove more ripple.

If I were doing it I would replace all as a set especially if they are being used as wave shaping or timing caps. As a DC filter any capacitance increase will usually remove more riple.

In any combination you like.

It's just a *maximum* voltage.

Graham

On Thu, 19 Oct 2006 13:50:51 +0100, Art Deco Has Frothed:

Either or.

I always opt for a higher max voltage if available. The capacitive value depends on what the cap does. Electrolytics usually can be subbed by higher values maybe 20% higher without side effects.

OK, thanks.

So in the third example above, where a 330uF 16V cap is replaced with a 470uF, 35V cap, it wouldn't be advisable to exceed 35V for the replacement?

if there are switching supplies involved make the replacements are rated for switching service or low/very low ESR

These largish value caps are usually used for smoothing and decoupling, where values a bit bigger are a non issue. The voltage rating is the max V the cap is rated to work with, so again replacing with the next voltage up is a non issue.

NT

It's not going to exceed 35V is it ? The voltage must be under 16V or the original cap wouldn't have been suitable.

Graham

Why do you have this idea about 20% ?

Any voltage higher than that required is just fine, although it would be silly to take it to extremes.

Graham

They are a bit better these days, but the capacitance value tolerance on electrolytics used to be +80% / -20%. An increased voltage rating, within reason, is of little consequece, provided the part is still of a suitable physical size to fit where it has got to go. Going up too far on the voltage, can alter other characteristics of the cap, such as its ESR, which may or may not be a problem, depending on the cap's function. For caps used as filters in switch mode or linear power supplies, an increase in value is often beneficial to the filtering function, but if you go too far, then the initial charging current from cold, may prove too much for some prior component, such as a diode or regulator transistor. As Meaty-boy says, going up 20% on the nominal value will not present a problem, and the values supplied in the kit should be just fine.

Arfa

I meant the voltage rating of the replacement capacitor. Would it make a difference here if a 470uF 50V capacitor was used?

Thanks, everyone. I now have a better idea how this all works.

Almost certainly not, provided it fitted the board ok. But note that if it's much bigger than the original part, that might bring it closer to some item that runs hot, and this will shorten the life of the replacement by a considerable factor...

Arfa

In power supplies the capacitance values are usually not critical, so anything up to 3 times the value is probably okay, actually, a good thing.

The voltage rating can also be higher, up to 3 times higher is perfectly okay. If you go above that, there's a chance the capacitor will over time lose part of it's dielectric thickness, which will increase the capacitance, which is harmless in 98% of the cases.

On Fri, 20 Oct 2006 06:20:26 +0100, Art Deco put finger to keyboard and composed:

The only things I would be concerned about would be whether the higher voltage types had higher leakage, which *may* effect some timing applications, or whether the higher voltage types had higher ESR values, which would result in poorer performance in switchmode supply applications (as stated elsewhere in this thread).

- Franc Zabkar