I have always believed that it was ok to replace a capacitor with a slightly higher voltage rated one. eg 16v with 25v.Going higher than this would risk the capacitance value being out at the working voltage. However I just read the following quote from this pdf
Electrolytic capacitors are non-ferroelectric with a very low dielectric constant. Their capacitance is derived from a very high surface area and nanometer thick dielectric layers. Their capacitance is not a function of applied voltage.‐Design Solutions For DC Bias In Multilayer Ceramic Capacitors (August 2010 Electronic Engineering Times)
So does this rule not apply to aluminum electrolytics? Eg will I have no issues replacing an electrolytic 25v 100uf with a 50v 100uf assuming both are low esr, similar footprint and good quailty?
Answer
Long story short, you can replace an electrolytic cap with another one of the same value but higher voltage rating.
Electrolytic capacitors usually have a very lose tolerance to begin with, and with age and heat their capacitance drops a lot. They also change their characteristics with applied DC voltage, but this effect is a less drastic as in multi-layer ceramic types.
Given that you can't really trust the exact capacitance longer than a few weeks, it is very unlikely that someone designs in an electrolytic capacitor and expects the value to be exact. That's why you will often find them in non critical situations where a capacitor of some ballpark capacitance is needed but the exact value does not matter much.
Such applications are buffer capacitors and smoothing capacitors in power supplies. In cheap audio devices you'll sometimes find them as DC blocking capacitors as well.
In all of these cases the performance of the device will not change much if the real capacitance changes by ... lets say ... 50 percent. So a change to a different voltage-rated part won't hurt.
Using capacitors with a higher voltage rating as strictly required is btw a good thing because the capacitor will not age as quickly as one working on it's limit.
No comments:
Post a Comment