3
\$\begingroup\$

Suppose I need 1 μF capacitors. I am expecting a voltage of 5 V. Now 1 μF is available in 25 V and 700 V too. Does it make any difference choosing one over other?

\$\endgroup\$
5
  • 3
    \$\begingroup\$ Package size would differ quite a bit. Also most likely different capacitor types. \$\endgroup\$
    – Velvet
    Commented Mar 1, 2023 at 7:19
  • \$\begingroup\$ No, both are electrolytic. So other than package size, no downside? \$\endgroup\$ Commented Mar 1, 2023 at 7:34
  • 2
    \$\begingroup\$ There is parasitic elements, temp range, life expectancy, and there is purchase resistance. (Total cost may include cost to have same value, different voltage items in production.) \$\endgroup\$
    – greybeard
    Commented Mar 1, 2023 at 8:23
  • \$\begingroup\$ Many decades ago, I read somewhere that electrolytics should not be operated below 20% of their voltage rating. I can't remember where I read it or what it exactly said but, in the absence of any other guidance in the pre-internet decade, it was a guideline I used. But I didn't sacrifice anything to stick to it. \$\endgroup\$
    – TonyM
    Commented Mar 1, 2023 at 13:41
  • 1
    \$\begingroup\$ Rather than voltage rating, I would consider what type of capacitor you need. For small values like 1uF and below, frequently an MLCC is a better choice due to lower ESR and better high frequency performance. Electrolytics have advantages too, but 1uF is on the smaller end of when you see them used. \$\endgroup\$ Commented Mar 1, 2023 at 16:13

3 Answers 3

6
\$\begingroup\$

Suppose I need 1 μF capacitors. I am expecting a [working] voltage of 5 V. Now 1 μF is available in 25 V and 700 V too. Does it make any difference choosing one over other?

Yes. For electrolytics, don't choose a voltage too far above the maximum expected working voltage. As the electrolytic's working voltage rises, so does the ESR, assuming that the capacitance is the same. So, for a 5V application, a 1uF/700V capacitor performs much worse than a 1uF/16V capacitor.

Besides, that 1uF/700V capacitor will cost quite a bit. "Normal" electrolytic voltage range tops out at 450V. Anything above that is considered specialty.

\$\endgroup\$
4
\$\begingroup\$

Many decades ago, I read somewhere that electrolytics should typically see a working voltage of at least 20% of their voltage rating.

I can't remember where I read it or what it exactly said but, in the absence of any other guidance in the pre-internet decade, it was a guideline I used and use.

The 20% isn't a hard-and-fast limit and I prefer to keep closer to the max. operating voltage, allowing for supply tolerances and a margin. For example, to decouple a board's 12 V supply I'll use a 16 V capacitor but avoid using a 35 V part. I'll avoid using a 16 V capacitor across a 5 V rail. Unless there's some significant benefit, such as part already used so a simpler BOM, easier stocking and potentially lower cost.

Obviously, this can't be applied for things like coupling capacitors, use in variable output PSUs. But it was a guideline, not a rule. I didn't sacrifice anything in the circuit to meet it, give up a notably cheaper part that broke the rule or anything.

I can't recall the stated consequences of not doing it. But I think it was about reduced life of the part.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Seems reasonable. It could be due to the increased ESR rating Kuba's referring to. \$\endgroup\$
    – Velvet
    Commented Mar 1, 2023 at 14:12
  • \$\begingroup\$ Then what's the deal with electrolytic capacitors in symmetrically powered audio circuits? Being operated at zero offset, they certainly seem underpowered, and apart from passive crossovers, nobody uses bipolar variants for that purpose. \$\endgroup\$
    – user107063
    Commented Mar 1, 2023 at 15:31
  • \$\begingroup\$ @user107063, I actually wrote then crossed out "if anyone's thinking of starting a long debate about how this applies to AC coupling circuits, use in variable PSU outputs and all sorts of things, please do re-read the part saying I can't explain the origin or reasons". However, I'll reword it and expand it a little. \$\endgroup\$
    – TonyM
    Commented Mar 1, 2023 at 15:47
  • \$\begingroup\$ The dielectric in electrolytic capacitors degrades, if I remember correctly. It can be reformed by applying voltage to the capacitor. Applying voltage to the capacitor anodizes the plate, which creates the correct insulating layer inside the capacitor. Chemistry! \$\endgroup\$ Commented Mar 1, 2023 at 19:18
  • \$\begingroup\$ @DietrichEpp Yes, that's rather important for power capacitors, like those used in photographic flash guns. Without occasional exposure to significant voltage (typical instructions recommend 15 minutes every 3 months), the isolation layer shrinks, voltage tolerance goes down and capacitance goes up. Capacitors not optimised for energy storage tend to be less particular about reforming conditions and occasional short-time exposure to reverse or overvoltage. \$\endgroup\$
    – user107063
    Commented Mar 2, 2023 at 1:48
2
\$\begingroup\$

Maybe you can provide more information on the application and the capacitor type (ceramic, electrolytic, ...).

But if you expect 5V, a capacitor rated at 25V will be good enough most of the time. The downside of higher rated capacitors is usually their increased physical size which takes away board space and adds inductance. So - besides cost - there might be reasons not to choose the highest rated capacitor available.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Both are electrolytic \$\endgroup\$ Commented Mar 1, 2023 at 7:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.