23
\$\begingroup\$

I want to know if the polarized capacitor has some advantage that they are used in some circuits?

For example, in a schematic of the BISS001 PIR controller IC, in some places, a polarized capacitor is used and in some places a non-polarized capacitor one.

Can I use a non-polarized capacitor with the same voltage and capacitance instead of these polarizing capacitors?

Reference Docs:

  1. BISS001 datasheet
  2. HC-SR501 PIR MOTION DETECTOR datasheet
  3. Grove - PIR Motion Sensor or EasyEDA link

What I've understand from your answers is why the electrolytic capacitors are used, and why these are polarized.

But the designers of this circuit could have used a non-polarized capacitor or even polarized tantalum capacitors. Is it true? As the (Grove - PIR Motion Sensor) module uses polarized tantalum capacitors.

I want to know if the polarized capacitors are being used for circuit protection or is there any other reason (regardless of the type of capacitor.)

Is there a problem if these capacitors are replaced with non-polarized capacitors in these circuits?

\$\endgroup\$
  • 17
    \$\begingroup\$ It's not because we want them polarized, but them being polarized is a consequence of how they'r emade. \$\endgroup\$ – Hearth Apr 7 at 21:48
  • 2
    \$\begingroup\$ Compare the prices of electrolytic caps with ceramic (or film) caps in the range of 10uF ~22uF (after taking bias voltage derating into consideration) and you will have one of the reasons. \$\endgroup\$ – Wesley Lee Apr 7 at 21:54
  • \$\begingroup\$ The very thin oxide layers, providing insulation for a certain polarity, provides a very compact energy storage module. Just view those useful electrolytic capacitors as early versions of self-assembled-nano-tech; the manufacturing process creates the very thin oxide layer: the oxide is GROWN or FORMED. \$\endgroup\$ – analogsystemsrf Apr 7 at 22:01
  • 4
    \$\begingroup\$ Is there a problem if these capacitors are replaced with non-polarized capacitors in this circuits? -- no, there is no problem. But you may find it hard to find non-polarized capacitors of the same specification. Non-polarized capacitors are usually in the nF range and are usually rated for low voltage use (5V or less). If you want anything in the uF or mF and/or rated at tens of volts you will have to make do with polarized capacitors \$\endgroup\$ – slebetman Apr 8 at 5:22
  • 4
    \$\begingroup\$ @slebetman "Non-polarized capacitors are usually rated for low voltage use (5V or less)" - the reality is actually opposite of what you say. \$\endgroup\$ – Dmitry Grigoryev Apr 8 at 8:03
19
\$\begingroup\$

Can I use a non-polarized capacitor with the same voltage and capacitance instead of these polarizing capacitors?

Electrically speaking, non-polarized capacitor is always better than a polarized one. Yes, you can always replace with a non-polarized capacitor with exactly same rating.

But there is an assumption hidden here: Provided you can find one that's physically small enough to fit on your board and cheap enough to fit in your budget. And the fact that you can't is the only reason we use polarized caps.

I assume that, if we ever learn to make non-polarized caps that are as cheap and dense (capacity-per-volume) as electrolytic ones, the polarized capacitors will vanish.

Side note - voltage and capacitance are not the only electrical parameters of a capacitor. They would suffice for an ideal capacitor, but real world brings other, ugly metrics. Like ESR, capacity coefficient with temperature or voltage, frequency response, etc. Circuits designed around quirks of particular tech can fail if the substitute differs there. Even being too good can cause trouble, eg. high-ESR caps naturally keep peak current in check so substituting with a theoretically superior low-ESR part can cause the whole thing to blow up. Adding ESR is trivial - but that's no longer a drop-in replacement, but rather a circuit redesign. So we don't replace electrolytics with something else not because polarization is important, it's just a nuisance. We keep them because of many other parameters, less obvious than C, V and polarization.

\$\endgroup\$
  • 2
    \$\begingroup\$ One thing that should also be noted is that in applications where the frequency response of the capacitor is important, you can't just replace one type with another, since they work well at different frequency ranges. Examples are e.g. noise filters. \$\endgroup\$ – nshct Apr 8 at 21:22
  • \$\begingroup\$ @nshct that's what I've meant by "other, ugly metrics" : ) Edited \$\endgroup\$ – Agent_L Apr 8 at 21:25
  • \$\begingroup\$ Suggestion: add to your answer a link about ESR : avnet.com/wps/portal/abacus/resources/engineers-insight/article/… \$\endgroup\$ – Overmind Apr 10 at 5:00
37
\$\begingroup\$

The physical size of a capacitor is a function of the thickness of the dielectric (among other things).

Early on, it was discovered that the oxides of certain metals (aluminum and tantalum in particular) made good dielectrics, and could be made very thin through a chemical process — orders of magnitude thinner than other dielectrics such as waxed/oiled paper and plastic film. Therefore, the electrolytic capacitor was invented to provide high capacitance in a reasonable volume.

Unfortunately, the chemical process requires that the voltage across the capacitor must have only a single polarity, so these capacitors are "polarized". Reversing the polarity degrades and eventually destroys the oxide layer. It's something we just have to live with in order to take advantage of this technology.


The ability to produce high-value capacitors in nonpolarized technologies such as multilayer ceramic means that it is now possible to use them where only a polarized capacitor would have been previously available. There is generally no problem with making this substitution, although you may need to consider some of the quirks of the technology you're switching to.

For example, some high-K (high dielectric constant) ceramics exhibit significant capacitance changes with voltage. This might be acceptable in a coupling or bypass application, but completely unacceptable in a filter design.

\$\endgroup\$
  • 1
    \$\begingroup\$ There is a specific use case where replacing tantalum devices with MLCC parts can have unfortunate effects: at the output of some (older) LDO regulators which require a certain minimum output ESR. It can be done, but some care is necessary. \$\endgroup\$ – Peter Smith Apr 8 at 9:03
  • \$\begingroup\$ Very nice answer! I'm trying to figure out if asking "Why does the chemical process used to make capacitor dielectrics from oxides of certain metals (aluminum and tantalum in particular) require that the voltage across the capacitor must have only a single polarity?" but I'm not confident it will be on topic. Is there any chance you could add a hint here? \$\endgroup\$ – uhoh Apr 8 at 10:28
8
\$\begingroup\$

Since you mention protection I'll add that polarized caps should not be used for reverse polarity protection. They will react on a reverse voltage very slowly (seconds or minutes), while typical sensitive components which are worth protecting will be dead within milliseconds. And once a polarized cap starts to absorb the reverse voltage, it may vent, explode or catch fire which (apart from the obvious problem with smoke and fire) can make it non-conductive again, exposing your circuit to the reverse voltage once more.

\$\endgroup\$
  • \$\begingroup\$ Good point here. \$\endgroup\$ – Overmind Apr 10 at 5:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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