3
\$\begingroup\$

It is obvious to me that decoupling capacitors are necessary in proper circuit design. I also realize that different types of capacitors can be more useful depending on the application. However, I have also read in several places (most recently in Op Amps: Design, Application, and Troubleshooting by David L. Terrell) that decoupling capacitors can completely be negated in their purpose if aluminum elecrolytic capacitors are used due to their higher impedance or the leads are not short enough. This ties directly into my question:

Why do I find aluminum electrolytic capacitors being used in decoupling techniques? Is it bad practice to use them or are these few sources I have read from incorrect in their explanations?

\$\endgroup\$
  • \$\begingroup\$ You have to consider cost. For small capacitors, ceramic chip types are very affordable. So for 0.1uF caps placed at VCC pins of IC's, ceramic is the way to go. It would be foolish to use electrolytic in that application. But when you start getting to 22uF or 47uF (which would be more bulk decoupling), ceramics are expensive. So you need to go to tantalum or electrolytic. Tantalum has had some supply issues in the past, and there are ethical concerns about how tantalum is mined. So some people don't like to use it. \$\endgroup\$ – mkeith Nov 20 '15 at 20:20
  • \$\begingroup\$ In the context that I have read these references in they are in no way concerned with economy. They are describing aluminum electrolytic capacitors as an inherently bad choice due to their high impedances. But I do understand what you are saying. \$\endgroup\$ – user92289 Nov 20 '15 at 20:28
  • \$\begingroup\$ In high volume production, cost is very important. If you are seeing electrolytic caps in high volume production boards, cost is the likely answer. You also need to understand that there is a difference between "bulk" capacitance that is added to boards to stabilize rails against low-frequency variations and bypass caps which are paced close to power pins to help reduce voltage fluctuations caused by high current spikes in digital switching circuitry. For the former, electrolytic is OK. For the latter, it is a bad choice. \$\endgroup\$ – mkeith Nov 20 '15 at 21:14
  • \$\begingroup\$ For computer circuits (processor and GPU) some very very low ESR polymer eletrolythics are used. They provide good density for such high current applications, and are often used together with ceramic ones. \$\endgroup\$ – Wesley Lee Nov 21 '15 at 0:27
2
\$\begingroup\$

In addition to their capacitance, all capacitors have additional 'stray' components that may have to be taken into account if the circuit is sensitive to them. These include leakage, equivalent series resistance ESR, series inductance, and dielectric relaxation, not to mention the ability to degrade with temperature and/or age.

Different technologies of capacitor will have different strengths and weaknesses. For ESR, alli electrolytic is highest, followed by tantalum, with plastic and ceramic the best. The inductance generally follows the size and lead length. For leakage, alli electrolytic is worst, though usually not a problem in decoupling applications, you just avoid using them for long time delays in a 555 circuit.

For decoupling your circuits, it is important to use appropriate sized capacitors in the right places.

Most op-amp circuits will be working with such low frequency components that almost any capacitor can be used for supply decoupling, as long as the leads are kept short, alli electrolytic will be fine.

Very fast op-amps, for example video capable, as well as digital circuits, will need small value, low inductance, short lead decoupling capacitors very close to the ICs. In addition, the board will also need bulk decoupling, an alli electrolytic is fine for this.

\$\endgroup\$

Your Answer

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