| Electronics > Projects, Designs, and Technical Stuff |
| Do we still need electrolytic capacitors (for basic decoupling)? |
| (1/1) |
| TheUnnamedNewbie:
This might sound like a big of a strange question, but I have spent the last few days diving deep into the concepts of decoupling capacitors, with lots of simulations and modeling. It got me thinking - Most LDOs have pretty good output impedance up to a few tens of kHz. Past the few tens of kHz, you are already into the regime where big electrolytic caps (especially when combined with trace inductance) are past self-resonance. So that makes me wonder - are big electrolytics still useful? (I realize some other applications where you have high pulse-loads are a different matter entirely). Do you still use them? Are you sure there is still a point to using them? |
| Siwastaja:
Great for damping resonances and peaking whenever you need to use large values of low-ESR ceramic capacitance - such as input circuitry of most switching converters. Say, you have 100uF of ceramic capacitance - put a 470uF cheap elcap in parallel. Boom, 100% input overshoot reduced to a few %. |
| David Hess:
--- Quote from: TheUnnamedNewbie on April 02, 2020, 12:54:07 pm ---It got me thinking - Most LDOs have pretty good output impedance up to a few tens of kHz. Past the few tens of kHz, you are already into the regime where big electrolytic caps (especially when combined with trace inductance) are past self-resonance. --- End quote --- Just because they are past self resonance doesn't mean they are not providing useful decoupling; their impedance is still low to usefully high frequencies. --- Quote ---So that makes me wonder - are big electrolytics still useful? (I realize some other applications where you have high pulse-loads are a different matter entirely). --- End quote --- They are hard to beat economically when bulk output capacitance is required. I only see them entirely replaced when long life and high temperature operation is required. |
| Navigation |
| Message Index |