Dear friends!
I wonder why many circuits use small capacity electrolytic capacitors (less then 100uF) as bypass and decoupling capacitors, rather then ceramic ones? After all, ceramic ones are smaller in size and have a comparable price with better characteristics.
And if ceramics are indeed better, then what type of dielectric is better to use for the stated purposes
I wonder why many circuits use small capacity electrolytic capacitors (less then 100uF) as bypass and decoupling capacitors, rather then ceramic ones? After all, ceramic ones are smaller in size and have a comparable price with better characteristics.
And if ceramics are indeed better, then what type of dielectric is better to use for the stated purposes
Last edited:
2 cents >
To my knowledge ceramic capacitors are not made in values as high as 100uF > that is why people use electrolytics for such a value.
Ceramics are very good & long stable life for 'very low values'.
'Polly-caps' are generally preferred for low-medium values.
https://www.google.com/search?q=wor...jKgB964ArIHCDAuMS4yOS4yuAfzOg&sclient=gws-wiz
To my knowledge ceramic capacitors are not made in values as high as 100uF > that is why people use electrolytics for such a value.
Ceramics are very good & long stable life for 'very low values'.
'Polly-caps' are generally preferred for low-medium values.
https://www.google.com/search?q=wor...jKgB964ArIHCDAuMS4yOS4yuAfzOg&sclient=gws-wiz
Panasonic polymer hybrid electrolytics (EEH-ZA or EEH-ZC series) have low ESR as in ceramic capacitors but don't have voltage-dependent capacitance as in class II ceramics.
Watch out. Ceramic capacitors can cause some circuits to become unstable, especially some regulators. You need some series resistance in those cases. They can also ring in some switching circuits with other capacitors.
These days I use Poly-Aluminum capacitors for bypass or higher frequency (switcher) filter capacitors. Electrolytic capacitors are sometimes more suitable in line frequency applications.
These days I use Poly-Aluminum capacitors for bypass or higher frequency (switcher) filter capacitors. Electrolytic capacitors are sometimes more suitable in line frequency applications.
Don't know about that. I've mostly used them as LDO output caps where low ESR is required and class II ceramics would need to be paralleled due to derating. E.g. https://www.diyaudio.com/community/posts/7728111/
Hi capslock,
High K ceramics don't make good coupling caps, NP0 / C0G are fine. I am undecided on Poly-Aluminum, I haven't seen any increases in distortion in experiments.
With coupling caps, series impedance usually isn't any concern. The circuit impedance is high enough to swamp any of that out. The voltage co-efficient might be important, I haven't looked at that. Electrolytics for coupling caps are fine.
In applications where the cap is a filter element, everything changes. Leakage current may also be important, and that you have to watch over temperature.
For decoupling (this application for the thread), different story again. Ceramic caps can be fine, as can Poly-Aluminum. I wouldn't waste the money and have a huge package like a Polypropylene foil capacitor. Nothing wrong with it, it's just expensive and huge.
High K ceramics don't make good coupling caps, NP0 / C0G are fine. I am undecided on Poly-Aluminum, I haven't seen any increases in distortion in experiments.
With coupling caps, series impedance usually isn't any concern. The circuit impedance is high enough to swamp any of that out. The voltage co-efficient might be important, I haven't looked at that. Electrolytics for coupling caps are fine.
In applications where the cap is a filter element, everything changes. Leakage current may also be important, and that you have to watch over temperature.
For decoupling (this application for the thread), different story again. Ceramic caps can be fine, as can Poly-Aluminum. I wouldn't waste the money and have a huge package like a Polypropylene foil capacitor. Nothing wrong with it, it's just expensive and huge.
High K Ceramic caps are microphonic as well. Not good choices in audio circuits.
Ceramic caps can be trouble in high impedance circuits with a bias voltage on them. I can't remember the threshold for metal migration but it was maybe just a volt or so. The extremely thin insulating layers are not perfect in these capacitors and the metal will migrate through even the smallest holes and whiskers will short the plates together and then get blown like a micro fuse causing noise spikes. It was a nightmare in a preamp I was involved with. There are ceramic caps specially made without these flaws, but they are very expensive.
I know that class II caps are bad news for coupling. I don't quite get why I shoud need to parallel one in a coupling application where the polymer electrolytic is not subjected to high voltage at high frequency. All it's going to see is the bias voltage and the LF AC below the high pass frequency.
Bohrek, why do I need one in its regular application? That was also not clear to me from the link you provided?
Bohrek, why do I need one in its regular application? That was also not clear to me from the link you provided?
Low ESR of LDO output capacitor is beneficial for bypassing high frequency noise and for fast load transients. Here is some information on this: https://www.ti.com/lit/an/snva167a/snva167a.pdf
In the example I gave the recommended output capacitance is 50uF and output voltage can be up to 30V. Most class II ceramics above 22uF top at 25V. And at that voltage the best ones (X7R) would need to be derated by over 50% (X5R probably closer to 70%) due to voltage-dependent capacitance. So to reach 50uF at 25V requires at least 3 paralleled 47uF/25V X7R class II ceramics.
If you are going to bypass the supply at high frequencies, it should be done right at the load.
Many voltage regulators are simply an amplifier and voltage reference when you boil them down. Undamped capacitance across the output make many unstable. That's when you can run into trouble with Ceramic capacitors.
Many voltage regulators are simply an amplifier and voltage reference when you boil them down. Undamped capacitance across the output make many unstable. That's when you can run into trouble with Ceramic capacitors.
Depends on the application and LDO. In my example the purpose of bypassing high frequencies is to suppress the noise of upstream DC/DC converters from propagating to the load. LDOs used in this application are stable with low ESR capacitors which should be placed close to the LDO. Naturally additional high frequency bypass capacitors should be used at the load if needed.
In that case, consider a choke (inductor) and bypass before the LDO. Regulators typically don't have great HF rejection and would certainly be improved by reducing the HF component before and not after. A low DCR inductor after your LDO followed by a bypass cap would reduce HF even further. If critical, this is the route I would follow.