10uF ceramic capacitors?

Dicky Mint

Senior Member
Hi There

I am using 10uF ceramic capacitors to decouple my PICAXEs and the thought just occured...

Are ceramic capacitors capeble of supplying the current resevoir that an electrolytic would?

They're convienient, fairly cheap in volume and, usefully, they have a 5mm footprint which makes it easy to run a couple of thinner tracks between the pins. Very useful in PICAXE download circuits!

Rick
 

hippy

Technical Support
Staff member
I'm not familiar with uF ceramic capacitors and usually raise an eyebrow when people suggest 10uF and low values as reservoir capacitors as I have always tended towards much larger electrolytics, 22,000uF and even higher on the input side of a regulator, 470uF on the output.

I am not sure what reservoir capability 10uF would give, but it's always seemed sensible to take a 'the more the merrier' approach to me, unless a circuit is battery powered or a regulator insists on something else.

Whatever is used I would suggest reservoir caps plus decoupling caps because they fulfil different roles; so yes you can use 10uF with 100nF. To me it would just be an issue of how well that 10uF performs at being a reservoir capacitor. I wouldn't have thought 10uF was suitable though for decoupling; that should be nearer 100nF.

I'm no analogue electronics expert so I think it may well depend on what your full circuit is.
 

techElder

Well-known member
Your post is somewhat implying that you are placing a "reservoir" capacitor close to the PICAXE power pin. Is that what you are saying?
 

Dicky Mint

Senior Member
Hi thanks for your replies

Yes Tex that is what I'm saying!

I have a 100nF in parallel with a 10uF ceramic as close as practically possible to the power pins of a PICAXE.

Sometimes I have rather long tracks on the power lines of my PCBs (~4") and therefore place the 10uF capacitor accross the power pins of the PICAXE to act as a small resevoir.

Rick
 

rq3

Senior Member
Hi thanks for your replies

Yes Tex that is what I'm saying!

I have a 100nF in parallel with a 10uF ceramic as close as practically possible to the power pins of a PICAXE.

Sometimes I have rather long tracks on the power lines of my PCBs (~4") and therefore place the 10uF capacitor accross the power pins of the PICAXE to act as a small resevoir.

Rick
This is excellent design practice. A farad is a farad, whether it is electrolytic or ceramic. The benefits of modern high capacity ceramics include:
1) They are small
2) They (usually) have a lower Equivalent Series Resistance (ESR) than an electrolytic
3) They are (usually) less expensive than an equivalent tantalum electrolytic
4) They are somewhat forgiving of short term reverse polarity
5) They tend to have better wide range temperature characteristics
6) They tend to have a longer lifetime, and tend to fail open, rather than shorted like an electrolytic
7) In surface mounted packages, they tend to have a higher self resonant frequency (SRF) due to the small size and lack of wire leads, so they tend to function better as high frequency bypass devices than the equivalent leaded tantalum. They should not be used as high frequency by-pass, but they generally don't hurt, unlike large leaded electrolytics (especially aluminum).

Drawbacks:
1) They can get really expensive in large values. REALLY expensive.
2) They are not as tolerant as tantalum electrolytics of excess temperature during soldering

The use of bypassing is a science that devolves into an art with more experience with circuit design. If you have very high current devices (many pulsed LEDS, motors, many fast switching digital devices) a local 10 nF and 10 uF at each IC package is a very good idea, with massive (>100 uF or more) directly at the high current devices.

Keep in mind that there are nuances to this. For example, many linear voltage regulators cannot tolerate a short on their input if they have more than about 25 uF on their output, (voltage dependant) since the caps discharge backwards through the regulator. This varies by device, and there are design methods that can (and should) be employed in such cases.
 

westaust55

Moderator
Another aspect that should be considered when using ceramic capacitors, particularly when intended to use larger uF ratings for bulk storage purposes in lieu of electrolytic and tantalum types is the reduction in capacitance as the working voltage approaches the rated voltage.

In simplistic terms, where the working voltage is way < 20% of working voltage there is no change in capacitance (uF) rating.
At the other end of the scale, if the working voltage is near/at the rated voltage the change in capacitance can be significant depending upon the series (dielectric type) with reductions of from 50% to 80% of the rated capacitance.

Thus as an example a 6V 10 uF cap used in a 6 Vdc circuit may have an actual capacitance in the range of ~2 uF to ~5 uF.
Similarly with a 6 V 10 uF ceramic cap used in a 5 Vdc circuit may have an actual capacitance in the range ~2.5 uF to ~5.5 uF.

The graph at page 143 in this datasheet shows the change in capacitance versus working voltage for a couple of ceramic capacitor series.
https://datasheet.lcsc.com/szlcsc/10uF-106-10-10V_C77044.pdf
 

rq3

Senior Member
Another aspect that should be considered when using ceramic capacitors, particularly when intended to use larger uF ratings for bulk storage purposes in lieu of electrolytic and tantalum types is the reduction in capacitance as the working voltage approaches the rated voltage.

In simplistic terms, where the working voltage is way < 20% of working voltage there is no change in capacitance (uF) rating.
At the other end of the scale, if the working voltage is near/at the rated voltage the change in capacitance can be significant depending upon the series (dielectric type) with reductions of from 50% to 80% of the rated capacitance.

Thus as an example a 6V 10 uF cap used in a 6 Vdc circuit may have an actual capacitance in the range of ~2 uF to ~5 uF.
Similarly with a 6 V 10 uF ceramic cap used in a 5 Vdc circuit may have an actual capacitance in the range ~2.5 uF to ~5.5 uF.

The graph at page 143 in this datasheet shows the change in capacitance versus working voltage for a couple of ceramic capacitor series.
https://datasheet.lcsc.com/szlcsc/10uF-106-10-10V_C77044.pdf
An excellent point, and illustrates why EEStore never went anywhere!
 

techElder

Well-known member
Personally, I don't think a "reservoir" capacitor is necessary at the power supply pin of the PICAXE in most applications. The PICAXE by design isn't capable of high current sink/source where a bunch of capacitance is needed right at the pin (especially with just 4" of lead.)

Sure, put a high frequency capacitor at the power pin, but if you build your power supply for your application correctly, you don't need to go around your circuit poking capacitors here and there like daises in a garden.

If you have high current sink/source areas of your application circuit, you should be sourcing them from a separate part of your power supply. Separate those high current sections from the logic circuit supply.

And building your power supply correctly means that even if you are using a battery pack, you should add "reservoir" capacitors. Batteries have some internal resistance, too.
 

lbenson

Senior Member
In dozens of (non-demanding) picaxe circuits I have used .1uF ceramic and 10uF (25V) electrolytic capacitors and no other capacitors with no known problems. But I claim no theoretical understanding and no broader practical experience. Usually this is with power from a plug pack.
 

AllyCat

Senior Member
Hi,

The PICaxe will probably "work" with a 10uF ceramic capacitor, and also with a single 100 nF ceramic, or a 100 uF electrolytic or maybe even with no capacitor at all. But it's not generally considered to be "best practice":

If a "reservoir" capacitor is needed, then as hippy says, 10 uF is probably too small, although I wouldn't normally go as high as 22,000 uF, except for high-current mains-frequency applications.

However, supply decoupling capacitors normally "fail" (to perform their intended function) when they become series resonant with the inductance of their own leads, or the PCB tracks etc.. Unlike parallel resonance, series resonance is a high impedance state, i.e. little current flows but the oscillating voltage can be quite high. That's not a good situation for a component which is supposed to be delivering current to stabilise the voltage on the supply rail.

If you make the value of the decoupling capacitor 100 times larger (than normally recommended) then you will reduce its series resonant frequency by ten times. So instead of decoupling the supply rail up to, say, 80 MHz, it will be only 8 MHz; not a good idea for an X2 at its default clock frequency. :(

Many years ago I worked with a colleague who was exceptionally talented at getting high switching speeds at fairly high power with quite modest components. He always maintained that the parallel connection of a ceramic capacitor with an electrolytic was "optimum". Not only does each perform its required (decoupling) function at the intended frequencies, but the electrolytic (being relatively "lossy" at high frequencies) acts as a "damping" circuit (rather like a R-C snubber circuit) to the potentially "high Q" series resonance of the smaller ceramic capacitor with its own series inductance.

Cheers, Alan.
 

erco

Senior Member
In dozens of (non-demanding) picaxe circuits I have used .1uF ceramic and 10uF (25V) electrolytic capacitors and no other capacitors with no known problems. But I claim no theoretical understanding and no broader practical experience. Usually this is with power from a plug pack.
Ditto, except I use batteries (same batteries powering at least one servo). No complaints.
 
Top