Question About Capacitors
NWCCTV
Posts: 3,629
How does one know which voltage of caps to use? I see there are various voltages of standard and ceramic type capacitors. I am looking at purchasing a capacitor kit but I want to make sure I get the correct one. It seems that 25V are the most common ones available for standard caps and 50V for ceramic.
Comments
So if you are talking power supply decoupling caps just get something rated at twice the power supply voltages you might be using.
Most caps in a signal path will never see more voltage than the power supply so the same applies.
Of course if you are building some funky voltage multiplier or boost circuit you will need to use higher voltage ratings. Just now I have been building boost converters that generate 200v so I had to find 250v caps.
If you are building resonant circuits, coil and cap in parallel, they can also get up to higher voltages than the power rails.
Capacitors are just about the most diverse in types to select from and many are designed for a specific use or application. So of the retail solutions are cheap but not the best, other devices that are good are bulky in size, and there is a whole 'religious reverence' for the perfect audio or perfect RF capacitor that confuses the issue.
For digital work, I would just get caps according to mainly three tasks - [a] by-pass capacitors (ceramic 0.1uf are okay), ripple and power transient capacitors (10 to 100 uf electrolytics are okay), timing capacitors (for RC circuits and what to use is very dependent on how precise you desire ==> but temperature will always cause precision to be less than ideal).
Going for sizes below 25V for inventory can be counter-productive. You might just find that your inventory is too low voltage for a new and different project.
There are other categories. I suppose that I should mention that you might need a few picofarad ceramics for crystals, though the Propeller doesn't require these.
Oh yes,interesting, there is a statement of "fact" that I have never heard of before, sounds non-intuitive and demands further investigation.
I can't immediately get google to cough up any discussion of this but this capacitor life time calculator seems to disagree:
http://www.illinoiscapacitor.com/tech-center/life-calculators.aspx
For example: A smoothing capacitor operated at 5v and 20 degrees centigrade has a temperature rating of 80 degrees and a load life rating of 100. The calculated life time for some common voltage ratings is:
5V 6400 hours
6V 7680 hours
7V 8960 hours
8V 10240 hours
9V 11520 hours
10V 12800 hours
11V 12800 hours
After that life does not increase further with voltage rating.
Anyone know any more about this rated voltage vs life thing?
Everyone using electrolytic capacitors needs to be aware of this:
Life in the operational environment is dictated by the Law of Arrhenius, which dictates that the capacitor life is a function of temperature and DC voltage. As a rule of thumb, the life doubles for each 10 °C lower operating temperature.
So keep you electrolytics out of the way of warm things like regulators and power transistors.
P.S.
RDL2004, seems your rule of thumb is pretty good. It gets you almost up to maximum life expectancy. Going higher just brings physically bigger capacitors, more expense and perhaps other issues like inductance or whatever.
The is a lot to think about with capacitors.
This life-temperature dependence actually impacts how you should derate the voltage on the capacitor. Your first thought might be to increase the capacitor’s voltage rating to minimize the possibility of a dielectric failure. However, doing so can lead to a capacitor with a higher equivalent series resistance (ESR). Because the capacitor typically has a high ripple current stress, this higher resistance leads to extra internal power loss and increased capacitor temperature. The failure rate increases with the increased temperature. In practice, aluminum electrolytic capacitors typically are used at about 80% of their rated voltage.
That is to say, a higher voltage rating increases ESR. Increased ESR increases losses and causes heating. Heating shortens life. Ergo, don't use a higher voltage rating than you need. Also use a low ESR type cap in such applications.
Seems RDL2004 has alerted us to a real phenomena that may be important in some situations.
http://www.eetimes.com/author.asp?doc_id=1279791
Some things though are important. In an AC conversion to DC context, voltage ratings need to take into account that AC votages are 71% of actual peak voltages, due the RMS (root mean square) calculations. in other words, 24VAC needs a capcitor rated better than 33.8V, and even 18VAC comes in at 25.35 volt... an annoy 0.35 volts above 25volts.
Turns out the humble capacitor is a complicated thing. This business of voltage rating vs lifetime is a real effect that may need ones consideration. Then there are issues like the inductance of a capacitor. Or did you know that the capacitance of a ceramic cap changes with DC bias? That may be significant sometimes. The effect of temperature on an electrolytics capacitance is huge. And so on.
These are not fads but real measurable effects.
From what I can gather ... mylar, Teflon, polyethylene, and polypropylene are are some of the most stable. Electrolytic cap suffer from leakage, chemicals that age, and some chemicals used to be hazardous. Ceramics seem to have microphonic properties that make them unattractive to audio. Rolling up foils can create induction problems. Some change value over age and many have very wide range of tolerance for a specified value. Wanting higher capacitances and higher operating voltages in smaller packages are in direct conflict.
+1
Thanks for doing some research on the topic Heater.
Once again, I learned something new on the forum.
relative permitivities in the 1000's (most materials have relative permitivities in the range 1 to 10).
These are the highly non-linear and microphonic capacitor materials, as electric fields distort the arrangement of atoms in
the crystal (nuclei move, not just electrons, in response to the field). Such materials are very lossy at RF frequency.
There are other ceramics used for low value capacitors (your 22pF crystal load caps for instance) which are
stable, low loss and well behaved and used in RF circuitry everywhere.
For audio filter use high values of capacitance are needed without microphonics and with accurate temperature-stable
values - ie no BaTiO4 ceramics. Other ceramics are too small a capacitance value, generally. Plastic film caps are
typically used because they are temperature stable, even though they can be bulky. Electrolytics aren't stable
in value, but are not microphonic so get used for coupling and decoupling in audio all the time.
For power electronics ESR, ESL and self-heating due to ripple currents are important - there are ranges of electrolytic
designed especially for these properties, and for high frequency switchmode ceramics are common (although they
can be lossy) due to the compact size and ease of surface mount.
If you want a quality capacitor for general purpose use (perhaps in some instrumentation), teflon and polystyrene film
are good choices (if not inductively wound), since they tend to tick all the boxes (except compactness and cheapness).
Polystyrene cannot be used surface mount (it melts), so PPS (polyphenylene sulfide) seems to be the material of choice
these days.
In particular teflon can be very low loss all the way up to microwave frequencies, and is used for high quality coax cables too
for that reason.
If you think chosing capacitors is tricky, inductors are far worse!