Capacitor as power supply circuit

[mrpsychotic]

I have seen those heavily discounted maxwell ultracapacitors on electronic goldmine lately, and with the holidays around the corner I plan on having my hands on one soon.

The two possibilities i might get are 1 3000F 2.7V capacitor, or 2 2600F 2.5V capacitors which can make 5200F 2.5V or 1300F 5V.

(A side question: capacitors cannot really be chained in a long series like resistors can without running into problems, right? Isn't there a problem with the energy not being distributed evenly?)

My main question is what circuit or ic I can use to turn this large capacitance into a useful power supply. Because of its discharge curve, If I wanted to run something that needed 2 volts, it would stop running although there is still plenty of energy. The other issue is the inherent low voltage of ultracapacitors in general. I hear this is because the activated carbon used to achieve such high capacitance has a low breakdown voltage and there isn't much they can do about it. If I wanted to create a power supply that delivered a constant 5 volts from a capactior with lower voltage, how would I go about doing this?Thanks!

[Simon]

why do you want to use a super capacitor as a power supply ? why not just use a power supply ?

I'm not certain but beware with large capacitors, they could draw a large amount of current on charge-up and blow things (once a diodes data-sheet would tell you the maximum capacitance it can charge)

[Feanor]

Boost converter is the answer. You can use a switch mode boost converter to get your required 5V out of the super capacitor even as it discharges.

Simon is right, be careful when charging or discharging these things! A small accident like shorting out your 9V battery terminals will not cause much damage but a 2300F capacitor is another story. Use a series resistor to control the current at least.

Generally the higher the capacitance the lower the rated voltage is, it all comes down to the formula for capacitance of two parallel plates with a dielectric between them. A is the area of the plates, Er is the relative permittivity and pi is well, that's pi.

C = (Er* A)/(4*pi*d)

The d is the distance between the plates of the capacitor. So if we make this distance really small we get a large capacitance. But if this distance is very small then the voltage required to breakdown the dielectric between the plates is also very small, because they are so close together. We also want to choose a dielectric with a high relative permittivity so we can not always choose one with a high breakdown voltage.

[Murphy]

Ultracaps are a great replacement for batteries in some applications because they can be charged(and discharged) extremely quickly, their ESR is very small. Hundreds to thousands of farads can supply significant power for a reasonable amount of time and then be zapped back to full in seconds.

Yeah you need a boost converter, like the ones in LED flashlights that suck a AA all the way down.

Not sure on the load balancing, usually it's not a big deal but ultracaps could be different. I know the humongous banks that Maxwell sells have some sort of active circuitry attached to them that looks like it's for load balancing.

[Simon]

the most efficient way of charging the caps would be a switchmode converter with an automatically increasing voltage, probably you could make it current controlled up to a certain max voltage so you don't dissipate heat in the resistors and have control over the speed of charge and current demand

[Zero999]

Super capacitors still have poor energy densities to batteries.

1300F charged to 5V stores 0.5*5²*1300 = 16.25kJ

But if the circuit only works down to 2V 2.6kJ will still be left over so the useful energy storage is 13.65kJ.

A 1.2V 2.5Ah NiMH AA cell stores 1.2*2.5*60*60 = 10.8kJ (assuming it's charged to 1.4V and discharged to 1V with a constant current load) but it's much smaller and cheaper than huge capacitors. Four little 1Ah AAA cells in series will store 17.28kJ and will power your circuit for longer than those huge capacitors will.

The only advantage of huge capacitors is they can be charged and discharged very quickly, otherwise they're inferior to batteries.

EDIT:
The leakage current is very high, 5.2mA so they won't hold their charge for very long like AA cells can.
http://www.goldmine-elec-products.com/prodinfo.asp?number=G17930

[mrpsychotic]

Thanks for all of the informative replies. I will look into the things discussed.

My motivation for using a supercap is to use it in energy harvesting applications. Charging many times and incomplete, fractional charges would wear the batteries horribly.

[Simon]

well lead acid batteries don't mind random charging cycles

[Mechatrommer]

bigger capacitor is not meant to become a battery, it is meant for storing higher energy for a short period of time to cope with high periodic (non continuous) pulsing energy demand. these link are not meant for EE geeks, they only care about the use of it, they are capacitor's fanatic! but you may learn something from them

http://www.audio-n-more.com/installation_accessories/capacitors.shtml
http://www.caraudiohelp.com/car_audio_capacitor_installation/car_audio_capacitor_installation.htm
http://www.diy-caraudio.blogspot.com/
http://www.diyelectriccar.com/forums/showthread.php/capacitor-banks-26137.html
http://www.physicsforums.com/showthread.php?t=327153&page=5

[nyo]

Careful with those caps, if you short the terminals with a screwdriver or wrench, it will easily melt you tool

[slburris]

It's interesting that the car audio guys show these super caps
connected right on the 12v supply right at their audio amps.
But if you have ever looked at the 12v power while a car is running, you
know that there's some pretty nasty noise spikes going on.
It's not a very clean evironment electrically at all.

So, how do the capacitors avoid getting damaged by overvoltage?
They certainly aren't rated to several hundred volts?

Having looked at some circuit diagrams for stuff intended for auto use,
the better power supply have filters including zener diodes to
clean this all up *before* anything gets to low voltage electronics.

Of course a better question is, do using caps like this in autos
really do anything or are they another form of audio snake oil? :-)

Scott

[Simon]

I expect it would depend on the specific circuit, caps on power supplies should always be over rated, using an inductor or resistor in series will help kill the spikes, the capacitor will appear as a very small resistor to high speed spikes, so even a small resistor in series or better an inductor will be large in comparison and take the brunt of the spike, of course TVS diodes are inexpensive and effective i hear

[Zero999]

Of course a better question is, do using caps like this in autos
really do anything or are they another form of audio snake oil? :-)

Yes, I think that's most likely to be the case.

The amplifier should have a good enough supply rejection factor that fluctuations on the power supply shouldn't matter and the impedance of the wiring from the amplifier to the battery will be negligible when compared to the speaker's impedance.

The best way to ensure good power supply quality to the amplifier is to connect it as close to the battery as possible with thick cable which doesn't have to be silver, gold plated or even oxygen free. 

[Mechatrommer]

It's interesting that the car audio guys show these super caps
connected right on the 12v supply right at their audio amps.
But if you have ever looked at the 12v power while a car is running, you
know that there's some pretty nasty noise spikes going on.

bigger capacitor is meant for storing energy, not filtering any noise, for filtering purpose smaller farads and higher volt are used, just like in EE circuit. but as i said, most car audio nerds know nothing about it, they only believe what the retailer said, at least nerdies in my place here.

those bigger capacitors will play their roles when the speaker goes "BOOM" at frequency usually at 10-50Hz (subwoofing bass) at close to 90-120dB sound pressure, at that moment no battery/alternator will be able to provide the energy needed. without it, the battery level will goes floored (shorted), which will damage the amplifier (unstable). mainly it will be coupled with super high Watt amplifier/speaker but battery and alternator space is limited. and its mainly only for competition, NOT a "comfortable cruising experience"

ps: i know coz my bro is sound competition nerd. at right this moment, he is on his way back from main city (hoster) from a sound competition. they got 3rd place in the state and they use 2-3 lorry sized LA batt in his sedan car he knows sh*t about EE though

[Simon]

The capacitor will behave as a resistor to varying voltages (AC/spikes) to the tune of R=1/(6.28 x C x F) as you can see the bigger the capacitance the smaller a resistance it will look like, so it will hold the 12V DC but bypass any spikes to a lesser or greater degree, of course over a certain value your just being silly. your best bet for filtration is probably a low pass filter using a large capacitor preceded by an inductor and put a diode in series with the inductor so that if the car suddenly draws power like at startup it will not upset the radio due to a voltage dip. I have a friend that has a diesel land rover and if the radio is on and the ignition is started it kills the radio due to the amount of power the plugs draw on startup which causes such a voltage drop that the radio "drops out" but a diode would stop the power running back from the capacitor and keep it going during those seconds of startup

[Mechatrommer]

So, how do the capacitors avoid getting damaged by overvoltage?
They certainly aren't rated to several hundred volts?

as the short spike is concerned, big caps are immuned to that, they are very slow adapting to the hi volt spike [dv = (i/C) dt] they appeared to be opened short circuit from the spike point of view. as i said, for a short time only, for a longer period, hi volt exposure will damage the caps... as usual.

[cybergibbons]

The best way to ensure good power supply quality to the amplifier is to connect it as close to the battery as possible with thick cable which doesn't have to be silver, gold plated or even oxygen free. 

Gold or silver plating are far more useful than OFC copper. Plating means reliable, corrosion resistant connections that can be reconnected many times. OFC copper, well, has no advantage at all.

[Simon]

So, how do the capacitors avoid getting damaged by overvoltage?
They certainly aren't rated to several hundred volts?

as the short spike is concerned, big caps are immuned to that, they are very slow adapting to the hi volt spike [dv = (i/C) dt] they appeared to be opened circuit from the spike point of view. as i said, for a short time only, for a longer period, hi volt exposure will damage the caps... as usual.

no they appear as a closed circuit and with another resistance in series for a potential divider with very little voltage across the capacitor

[Mechatrommer]

Gold or silver plating are far more useful than OFC copper. Plating means reliable, corrosion resistant connections that can be reconnected many times. OFC copper, well, has no advantage at all.

as the professional they are for. the pro wont want to bother with cleaning the connector rust each time they are unplugged/plugged. but if the copper is properly cleaned and tightly connected, it makes not much difference, it will come back to "surface area law". the keypoint for gold plating, is "durability and no rust", it has nothing to do with electrical property, well... you may say the resistance coefficient is better for the gold, ok, i wont argue about that as long as you have the money

ps: there are also silver wire/cable for sale, not sure about the gold. but only if you got a thick pocket

[Mechatrommer]

no they appear as a closed circuit and with another resistance in series for a potential divider with very little voltage across the capacitor

yup i screwed up on that , msg corrected.

[Simon]

Copper is actually a slightly better conductor than gold, however gold is a noble metal and will react with virtually nothing where as copper will very quickly tarnish, thats why gold is used in space missions and no doubt the reason god wanted moses and his people to give him all the gold they conquered but told them they could keep the women  

[Mechatrommer]

Copper is actually a slightly better conductor than gold, however gold is a noble metal and will react with virtually nothing where as copper will very quickly tarnish, thats why gold is used in space missions and no doubt the reason god wanted moses and his people to give him all the gold they conquered but told them they could keep the women  

haha! this is good! gold can substitute women!
or gold are meant for women!

[Simon]

Copper is actually a slightly better conductor than gold, however gold is a noble metal and will react with virtually nothing where as copper will very quickly tarnish, thats why gold is used in space missions and no doubt the reason god wanted moses and his people to give him all the gold they conquered but told them they could keep the women  

haha! this is good! gold can substitute women!
or gold are meant for women!

no God wanted the gold for his technology and let the Israelites keep the rest of the spoils of war (hm a peaceful god telling his people they could conquer others)

[Mechatrommer]

man! the same thing here! why i saw a "µ" in the OP! ... nevermind!

[Zero999]

The oxygen free, gold or silver statement was sarcastic.

those bigger capacitors will play their roles when the speaker goes "BOOM" at frequency usually at 10-50Hz (subwoofing bass) at close to 90-120dB sound pressure, at that moment no battery/alternator will be able to provide the energy needed. without it, the battery level will goes floored (shorted), which will damage the amplifier (unstable). mainly it will be coupled with super high Watt amplifier/speaker but battery and alternator space is limited. and its mainly only for competition, NOT a "comfortable cruising experience"

That's not true, 10Hz to 50Hz is above the frequency range of such a large capacitor which will have a self resonant frequency of less than 10Hz.

As I said before, the impedance of the battery and wiring should be negligible compared to the speakers. Suppose you have two front speakers, two back and a dual coil sub in the boot, that's 6 x 4 Ohm loads in parallel, a total impedance of 4/6 = 0.667 Ohms. The maximum theoretical peak current will be 12/0.667 =  18A, the maximum RMS power will be just over 100W and this is at clipping with the alternator running.

Your headlights should consume more than that continuously.

To get more power, you need to connect more speakers in parallel or install a boost converter.

[mrpsychotic]

O.K. all that stuff is nice, but do you have a specific chip in mind that is a solution to the original post. For a 5v boost converter in a dip package, the following criteria are arranged roughly in order:

-Very low minimum voltage for the input
-Efficiency
-Maximum load able to deliver.

[Zero999]

You can get DC-DC converter ICs which output 5V and run down to under 1V, check out the National Semiconductor, Linear and Maxim websites.

Your first two requirements make sense but your last doesn't. What do you mean? Are you driving an LED which uses 20mA or a motor which uses 20A?

[Murphy]

It doesn't sound like this is intended to power any specific device.

To figure out maximum load for the boost converter I guess you could look up the datasheet for the caps and see how fast they're rated to be discharged, then aim around there.

I like searching for parts on Mouser, their web site is quite good.

[mrpsychotic]

Your first two requirements make sense but your last doesn't. What do you mean? Are you driving an LED which uses 20mA or a motor which uses 20A?

Sorry I meant to say:
"Maximum load able to deliver around 250mA."

[scrat]

As I said before, the impedance of the battery and wiring should be negligible compared to the speakers. Suppose you have two front speakers, two back and a dual coil sub in the boot, that's 6 x 4 Ohm loads in parallel, a total impedance of 4/6 = 0.667 Ohms. The maximum theoretical peak current will be 12/0.667 =  18A, the maximum RMS power will be just over 100W and this is at clipping with the alternator running.

Your headlights should consume more than that continuously.

To get more power, you need to connect more speakers in parallel or install a boost converter.

I think that every high power car amp has to have a boost, otherwise its power will be very limited. So current on the 12V rail could be much higher than those 18A. They speak of thousands of rms Watts...
I'm not a car audio nerd, so I can't be sure.

[Zero999]

I think that every high power car amp has to have a boost, otherwise its power will be very limited. So current on the 12V rail could be much higher than those 18A. They speak of thousands of rms Watts...
I'm not a car audio nerd, so I can't be sure.

That's sort of my point.

If the boost converter is well designed, it should have plenty of decoupling capacitors both on the input and the output. The amplifier shouldn't be unstable, even if the voltage does drop a little on the input because the output should be well regulated.

[scrat]

Back on the original topic...

Supercaps are widely used for energy scavenging, especially in those cases where no maintenance is possible. Their life cycles and power (current) are not equaled by batteries.
They are extremely sensitive to overvoltage, so when using more of them in series one should beware of unbalancing, by measuring voltage or providing another means of protection (big assemblies come with integrated protection).

On the energy exploiting side, AFAIK the ICs which can boost from the lowest voltage are the TPS61200 series from TI. They can power up from voltages down to 0.3V (while providing step-down regulation too). This enables charging a battery from a common peltier cell at low temperature difference, for example.

High currents are usually an issue for the charging circuit only, not for the cap, so one could charge at the highest current possible (and available). Even with integrated regulators (converters) the usual feedback can be altered to achieve current regulation from voltage regulation.

[saturation]

I'm not sure what they cost these days, but ultracaps still cost more overall to design as a portable power supply over a LiIon or NiMH battery, to power a device.

For practical devices, I do own a power screwdriver that uses ultracaps which charges in <=1 minute, and it puts out enough power to run the driver for most any practical job around the house.  Given the charge time is short, its very practical in this design, and uses no voltage regulation, the caps are simply tied directly to a DC motor.  From near zero volts, to 5+ V, it will charge in 90 seconds, but the torque is pretty useless once below 2.5V, so for practical purposes the caps as useful down to there only.  Charging is exponential.

http://www.flashcellscrewdriver.com/

The main advantage of ultracaps is that its rated life is not tied to charge cycles like LiIon or NiMH, or aging in 5 years like any Li cell.  The expected lifespan of a cap is 10-20 years, due to deterioration of its electrolyte, and its independent of charge cycles, so use it as much as you can.  I also use it to power a power vac, it gets about 1 minute of use before it dies, but its so easy to charge up thereafter.

Given this device is out of production since 2007, if you bought a 2nd hand unit, unless the motor burns out or gears strip it should last at least to 2014, to 2021.

[sonicj]

"stiffening capacitor" is the term commonly used for large capacitors in parallel with the power supply of car amplifiers.  its intent is to supplement the power supply during brief transients that consume more power than is otherwise available. larger/more batteries and a high output alternator are likely better solutions if the audio system needs have outgrown the car's electrical system capabilities. the topic of many heated discussions over the years...
-sj

[qno]

Generally speaking using a Capacitor as a voltage source is not a good idea when you look at the discharge characteristic.