Supercapacitor / Ultracapacitor application for driving LED

[vicbin]

Hi all,

Does anyone ever play with this kind of capacitor for flashlight or DC LED circuit ?

This super/ultra capacitor has monstrous capacity starting from 1 Farad to hundredth Farads (1 Farad = 1 Million uF ).

I was wondering if this cap can be use for powering small LED flashlight with DC-DC boost converter. Think of the possibilities compared to ordinary rechargeable battery (Ni-xx,Li-Ion):

- Small and light weight, some 1 Farad model is smaller than CR-123.

- Recharge time in SECONDS, e.g. within working voltage controlled by stepdown circuit from high current source like car battery the charge time can be less than 5 seconds.

- Very high count or unlimited charge/discharge cycles compared to ordinary rechargeable battery.

- It is not possible for the cap to get overcharged (in it's working voltage range). Leak or self discharge current in micro Amp (uA), i.e. it is safe to leave it plugged at car cigar ligther jack without worrying it drain the car's batt.

Now, questions :
btw, Mr.Al are you there ?

- How much energy stored in 1 Farad ? e.g. How long for 1 Farad cap can supply 3 x 20 ma LED ? or better, drives LS in acceptable time frame like 15 or 30 minutes ? Remember, re-charge time in seconds.

- How this cap affects the perfomance for DC step-up boost circuit ? Coupling it parallel with battery or C-out ?

Some links on super cap :

Power Stor supercap data sheet with excellent application notes on it's advantage for DC powered circuit.
Power Store Supercap data sheet and application notes

AVX Supercap data sheet

Cheap super cap (thanks to Marked for the finding)
All Electronics Super Cap

Vic

 

[Marked]

Here is a Co. using ultracaps
http://www.solarcap-lighting.com/
with LEDs.

 

[vicbin]

Wow, this is great, thanks again Marked !

Edited quote from the site's FAQ ...

First, they will save money,..... Unlike rechargeable batteries, which must be replaced every two to three years.......which is guaranteed for at least ten years, and may last almost indefinitely! ..... unlike rechargeable batteries, their ability to store and deliver that energy declines very little with age and use; thus the tiles are almost maintenance free for life. .....can be permanently installed in unserviceable locations and then forgotten about. The solar cells built into these tiles will gather enough energy, even if not located in direct sunlight or on rainy days, to power the lights for at least 12 hours. So they also save energy because they do not need to be connected to a manmade energy supply. They are much easier to install than any other outside lighting, since they do not need to be wired to a power source; or have rechargeable batteries replaced periodically.

<font size="2" face="Verdana, Arial">Mind share any experience/knowledge on this montrous cap ? Again, thank you !

Vic

 

[KenBar]

I traded for some of these a year or so ago. I made it so that it would charge up when I pressed a button. Upon releasing, I would have bright light for about 10 minutes...and then gradually dimming light for 30 min.

I used it for a night light untill I bought a Snzzzlite. I ended up giving it away to a kid who fell in love with it.

These caps look like they are "better" plus I like the solar idea.

Looking forward to seeing what others do with these.

If I am remembering right, I think I used 2 lights...a red and a blue. One would "drop off" before the other one. I think it was the Red that would keep on running for some time.

I sort of wish I had it back! LOL

 

[carbonsparky]

These are neat capacitors. They have been using somthing similar for years for memory back up in phones and computers. The old ones could handle only low voltages and currents, these have changed that. I read about some of these they plan on using in hybrid electric cars.
The biggest drawback for driving led lights is the discharge curve of a capacitor. The voltage output will drop in a almost linear matter as the capacitor discharges. This makes it hard to produce a constant output from led. A dc-dc step up circuit may be just the thing.

 

[Albany Tom]

Originally posted by vicbin:
- How much energy stored in 1 Farad ? e.g. How long for 1 Farad cap can supply 3 x 20 ma LED ? or better, drives LS in acceptable time frame like 15 or 30 minutes ? Remember, re-charge time in seconds.

<font size="2" face="Verdana, Arial">Oh boy. Doing this from memory again, so I'm bound to get into trouble. A Farad is a unit of charge equivalent to the storage of 1 amp of current flowing for 1 second. So a 1 Farad capacitor is a 1000 mAs (miliamp-SECOND) battery. When I was in school we used to make jokes about 1 Farad capacitors, because in those days they didn't exist, and if you built one it would be the size of a Toyota. Times have changed.

So it should drive that 20mA LED for about 50 seconds.

The ways to kill a capacitor are to over-voltage it, reverse voltage it, or over-heat it. To overheat, you discharge and charge too fast. These caps could be charged very quickly, in seconds, but you probably need to limit the current to keep from hurting it. Internal resistance of a cap is usually very small.

You'd also have to regulate the voltage. Using a 3 terminal voltage regulator would probably take care of the current limit, too, as I believe most have internal current limiting.

 

[Jonathan]

Originally posted by Albany Tom:
Oh boy. Doing this from memory again, so I'm bound to get into trouble. A Farad is a unit of charge equivalent to the storage of 1 amp of current flowing for 1 second. So a 1 Farad capacitor is a 1000 mAs (miliamp-SECOND) battery.

<font size="2" face="Verdana, Arial">Close, but not quite correct. A Coulomb is a unit of charge equivalent to 1 amp for 1 second. A Farad is a unit of capacitance which is the number of Coulombs stored up per applied volt. If you apply 1 Volt to a 1 Farad capacitor, you will store 1 Coulomb of charge. 2 Volts stores 2 Coulombs, etc.

As the capacitor discharges, the voltage drops, but when the voltage is higher, more energy is delivered per coulomb. The _energy_ stored in a capacitor is 1/2CV^2. If you had a 1 Farad capacitor charged to 4 Volts, the energy stored would be 8 watt seconds. You could run a nice 1/20 watt LED for a couple of minutes.

-Jon

 

[vicbin]

Thanks gentlemen for explanation, now forget about using it as a power source.

How about for combining it parallel with the battery for step-up boost DC converter ? Is there any "big" difference especially when the battery voltage is close to the minimum start up voltage requirement, assuming there is no soft-start mechanism.

Battery Aerogel Capacitor Hybrid Systems for High Pulse Power Applications

From above document, the "Hybrid Power Pack" (battery + supercap) can deliver longer run time (six fold) for high current application like cell phone. Is this applicable for high power LED like the LS 5w step-up circuit ? Or its just marketing hype ?

Vic

 

[Jonathan]

Using a battery/capacitor hybrid is not marketing hype, but you would need to _very_ carefully consider your application.

Basically batteries can trade off energy storage for internal resistance; some battery designs (notably many lithium batteries) store enormous amounts of energy, but have high internal resistance, meaning that you can't discharge them very quickly.

Capacitors are essentially all the way on the other side of the spectrum. They store very little energy, but can release it very quickly.

Some of the supercapacitors are more in the middle, with more energy storage than most capacitors, but also higher internal resistance, but still lower energy storage and lower internal resistance than most batteries.

So if you had a light that you wanted to 'flash' (turn on for a couple of seconds every minute), then a capacitor would let you use a high energy, high resistance, low power battery. But it wouldn't help at all if you wanted to turn the light on, and leave it on. The reason for the longer apparent run time with cell phones is because cell phones are really intermittent applications, pulsing high drains infrequently.

-Jon

 

[KenBar]

Sat-Cure circuit time?

Any thoughts how this would work driving a SatCure circuit?
I originally put 4 of the old style memcaps in parallel to drive led's directly after charging them up.

I still have a hard time wanting something for nothing as far as LED's go.

 

[MrAl]

Hello,

Here are the results of a simulation run using a
1 Farad Capacitor charged up to either 4 or 8 volts,
with either a Nichia 20ma LED and a series resistor
or a Nichia 20ma LED with a constant current regulator.

Notes:
The times shown below (in seconds) are the logged times after
the Nichia is turned on. The currents below are the currents
measured at that particular point in time.
The voltages shown below are the voltages the cap is charged up to.
The Nichia drops approx 3.3 volts at 20ma.
The currents at time 000 are the initial currents, which would
be measured at the instant the Nichia is first turned on.
A typical Nichia still gives some useable light at 1ma
but not much after that, so most of the experiments are concluded
when the current drops to 1ma.
An ideal switching regulator would keep the LED lit even longer
at a constant 20ma.

4 volts + series 1 ohm resistor
000 sec 50ma
010 sec 30ma
036 sec 10ma
200 sec 01ma

4 volts + series 10 ohm resistor
000 sec 30.0ma
003 sec 30.0ma
040 sec 10.0ma
100 sec 03.0ma
220 sec 01.0ma
400 sec 00.5ma

4 volts + series 20 ohm resistor
000 sec 25ma
009 sec 20ma
040 sec 10ma
243 sec 01ma

8 volts + series 40 ohm resistor
000 sec 87ma
011 sec 70ma
060 sec 30ma
100 sec 16ma
200 sec 04ma
378 sec 01ma

4 volts + series 20ma constant current regulator
000 sec 20.0ma
040 sec 20.0ma
050 sec 10.0ma
100 sec 04.0ma
500 sec 01.5ma

8 volts + series 20ma constant current regulator
000 sec 20.0ma
229 sec 20.0ma
300 sec 03.7ma
500 sec 01.6ma

 

[MrAl]

Hi,

Wow, 4000F ? Pretty cool all right.
How much are they? US dollars?
Can you get higher voltage rating,
and how big are they? Height, width.

Al

 

[MrAl]

Originally posted by vicbin:
Thanks gentlemen for explanation, now forget about using it as a power source.

How about for combining it parallel with the battery for step-up boost DC converter ? Is there any "big" difference especially when the battery voltage is close to the minimum start up voltage requirement, assuming there is no soft-start mechanism.

Battery Aerogel Capacitor Hybrid Systems for High Pulse Power Applications

From above document, the "Hybrid Power Pack" (battery + supercap) can deliver longer run time (six fold) for high current application like cell phone. Is this applicable for high power LED like the LS 5w step-up circuit ? Or its just marketing hype ?

Vic

<font size="2" face="Verdana, Arial">Hi there again Vic,
Sorry i didnt see your question sooner.
In general, a capacitor can not increase the
ampere hour rating of a battery. It can
however seem to provide an increase in run
time if the application draws large amplitude
pulses which often drain regular batteries
down faster then usual. This only happens though
when the system is set up to take advantage
of the capacitors storage. To see any extra run
time, it would be necessary to insure that the
capacitor is charged up slow relative to the
pulse amplitude. This of course would vary
according to how high the amplitude is and
the type of battery.
Now it might just work
out that the batteries' internal impedance is
great enough to recharge the capacitor slow
enough and when a current pulse is in demand,
the lower impedance of the cap takes over,
supplying the bulk of the current. This would
make it possible to simply parallel one of
these caps with the battery, and would certainly
be the best thing to try first. The impedance
of most batteries is something like 0.1 ohms,
so during the pulse i would expect to see the
capacitor delivering at least 5 times as much
current as the battery. Then, when current returns
to normal, the capacitor is then recharged by the
battery.
Note that there would be some duty cycles that
would work better then others, but as long as
the capacitor averages out the current pulse
i think we would see some increase in the run
time for high pulse current apps in general.
In any case, the specific application would have
to be tested with a given cap and a given
battery type to see what advantage would be
obtained.
I would think that in apps like digital cameras,
the caps would increase the run time. Just how
much would be very difficult to determine without
knowing a whole lot of information about the
camera and the type of batteries used.
It would be easier to try it and see.
Of course there is always the possiblity that the
newer cameras might already have a large cap
installed, so that adding another one might
not do anything at all.

For switching regulators, it would again depend on
the amplitude of the current pulse in relation
to the average input current. The larger the
pulse, the more effect the extra capacitor will
have on the run time.

The efficiency of the circuits shouldnt change at
all with the addition of a capacitor on the input.
The apparent efficiency (noted by the run time)
should increase though.

Good luck with your LED circuits,
Al

 

[camisdad]

So, if we could find an Ultracap like Inretech's 4000F but good for twice the voltage (5.0V), if charged at 4V, we would have 32,000 W.sec, i.e., for a constant current of 20mA, 111 hours of juice. Or, for 800mA (a good size power for a new Lumiled), about 2 hrs and 40 mns of juice.

Does this make sense? What is wrong with this result? It just seems too good to be true.

Is Inretech's 4000F capacitor real?

 

[Albany Tom]

Originally posted by camisdad:
So, if we could find an Ultracap like Inretech's 4000F but good for twice the voltage (5.0V), if charged at 4V, we would have 32,000 W.sec, i.e., for a constant current of 20mA, 111 hours of juice. Or, for 800mA (a good size power for a new Lumiled), about 2 hrs and 40 mns of juice.

Does this make sense? What is wrong with this result? It just seems too good to be true.

Is Inretech's 4000F capacitor real?

<font size="2" face="Verdana, Arial">I would think it would have to have wheels on it, but technology keeps moving faster.

I would also treat such a capacitor about like I would a hand grenade. Flash capacitors for handheld flash guns can run around 100 WS, and they can cause stuff to go "boom". Somewhere there's a web page about people that play with big caps...have to track it down...

here you go:

http://www.eskimo.com/~billb/amateur/capexpt.html

 

[camisdad]

Jeez Louise That is some spectacular experiment...

Of course they are charging that capacitor with 150,000V

 

[INRETECH]

Supercaps are cool, I have 2 100F caps
and this one !

http://www.inretech.com/pictures/cap4000F.jpg

 

[INRETECH]

Yes the 4000F capacitor is real

Elna made 10 of these as Prototypes to show at Trade shows, my friend quit Elna and gave me the cap

My cap is remarked to a MAX of 2.3V at 4000F

 

[INRETECH]

The 4000F @ 2.3V CAP is 3.550 Inch Diameter, and 5.500 Inch tall (not counting terminals

Weight, about 5-6 Lbs

It takes a LONG time to charge it

I once did a quick discharge on the 100F cap (let me try and find the graphs) and it put out about 150A for about 20ms

 

[Floating Spots]

I've looked at these before.
Nesscap has the biggest ones out there.
They have a production 5000F at 2.7 volts.
The storage capacity goes up exponentially with voltage, so 0.4 Volts over teh competition is huge.
There is some excellent info at http://www.nesscap.com/prod/prod.htm

LOOK AT THE RATED CURRENT!!!
WAAHOOO!! 950 Amps discharge from Vrated to 1/2 Vrated for 5 seconds!!!!

Internal Res 0.4m Ohms
Size is ~ 6.5" x 2.4" x 3".
Stores a little over 5 Watt-Hours ~ 18KJ ~ 18KWSecs

I had one quoted a while ago, but it was out of my price range. I'll see if I can dig it up.