Dynacap Double Layer Capacitors.

[X-CalBR8]

Jackpot!!! I finally found them! Here is the link on the Elna homepage for distributors for the Dynacap Double Layer Capacitors and they are cheaper than I expected them to be. http://www.elna-america.com/Distributor.htm#top I searched for nearly 2 days for these things and finally found them. Check out the following site if you are new to Super Capacitors as they describe all the many advantages of using them in place of rechargeable batteries. http://www.chipcenter.com/eexpert/akruger/akruger008.html Elna's Dynacap capacitors have one of the highest charge/discharge ratings of all the other super capacitors that I've checked into at over, an amazing, 1 million cycles! At that rate they should last the rest of my life, and might possibly even outlast the LED that I use with them. I never dreamed when I first started this quest that I would find any power source anywhere that might actually have the potential to outlast a long life LED.

 

[X-CalBR8]

I found out that Elna's Dynacaps go all the way up to 100-Farads! The 100-Farad cap is a bit pricey though so I might have to go with 3 50-Farad caps instead of 3 100's in my solar powered flashlight that I'm building. From what all I've read, all I need to worry about is to make sure that I don't exceed the rated voltage and I should never damage them when charging. So if I were to buy 3 50-Farad capacitors each rated at 2.5 volts, then connect them in series, I should be able to send a total of 7.5 volts to them without harming them, correct? I need a second opinion on this to make sure that I don't make a stupid mistake and blow the new capacitors just as soon as I get them in. I'm planning to connect 2 3-volt solar panels in series, for a total of 6 volts, then send the 6 volts to the 3 2.5-volt capacitors connected in series. The capacitors would need to be connected in series to get it up to where it could tolerate up to 7.5 volts, right? The 6 volts that would be sent by the solar panels to charge the caps would be just perfect for this because the LEDCORP bulb that I already have is the one rated for use at 6 volts. From my calculations, 150 Farads would run at 6 volts 40ma draw for a little over one hour before dropping to 5 volts. So, from this, if I were to use 3 50's I could run a LED at full power for 1 hour or with 3 100's I could run at full power for 2 hours and then have a weaker light for at least another 2 hours. Not too bad for a power source that should last a lifetime and with a fancy power regulator circuit it should be able to run all night long. What other parts should I get to make this work? Do I need to put Diodes anywhere in the circuit and if so, where, and what kind of Diode should I use? All advice, comments, and ideas are very much welcome so please give me some input and opinions on this project. I want to make this project as perfect as I possibly can.

 

[X-CalBR8]

KenBar: You probably got better results when using a higher voltage capacitor because it would have a larger total capacity than a lower voltage capacitor. It's a wonder that you didn't blow the LED when you used such high voltage caps. The thing to remember about dealing with these new "Super Capacitors" though is that they have a capacity so large in comparison to the old kind of capacitor as to make a normal capacitor's storage ability practically insignificant in comparison. With the new super caps you would want to use a voltage very near what you want to run the LED at. Also the new caps are probably more akin to rechargeable batteries in most ways than to an old type capacitor, except of course, that they don't destroy themselves through normal use like rechargeable batteries do. If you get the voltage way too high with the super caps you will probably blow the LED. Another very important thing to remember with the super caps is that if you want to charge them with a hand crank device such as the Russian flashlight you will have to limit the voltage somehow so that it doesn't go over the capacitors rated voltage. From everything that I've read on them so far, they are very tough against most everything except being charged at a voltage higher than what they are rated to take, which they warn not to do. I just don't want you to accidentally blow a pretty pricey capacitor while experimenting.

 

[X-CalBR8]

KenBar: LOL. I know what you mean. I live in the middle of nowhere also and it takes a longer than usual time to get anything in here to.

 

[Chris M.]

if I were to buy 3 50-Farad capacitors each rated at 2.5 volts, then connect them in series, I should be able to send a total of 7.5 volts to them without harming them, correct? .....

.... From my calculations, 150 Farads would run at 6 volts 40ma draw for a little over one hour before dropping to 5 volts. So, from this, if I were to use 3 50's I could run a LED at full power for 1 hour....


If I remember right from the first year of my Uni degree, capacitors behave in the exact opposite way as resistors when connecting them in series or paralell. If you hook 3 identical resistors in series the total resistance is 3x the individual (or just add them up if they are not identical), whereas if you hook 3 identical ones in paralell you get 1/3 the individual resistance. Non-identical values you need to use the product-over-sum formula I forgot about 2 years ago.....

So anyway if my thinking is correct, if you hooked 3 identical capacitors in paralell you would get the opposite effect to resistors- ie a total capacitance 3x the individual, but hook 3 identical ones in series and you would only get 1/3 the individual capacitance in total.
I might be wrong though and I`m sure someone round here knows for sure.

I reckon the voltage handling of such a series connected array might indeed be 3x the voltage handling of the individual ones, but you would lose a lot of capacitance.

If you could get a higher voltage higher-farad-rating rating cap I would reccommend spending the extra to get one. Also it would probably cost less in total and take up less space than 3 smaller ones hooked in paralell.

 

[X-CalBR8]

Chris M.: Thanks for the advice. All advice is much appreciated. My reasoning with connecting the capacitors in series was this. If you connect 3 2.5-volt rechargeable batteries in series you would increase their voltage to 7.5-volts and these new Super Caps act so much like rechargeable batteries that I thought the results might be the same, but like you said, these are capacitors that we are dealing with, so who knows?

So what about it guys? Would capacitors act like batteries when connected in series or not? I'm afraid I have little experience with such things.

 

[Steelwolf]

X-CalBR8:
Thanks for doing so much hunting for us.

Now, from my textbook,

Capacitancs:-
Parallel: The total available capacitance is the sum of the individual capacitance. ie
Ctotal = C1+C1+...+Cn

Series: The total available capacitance is the sum of reciprocals. ie
Ctotal = (C1*C2*...*Cn)/(C1+C2+...+Cn)

Voltage:
Kirchoff's voltage and current laws hold for circuit design.
Thus, for parallel connected capacitors, the available voltage is the lowest voltage drop across the capacitor.
For series connected capacitors, the available voltage is the sum of the individual voltages.

Resistance:
This is response and time dependent. As the capacitor is charged, its resistance increases until it is completely charged. Then it behaves like an open circuit. Resistance at t=0 can be calculated. If I recall my calculations correctly, the resistance behaviour is similar to standard resistors. ie in series R1+R2+Rn, in parallel (R1*R2*Rn)/(R1+R2+Rn)

Hope that makes sense to someone.

 

[X-CalBR8]

Wow! Thanks Steelwolf. That is just the sort of cold hard facts that I was looking for. Thank you very much for taking the time and trouble to look it up for us. Now let me see if I can translate from electronics jargon to plain ol' English with a real life example and let me know if I get it right. Whew, it's been a long time since I got this deep into hardcore Electronics.

O.K. for 3 2.5-volt 100 Farad caps in parallel it would result in 300 total Farads at 2.5-volts, right? For 3 2.5-volt 100 Farad caps in series it would result in Total Capacitance = 100F*100F*100F/100F+100F+100F= 1,000,000/300= 3333.33 Farads at 7.5 volts.

Can this possibly be true!!?? If charged at 6-Volts from solar cells, that would be enough capacitance to run a LED at 40ma for 23.15 hours before dropping 1 volt from 6-Volts to 5-Volts using the formula posted in another thread. "A current of 1 Amp will cause a rate of change of 1V/second in a 1 Farad capacitor." Using this formula, 1000ma/40ma = 25, so with a 40ma draw it would take a 1 Farad capacitor 25 seconds to drop by 1 Volt, so if we have 3333.33 Farads available charged at 6-Volts, it would take 3333.33F*25seconds = 83333.25seconds/60seconds = 1388.89minutes/60minutes = 23.15hours to drop from 6-Volts to 5-Volts.

Please check my math and let me know if I got it all right, because this is just too good to be true. If this is correct, and I didn't make a mistake anywhere, then 3 2.5-Volt 100Farad caps would be more than enough to make the ultimate solar powered flashlight. I could even look into using more than a single LED if this is the case. Please let me know if I did the series capacitance calculations correctly to come up with 3333.33Farads.

 

[Chris M.]

Hmmm......something`s definately not right going on up there in those crazy sums of yours

. No way could you get that much capacitance out of a series array- I`m sure it goes down not up. It`s sort of like you`re getting capacitance from nothing in that calculation

.

But the sums add up- you do indeed get 3333 farads. I think the formula is wrong. I seem to recall there was a reciprocal or 2 in there somewhere....or was there? I`ll have to delve into the archives and see.

 

[snakebite]

decimal point in the wrong place!
33.3333-
i will get some of these to play with.
anyone know their long term charge retention?

 

[X-CalBR8]

kc8adu: The retention rate is well over one month, if I remember correctly. This wouldn't even be an issue if you have solar panels recharging them every day, but in other LED flashlight projects, it might be handy to have such long retention rates.

Chris M.: Please help me out with the equation if you can find anything.

Steelwolf: What about it? Is the equation for series capacitance correct or is there a part that is missing? Something must be wrong, right?

Thanks for all the input guys. I'm enjoying reading each post and am learning a lot.

 

[X-CalBR8]

I just found out that Mouser Electronics carries these also, but they are currently out of stock. This is also the cheapest price that I've found so far. Only $23.48 (18.79 in quantities of 10 or more) for the huge capacity 2.5-volt 100 Farad Dynacap. When they come back in stock, maybe we should get together and order one big order to get better pricing on them.

 

[php_44]

How to correctly use caps in series....

For capacitors in series:

Ctotal= 1 / [ (1/c1)+(1/c2)+(1/c3)...]

I read an application note on these "super caps". Because the capacitance value can vary widely from unit to unit (-20 to +80%) and with temperature - you have to be careful connecting them in series. If two units were near +80% and one near -20% then the latter cap would end up with most of the voltage on it and would either be ruined or suffer shorter life.

They suggest resistors across each cap to equalize charge - but this is no good in a solar or crank flashlight. You could place a zener diode (or series of silicon diodes) across each cap so that if it got charged before the others the diode would shunt away the excess voltage till the others caught up. For a 2.5V cap you'd have to use four regular silicon diodes, for 3.3V (and up) caps you could use a single zener diode.

Hook as many of each of the above circuits in series as you like. Note the direction of the diode connections is different for each circuit. The top circuit is depending on the fact that those diodes dont start to forward conduct until around 0.6V each (4*0.6=2.4V). The bottom circuit uses a zener diode which operates by breaking down at a low reverse voltage - sounds bad but they are designed to work that way continuously. So a 3.3V zener will never allow more than 3.3V reverse bias on it. It will conduct enough current to keep the voltage at 3.3V - keep in mind you have to respect it's wattage rating or it will get hot and possibly burn right out.

Normally - an EE designs circuits with caps allowing a lot of margin. If a circuit operates at a given voltage - we choose caps rated at 150-200% of that voltage for reliability. In this flashlight application you'd be pushing the caps hard - but given the current levels and voltages the caps will not fail explosively

like they might in a TV or other line operated appliance....

 

[X-CalBR8]

php_44: I want to thank you for helping us out with this complex problem. I've been puzzling over it for quite a long time now. O.K., let me see if I got it right using the correct Series Capacitance Formula. Still using the previous example, 3 100Farad 2.5-Volt capacitors in series would come to: Ctotal = 1/ [(1/100F)+ (1/100F)+ (1/100F)] = 33.33Farad @ 7.5-Volts, right? Dang, I hope this isn't the case, because if it is, it is very disappointing.

At 33.33 Farad, it would be about worthless for this project. I guess I'm left with only a couple of good options unless you guys can help out with some cool new ideas that I haven't thought of yet.

1. Go with a parallel setup and use 3-4 caps to get 300-400Farads at 2.5 volts and perhaps use the Satcure circuit to get the voltage I need to run a LED. Of course then I would have to find a solar panel that put out 2-2.5-Volts or try to figure out how to limit my 3-Volt solar panel to put out a half-Volt less. With the Satcure circuit we would almost certainly have a LED that would run all night on 300-400Farads, but would it be bright enough to make a good emergency flashlight? Anyone care to comment that has built the Satcure circuit? How bright is it and does the LED flicker because of the alternating current? How much current does the Satcure circuit typically draw?

2. Try to find a different Super Capacitor that is rated at a higher voltage and then connect them in a parallel configuration without the need to do a step up on the voltage. I'll try to find some other Super caps anyway, but it will be tough to find one that has a high voltage and high capacitance from what I've seen so far. I recall one company that produced a super cap that was rated at 6.3 volts that would be absolutely perfect for this project, but like all the other super caps it was practically impossible to find anyone that sold it retail.

I will keep at it though because I'm very determined to do this project because I have such high hope for it. I really think that this solar light design has the potential to be the best, most reliable, emergency light out there that would have the potential to last through a lifetime of use. This design would also be the safest because, unlike the NightStar (which is a very reliable flashlight design BTW), this design would have no killer magnets to worry about getting near your electronics and magnetic storage media. LOL. My small room is just full of stuff that such a powerful magnet would chew up and spit out. Keep the suggestions and formulas coming in. I'm reading each and every post on this project with very much anticipation. Thanks guys for all the help and contributions so far.

 

[X-CalBR8]

KenBar: LOL. You must be psychic man. I was, only last night, looking at the exact same charger, thinking the exact same thing, that it might be a good solar panel to use on this project. We must have been working on the same project for so long that we are starting to think alike.

Yeah, what I was thinking was this: Many of the different Super Caps run at a voltage of 5.5 volts, so a solar panel that charges at 5 volts would be just perfect! Or, even cooler, maybe we could find one of those solar panels that allow you to change a setting and change the voltage so that it would be more versatile for use in more experiments. Now I just need to find some Super Capacitors for sale somewhere that meet the necessary criteria. From all of the most recent calculations, I don't think we should even consider any caps below 50-Farads because we would have to add so many of them that it wouldn't be cost effective. I know for a fact that some of the companies have caps that go up to something crazy like 3000-Farads! If we could find one of those monstrous caps, it would be more than enough to power the entire flashlight, but I haven't been able to find one for sale yet. If we could find a cap that big, we could do a multiple LED configuration and still have enough light to burn all night long. Let me know if you guys find any large capacity (50Farad or better) super caps for sale that I may have missed.

 

[php_44]

X-CalBR8:

You've got the calculations 100% correct. And, yes, it is dissappointing! However, there is an alternative....

You could use your 100F 2.5V caps in parallel for a 300F (!!) cap, and choose a super bright AMBER LED. Before the white LEDs came along, I used amber LED's extensively - and I still do for two cell applications (see the string on kid's lights). You could also choose red, orange, yellow, and some green - though I find yellow and amber to be the best. These LEDs only require a forward voltage between 1.8-2.2Volts or so. They would be perfect for a 2.5V super cap. Look at this web page for some outstanding LEDs that are easily obtained. Don Klipstein's brightest and most efficient LEDs

By the way - use the circuit I posted a couple of notes back with any voltage solar panel to shunt away excess energy from the super caps. You'd probably want a silicon diode in series with the panel to prevent discharging the super cap when the panel got shaded.

 

[Mike]

I'd recommend the solar battery charger that Kenbar was referencing above. It actually works as claimed. I've bought others that have been dissappointing.

 

[X-CalBR8]

php_44: Thanks for getting back to me and letting me know that my calculations were right on the series capacitors. It's been so long since I had any hard-core Electronics classes that I have to second-guess myself sometimes. I hate that using the caps in series would produce such poor results, but I still have hope of finding one of the higher voltage caps for sale somewhere. Also thanks for the idea to try an amber LED. I will keep that in mind and may do a version like that. Like you said earlier, I would like to try to allow like %150-%200 of Voltage play on the caps anyway to make them last longer. I'm still trying to find some of those super caps rated at 6.3 Volts.

KenBar: Is Ccrane the cheapest place to get that charger that you mentioned? What brand is it so I can run a search on the web for a cheaper price. I've noticed that, even though Ccrane has some of the coolest stuff, they don't usually have the cheapest prices. Also what are the dimensions? Could you take a few minutes and measure it for me? It doesn't have the physical measurements in the Ccrane catalog. Thanks.

Ted the Led: I was planning on a design that would have a jack that would allow you to plug in different power sources to charge the caps. The charging options are almost limitless. Hand power, solar, house current, and car battery are just a few examples. Another kick but thing about the super caps is the fact that if there is enough power available (such as with house current), according to the Dynacap web site, you can charge the caps from 0% to better than 80% in under 2 minutes! LOL, just try that with rechargeable batteries!!!

 

[X-CalBR8]

Here is the URL to order the 50Farad 2.5-Volt Dynacap super cap from Newark Electronics. They are listed at $11.26 apiece there. I, personally, would wait to find a higher voltage cap before making a purchase though. I just thought I would include the link in the thread in case anyone wanted to get one just to play around with.

 

[Mike]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR> KenBar: Is Ccrane the cheapest place to get that charger that you mentioned? <HR></BLOCKQUOTE>

The $15 Ccrane charges was the best deal I could find when I was looking a couple years ago. I believe that included shipping.