This is the first time I have been able to keep a Joule Thief type device running a CFL on just an electrolytic capacitor.

Video: YouTube - Joule Ringer! Keeps running with battery disconnected.

Lasersaber,

What an excellent experiment you have there. I can't wait for your instruction video!

Lasersaber,

That is amazing. What is the electrolytic capacitance?

Lasersaber,

A possible explanation for the poor performance of your supercap is that it has a high value inner resistance, ESR (Equivalent Series Resistance), compared to any good normal electrolytic capacitor you seem to use. Especially high series resistance is "embedded" in the so called memory back-up supercaps where the load current is but a few microamper or even less and the voltage drop hence dissipation across the high ESR is a very low value. Possibly you current consumption is higher than a few microamper...
Another possible explanation is that your supercap you showed in the video has a deteriorated capacitance value, maybe by abused use. They are very sensitive to higher than specified working voltages exposed to.

Could you show a schematic?

Thanks, Gyula

I have over 20 of those little super caps and they all show the same effect. If I connect over three of them in parrel then they start driving the circut.

I am planning on doing that soon. I will be building another unit soon and will making the schematic at that time.

Lasersaber, You have very well made descriptive videos of all your projects. On
this last one, are the capacitors in series or parallel ?



FRC

The ones in the video were 10000uF 10V. I will attach the specs. I should make a video showing runtimes with different size capacitors. I have tried many different types of electrolytic capacitors and they all seemed work about the same.

Attached Images

e-rs.pdf (47.8 KB, 632 views)

Well, then this means that high inner resistance is involved, so paralleling them reduces it. A supercap manufactured for memory back-up usage has any ESR between 40-100 Ohm, an unusually high value one would not think of.

Gyula

They are in parallel.

kool i never thought to try ferrite..
but i had a mini tesla that emdevice posted a diagram of a long time ago made up the same way with a feedback coil to the input but it would only running for 35-45 seconds with no battery.. i always felt like i was pulling to many ma's to get it to run longer..
so you have showed me that i was wrong..
very well done .. sir

robbie

Very well done Lasersaber!

Out of curiosity, what is your ''record duration'' for no batteries?

Have you tried this with an exciter, or is it only suitable for the joule thief?

Thanks for sharing

I have not really tried to do a record duration. I did do a test duration run once with a fluorescent tube and my 10 volt caps filled to 8 volts. The tube stayed lit for 48 minutes. The last 20 minutes the circuit was still ringing but the tube was lit very dimly.

This certainly looks promising, but could it not be a sort of weird "capacitor relaxation circuit" of a kind? hopefully its something more exotic than that.

my thinking is that with 6 or 7 x 10,000uF caps in parallel, thats a very large capacitance capable of storing a lot of energy and which via the right circuit could be made to light a cfl for a long time without anything "exotic" going on.

Hopefully its not thw above and we can look forwards to some interesting developements

The more you parallel the less resistance as you say. It makes perfect sense. The problem is why use super caps when the normal electrolytic caps work so well. Cost wise it makes better sense to use the normal caps.

@ LaserSaber... When are you gonna share your circuit with us.. I am so ready to replicate this lol..

You may be right, but in my experience prior this I have never been able find the "right" circuit that could light a CFL like this. In fact I could never get a CFL to light for more than a couple seconds on one 10,000uf 10v capacitor. With this circuit I can easily light one for over 1 minute on a 10V 10,000uf cap.

Try it. You will see it is not so easy to light a CFL for 20 to 40 minutes from that much capacitance. At least for me it isn't.

Hear hear!

can anyone else provide a circuit that keeps a CFL lit for 20 mins from just an electrolytic cap? (not even a super cap!)

I sure cant! I tried my exciter on caps and got nowhere.

Jonnydavro? Lidmotor? anyone else who manages to get things to work?......

I dunno about you Xee, but im looking forward to schematics from lasersaber. And I know, the guy just cant resist sharing

Merry Xmas and New Year to everyone!!!!

Laser, could you try replacing the CFL with capacitor of an equal capacitance to the input and see how much gets transferred?

Hi Jbigness5,

I underlined in your text I want to answer. there are several types of supercaps manufactured for different purposes. Memory backup types has an inherent high inner resistance because the current draw demanded by modern memory chips is in the order of a few microAmper (say 3-10uA) or lower, hence the inner voltage drop of the supercap across their 50-100 OHm resistance is negligible. See such memory backup type data sheets in Page 43, at the bottom, for instance http://www.panasonic.com/industrial/...ide_052505.pdf

But other types have an inner resistance of under 0.1 Ohm or less like in case of a good rechargable battery and several hundred Amper may flow when you short the pins of a charged up supercap with a heavy wire piece.

http://www.avx.com/docs/Catalogs/bestcap.pdf

So make sure what purpose a supercap is manufactured for and choose accordingly.

rgds, Gyula

Still the question is cost and it seems the capability of the cheaper electrolytics are better suited to this application. Even with one cap he gets it to work but it takes three super caps to start the process. Cost wise this would make the electrolytics a better choice for this application, don't you think?

i think the main reason that super caps don't work as well for this application is the fact that they have such a high... well capacity. as you move up in capacity the ability to catch that low amperage flyback or ring is diminished. I think the super caps discharge so quickly and have such a little ability to catch that ring that it acts just like a battery in the circuit with a straight discharge, Cause super caps aren't the best at high speed operation and i would imagine even worse at charging and discharging high speed.

I found a 10,000 uF capacitor and did a test with my most efficient circuit for lighting fluorescent tubes (able to light tube with 9 mA at 1.5 volts). With the capacitor charged to 6 volts it would only light tube for about 5 seconds.

Dear Jbignes5,

Yes, cost wise I fully agree with you, though a supercap with high inner resistance like 75 Ohm is comparable in price to a 10000uF normal electrolytic cap, you have to buy several such supercaps and connect them in parallel to get similar performance as lasersaber showed it, this boils down to: supercap is more expensive.

Here is some 10000uF (0.01F) 6.3V normal electrolytic cap types and pricing at mouser:
Aluminum Electrolytic Capacitors - Leaded

And here is some supercap with innere resistances between 50 to 100 Ohm, 5.5V and with cap values between 0.1 to 1F:
Supercapacitors

However, supercaps in the range from 0.06F to 0.2F with comparable inner resistance to a 10000uF normal electrolytic cap are much more expensive, especially for 5.5V voltage: Supercapacitors

(Of course pricing may differ at another component seller, pro and con, tendency remains: supercaps are more expensive than electrolytic caps.)

All I wanted to communicate is that there are big differences in the inner resistances of supercaps, depending on what purpose they are manufactured for and what type you happen to have and then make performance comparison to electrolytic caps.

Respectfully
Gyula

I see your points and they agree with mine. For now for simplicity sake we should just use the electrolytic caps unless there is a reason for using super caps?

I am wondering how Ultra caps would work in this.

Well, my understanding on this video so far has been that for lasersaber has no reason for using his present supercaps, unless he obtains a type with 0.1 Ohm or less internal resistance for 5.5V working voltage. He was unlucky to have types made for memory back up purposes, had he had just a better one I showed in the earlier mouser link, he would not have concluded his finding a much less working time with supercap as bizarr. With a 0.1 Ohm or better inner resistance supercap his circuit can surely work much longer than with the 8 electrolytic caps now (provided this supercap had a capacitance of say 0.47F or higher at 5.5V).

Yeah it's not something they tell you out right. But it does work and he said that it did it just took 3 to get it to start working. But since his works sooo well already if we could get it to extend the running time to say days that would make the super caps run for a week I bet.

Baby steps... I guess... Dang Laser when you gonna show us something to replicate.. I'm burning to do this...

I also go myself a 1.5 watt fluorescent light to mess with in the meantime. It's so tiny..

I have now been able to make successful replications of my Joule Ringer circuit! I reduced everything to the minimum number of parts needed. I found out the the number 2 coil can be an air core coil. I will post a DIY video with lots of new details and schematic later tonight after I get back from Church.

Attached Images

JouleRinger.jpg (83.8 KB, 322 views)

Nice I have been doing my own experiments with ringing coils and caps with limited success. I am eagerly waiting your circuit and diy video. Thanks for all the hard work my friend.

I'm literally on the edge of my seat!!!

I spent many fruitless hours last night trying to get a capacitor to keep a CFL lit for more than a second on my exciter, but no luck. This is a really exciting development and i congratulate and thank you for letting us all know about it.

happy new year!

@ Lasersaber:
Have you tried to use Leds with this circuit, as it would seam more practical and more efficient than using the gutted CFL? And, do you think that a modified Joule Thief circuit can be made to work instead of the coil set up that you have?
I can see similar result when using the parts mentioned above, connected to a 10 capacitor Captret. Without the Jt in the circuit the Leds light for only a couple of seconds, (without a battery), with the Jt it last a few seconds longer, but still not like your set up. It does show some similarities, just doesn't have much run time.
I've seen that in your video you are tapping the batteries to the circuit, several times during the short video. If the device lasts for as long as 40 minutes, or longer, why is this necessary, or is it?
Thank you for showing your new circuit. It looks great. We are all waiting to replicate it...
Thanks again,
NickZ