first joule thief :p

here my first try at a joule thief

it seems pretty good, i started with and alkaline at 0.9V and drained it during the last week, to 0.48V ...
this seems to be the best it can do ...
**broken link removed**

i used circuit B from
http://talkingelectronics.com/projects/LEDTorchCircuits/LEDTorchCircuits-P1.html
with modification on the toroid, mine has a lot less winding than the original circuit ...

i really need a scope to see what is happening and to try few tweaks.

is there a way to drain more juice from those old batteries ?.
i guess some kind of multistage with some caps (maybe mixed with a miller engine)
of course it will not light a led but it might be good to pulse it from time to time...

.48 is good if you ask me, that's bellow a common diode drop. The cell is more than dead, the last dredges aren't worth trying to recover. Oft times recovering the first dredges aren't worth it.. But fun to see! The amount of energy required to light a red diode so you can see it in low lighting are incredibly low, the amount required to do something more useful goes up very quickly.

You'd need a solid performing scope to investigate a Joule Thief, their frequency is relatively low often in the audio to 200khz range, but the rise and fall times are not sine waves so you need a scope with a good 10-20mhz to really get a good look at them functioning.

New application?

This is very interesting. I'd like to discuss this circuit from a different perspective, desulphating lead acid batteries.

It is claimed by some, including by commercial manufacturers, that pulse charging can reverse desulphation. I have tried commercial units and there does seem to be some improvements in the batteries.


Anyhow this flyback design implies that with a suitably beefy inductor (using a cheap 12VDC PC pwr supply for the basic safe supply voltage) and a large (22000Uf 50V) low esr capacitor I can build up a 40V+ charge on the capacitor and then fire it off thru a bank of zener triggered ( @ 40V+) hi pulse current, low resistance NFETS.

Perhaps these FETS:
https://www.electro-tech-online.com/custompdfs/2012/03/irlb3036pbf.pdf
and this cap:
https://www.newark.com/cornell-dubilier/380lx223m050a052/capacitor-alum-elec-22000uf-50v/dp/12C7441
This Zener:
https://www.newark.com/vishay-semiconductor/1n5260b-tap/zener-diode-500mw-43v-do-35/dp/18M3571



Lead acid auto batts can have internal resistances in the .01 to .03 ohm range. Bigger marine batts can be in the .01ohm or even less range.

A sulfated batt may be in the .1 ohm range.

The charge pulse resistance circuit is then:
Cap ESR + FET ON resistance( 4 in parallel) + Batt resistance + Wiring Resistance.

Perhaps:
.02+ .002/4 + .1 + .02 (est.) = .1405 Ohms.
A discharged batt V is 11.7,
https://en.wikipedia.org/wiki/Lead–acid_battery

Ohms law => I = (43 - 11.7)/0.1405
= 222Amps
A charged battery at .03ohm will incur a pulse current of: (43-14) / .0705 = 411 Amps.

The unknown at this time will be how fast the cap will charge to 43V based on the flyback efficiency and current drive..

Not having wound inductors I could use advice here. I have a couple 800W microwave transformers available which I could use, or rewind. Plus I have a few decent 5A toroidal inductors as well as toroids harvested from PC pwr supplies.

I'd hope to achieve 100Hz or so. max as this would deliver max power and not cause significant battery reactance to get involved.

thanks

Batteries are so cheap and so important that it is not worth taking the risk to deal with a battery that has any sort of problems.
As soon as my car battery fails, I replace it.
I tried nursing a battery and got caught. It totally failed. I will never take the risk again. The cost of getting a lift from the car and back again was 5 times more than the cost of the battery. And hours of wasted time.

I am going to put that Joule Thief pic on my website. We get a lot of feedback from readers who have buit the circuits but few get around to actually photographing their project.

Mosaic.... This has been gone over endlessly on forums both here and on the web in general. There is no way to desulfate a battery fully. You can jog it back to life but this is BRIEF and the subsequent failure is absolute and irreversibly worse over time, the entire battery would have to be rendered down into it's constitute components and reconstructed, which is what they already do!

If you look at the chemistry involved in sulfation that eventually kills all Lead Acid batteries can't be reversed electrochemically.

The only possible sollution to avoiding sulfation of Lead Acid cells is to never use them and keep them at a universally perfect float charge forever. This is obviously not practical. Any use of a lead acid battery over time regardless of any methods to prevent it will eventually end in sulfation failure. It is a fact of the chemistry.

I am going to put that Joule Thief pic on my website. We get a lot of feedback from readers who have buit the circuits but few get around to actually photographing their project.

Click to expand...

colin!!, why didn't you post mine when I made it 3 years ago!
As I recall it may be perhaps I disagreed rather strongly with some modifications to the simplest joule thief circuit for efficiency that didn't have substantiated support data about how the power in/out was measured. These are very simple devices the waveforms they produce however are not.

The original toroid I tried is laying unplugged on the right, I found that I could probably have done far better if I could find smaller toroids, I even had a few promising potted inductors I'd pulled out of old boards but I didn't have anything smaller than 30 gauge wire for wrapping so I didn't go any further with it.

To do any better on a hobby level would require some serious investment in time and learning how to physically construct things that small being the biggest problem.


Voidptr, if you read the link I posted above about the Joule thief I made back then (it's all coming back to me now) You'll see that I managed .4volts, your .48 volts is so close it's not worth mentioning probably a slight difference in the transistor itself or the closer winding on the smaller toroid I used. You could probably seriously increase the efficiency of your circuit if you rewind the toroid and spend a lot of very careful time getting that winding as close together as possible, even the smallest gap decreases inductance a lot. Don't worry about covering the entire diameter of the toroid either, the closeness of the winding's is far more important, overlapping the inner coil is fine just don't go backwards, if you pick the right toroid inner outer diameter you could fully wrap the outside of the coil while stacking the inside of the coil, but... you gotta want to really squeeze every last bit out of the components to got that far for efficiency and size/weight.

Sceadwian said:

.48 is good if you ask me, that's bellow a common diode drop. The cell is more than dead, the last dredges aren't worth trying to recover. Oft times recovering the first dredges aren't worth it.. But fun to see! The amount of energy required to light a red diode so you can see it in low lighting are incredibly low, the amount required to do something more useful goes up very quickly.

You'd need a solid performing scope to investigate a Joule Thief, their frequency is relatively low often in the audio to 200khz range, but the rise and fall times are not sine waves so you need a scope with a good 10-20mhz to really get a good look at them functioning.

Click to expand...

i do it for the freedom of all electrons !!!
those little guys were stuck in batteries and need to finish their "life" free (in a way of speaking) !!!
and... freeing them can be fun for me

i got a box full of those 1V "dead" batteries that are to dead for other electronic devices who cannot accept rechargeable batteries... so ... i need more experimentation time

yes .48V seems not bad at all for my first try (and for my first toroid/winding) but maybe there is ways to free more of those little guys

yes im currently thinking to buy a cheap scope maybe 50Mhz or 100MHz so i can try to see what im doing in my silly hobby
of course i'll still be a noob but i'll be a noob with some tools and flashing leds !

i do it for the freedom of all electrons

Click to expand...

You made me laugh good saying that! and I agree with you!

I have more than a box, I have an entire drawer that's getting on into the 30lb range, a lot of them probably don't even still hold a charge I just bag them for eventual recycling. I've thought about it a bit and I think the best method to approach this from would be to feed an active circuit from many cells with mosfets doing the switching. Many cells in series pulsed for a short duration with caps to buffer the actual cell, as at that discharge point their horribly high resistance. The active circuit itself can be boot strapped from the cells, you could bypass each cell section with a schotky diode to allow higher charged cells to bypass the series Resistance adjacent cells. But again... This is getting into weird territory and cost/effect and rationality don't really add up.

Would be neat to see though.

If you can get a 50-100mhz scope it would be good. I had a 500khz scope two years ago that I miss desperately now, but I can't afford even the thought of a cheap scope right now.

20mhz will get you something passable, 50-100 will get you into the more usable range for a good scope, but keep an eye out for the sensitivity too, the 500khz scope I had wasn't anything to speak of but it was sensative down to 10mv a division within it's full range.

Mosaic said:

This is very interesting. I'd like to discuss this circuit from a different perspective, desulphating lead acid batteries.

It is claimed by some, including by commercial manufacturers, that pulse charging can reverse desulphation. I have tried commercial units and there does seem to be some improvements in the batteries.

ZAP
thanks

Click to expand...

i really to noob in chemistry to give you advice on that side of your project.

i can only say this has a noob ...
if your goal is to study the chemistry process and you are a noob like me in power electricity / electronic there is probably some lot faster and safer way to build your pulser ....

take per example a cheap 120v flux core welder, it will give you i guess tension and a lot of amp you need, i guess you can even use it like that or used the same approach, big bad transformer to lower 120v and output more A or there new inverter approach ...

having your pulser near already working can help you a lot in your project !

**broken link removed**

Excellent link voidptr, goes to show you that even though most hobbyists and many persons that have gone through professional engineering schools learn and know the details of the electrical side of things, that it is not always easy to come up with an efficient solution to a specific problem.

colin55 said:

I am going to put that Joule Thief pic on my website. We get a lot of feedback from readers who have buit the circuits but few get around to actually photographing their project.

Click to expand...

@colin55
i didn't know that you were the owner of talkingelectronics.com
thanks for sharing
if you want a better pic, i can give you one,
on this one we merely see the resistor but it has the advantage to blur out my noob winding technic !
on that subject i should say that i only put total 40 turns but i kept the same ratio as on your website.

i had to review my old electromagnetic notions before attempting that coil. i was also a bit lazy so indeed the wire should be in better contact with the core, next one will be better i hope.

@Sceadwian
we will have to have also switching bank of batteries to put them on : On or Pause
i did few full discharge tests of alkaline and found that when they are nearly dead, they can regain a bit of charge when they are left unplugged for a while, the gain of current doesn't last very long tho before returning to previous very low current ...


i didn't search or read on the JT subject on this forum before, it seems it was a beaten battle subject in many ways :-/

i just want to add that ... from my noob point of view i can learn so many things just playing around with things , tools , technics and processes
maybe i will not build a better JT but i will learn few things about those little electron guys behavior...

Sceadwian said:

Mosaic.... This has been gone over endlessly on forums both here and on the web in general. There is no way to desulfate a battery fully. You can jog it back to life but this is BRIEF and the subsequent failure is absolute and irreversibly worse over time, the entire battery would have to be rendered down into it's constitute components and reconstructed, which is what they already do!

If you look at the chemistry involved in sulfation that eventually kills all Lead Acid batteries can't be reversed electrochemically.

The only possible sollution to avoiding sulfation of Lead Acid cells is to never use them and keep them at a universally perfect float charge forever. This is obviously not practical. Any use of a lead acid battery over time regardless of any methods to prevent it will eventually end in sulfation failure. It is a fact of the chemistry.

Click to expand...

While I appreciate your advice, what cannot be done today and is impossible and is a 'fact' , often becomes be quite doable and run of the mill with a bit if ingenuity.

Human history is a case study of 'facts' becoming fallacies. Therefore I intend to try for myself. I am quite certain that your abilities and mine are different and whatever u think is not what I think. Therefore what I try to do may very well yield different results from other attempts. Often times dogma and doctrine of current science understanding is the anchor that limits us.

Just 20 years ago 'Dark matter' did not exist and astrophysics had a different belief structure. Perhaps 20 years from now Lead acid battery technology may be able to reverse sulphation.

Hi,

I built a "desulfator" as an experiment a while back, not too long ago. It seemed to help a little but nothing good enough to make me think that it was going to restore the battery to some sort of new like state. I think all it really did was charge the battery a little more.
If it worked i would still be using it today, but im not.

As far as the original topic, i've built a "Joule Thief" as a DC to DC converter rather than LED driver in the past using a transformer with multiple windings. I needed a low power converter that would be able to supply a DC voltage to a circuit that had to be isolated from the main power source.

What kind of frequency and power per pulse did u deliver to the battery?

colin!!, why didn't you post mine when I made it 3 years ago!

Click to expand...

Your photo had extra components and it would have confused the issue for those beginners making the project. Things have to be simple and easy to understand. The project is already a very sophisticated and complex design.
I don't see the capacitor in your photo and I explained how it improved the efficiency of the circuit considerably.

Mosaic said:

What kind of frequency and power per pulse did u deliver to the battery?

Click to expand...

Hi,

Sorry i dont remember much now, but i did try different pulse widths too.

Well I hope to give it a try. Cadex is a relatively well respected name and they describe how sulfation may be reversed.
**broken link removed**

Sceadwian said:

Excellent link voidptr, goes to show you that even though most hobbyists and many persons that have gone through professional engineering schools learn and know the details of the electrical side of things, that it is not always easy to come up with an efficient solution to a specific problem.

Click to expand...

SC you will observe that this suggestion overlooks the matter of sub 100Hz pulsing. Thus it is not at all a solution for the subject at hand.

While I appreciate your advice, what cannot be done today and is impossible and is a 'fact' , often becomes be quite doable and run of the mill with a bit if ingenuity.

Click to expand...

I'm sorry but all this post does it to show that you know nothing of the chemistry involved and the realities of what occurs the the materials that make up a lead acid battery age over time.

Desulfation can help something like a marine battery that is deeply discharged or cells that's been left idle for too long and have completely drained, if a desulfation cycle is run before it's recharged it will help prolong usable life and can allow charging of cells that have been sitting idle far too long, but it can not infinitely extend life, and lead acid packs that are used in say automobile applications were they're kept at a proper float charge and used for starting or just power buffering will eventually die from the lead paste inside the plates becoming inert over time and desulfation can not fix this. Even deep cycle batteries will eventually die and desulfation again can not fix it because it's not sulfation in the end that causes the battery to fail it's passivization of the active lead compound that lead acid cells require to work.

If given those two possible uses you still want to build one then simple Google a bit I'm sure you can find recommended current pulse trains, it's not really relevent to a Joule thief however.