Big Joule Theif

[Watson]

Incandescent Lamps

@Kooler

Nice circuit ! keep the good work!

@all

I just finish a new video showing my last JT construction , but also the bulb lighting from the battery the E-C connection and the 3 different secondaries .

The E-C on this one is really strong , but this design is flawed , with parallel layers the load is transparent but not on this one ...

Its a good one anyway.

YouTube - MK1 Micro TP Joule Thief

The best is still to come !

Mark

The video is easy to view, not all jumping around like so many others. Really helps with seeing what the coil looks like.

I'm seeing a lot of posts where the light is an incandescent lamp, not a LED. First time I've seen this used on a JT. A Tungsten filament has a positive temperature coefficient (tempco) so it acts different than a regular resistor. If you put a voltmeter and ammeter on a lamp, the current will not go up linearly with the voltage like it does with a regular resistor..

A LED is nonlinear too, but in the opposite direction. It won't conduct until the voltage gets up to a certain point, 2.5 or more volts for a blue or white LED. Then it starts to conduct heavily. The incandescent lamp is the opposite. At low voltages it will conduct very heavily (low resistance) and as the voltage rises and the filament heats up, the current then rises only slightly, like the resistance is increasing. It is!

This is probably why you never see the lamp used on a JT. IF you want to run a lamp from a low voltage supply, you just use a low voltage lamp. But one resulr\t of the different V-I characteristics of incandescent lamps is that the brightness of the lamp can fool the eye into thinking that the effect is due to the JT making it bright. If you use a 2 AAA cell lamp on a JT with the transistor disconnected, it may look bright just from the 1.5V battery, without the JT adding anything. You think the JT is running but it's not.

The JT likes to have a load that is like a LED, where the load doesn't conduct at low voltages. That allows a lot of feedback and strong oscillation. The opposite is happening with the lamp: the lamp loads the JT at low voltages, and tends to damp oscillation.

If you connect the lamp to a winding that is separate from the primary or feedback winding, then the load doesn't occur until the magnetic field is changing, which to me means that the damping effect is not as great, if at all. But then the circuit is not really a JT, it's a DC to AC converter.

I'm aware that people may be bored and turned off with this theory stuff, but if you are aware of what is going on in the circuit, it will help you immensely when you experiment and allow you to get the most out of your project.

Back to experimenting...

[Watson]

hey guys
this is a update on my backwards joule theif.. inductor installed..
found that even since the transistor is backwards it still has bemf in the same direction..
use the secondary for leds..
check volts on battery after hours of running..

Nice drawing, a bit on the small side, but really good. What did you use to draw it?

It's a good idea to put a small resistor like 470 ohms in series with the pot. This prevents excessive current when the pot is turned down to its minimum, and the full battery voltage is put on the base of the transistor. Without the resistor, you can connect the battery directly to the base, and the excessive current can damage the transistor.
Thanks.

[Watson]

hey guys
this is a update on my backwards joule theif.. inductor installed..
found that even since the transistor is backwards it still has bemf in the same direction..
use the secondary for leds..
check volts on battery after hours of running..

I looked at the circuit closer and I now see what's going on.

It's like a regular JT. But the feedback winding is connected to common instead of the positive. If you connected the feedback winding to the base with a resistor, like a normal JT, there would be no positive voltage to bias the base. So you had to add the DC blocking capacitor and the pot to the positive to supply the base bias.

I'm not sure why the 100 mH choke was put in there. What does the plus sign at the end mean?

THanks..

[kooler]

watson
the plus sign was a short way for me to say ( or greater than )
the inductor gives you better adjustment.. easier to fine tune the circuit thru the pot.. works great with 100 uh to 220uh
i used tinycad and ms paint to do the schematic.. it got smaller once i added it as a attachment to the website..
if need be .. i can post it on imageshack if you want

let me know

robbie

[Watson]

watson
the plus sign was a short way for me to say ( or greater than )
the inductor gives you better adjustment.. easier to fine tune the circuit thru the pot.. works great with 100 uh to 220uh
i used tinycad and ms paint to do the schematic.. it got smaller once i added it as a attachment to the website..
if need be .. i can post it on imageshack if you want

let me know

robbie

Above you say 100 uH. It looked to me like the schematic was marked 100 mH, but it's too blurred to read. Do you mean that the value in the schematic is 100 uH? It would make more sense than 100 mH, but it looked like an 'm' to me.

Thanks

[kooler]

100 uh

watson
that was indeed a 100 uh inductor but you don't need it if you have a real good pot..
my little cheapo pots are funny acting.. so i went with a better made one and i didn't need it to fine tune it no more..there is a real good schematic on ou site
let me know and i will upload it to a image site

robbie

[kooler]

backwards jt 2

this is my second video of the backwards jt with a 265,000 mcd led on the battery rail..
just needed to try out a new camera

YouTube - koolerization's Channel

fun circuit since it runs xtra leds with out using anymore ma's

robbie

[kooler]

induct - 0

i made a new circuit
video is here
YouTube - kooler's induct-O

some of you may find it interesting

robbie

[Watson]

My Version

watson
that was indeed a 100 uh inductor but you don't need it if you have a real good pot..
my little cheapo pots are funny acting.. so i went with a better made one and i didn't need it to fine tune it no more..there is a real good schematic on ou site
let me know and i will upload it to a image site

robbie

I put it together on a proto board, but I had a hard time getting anything but a sneeze out of it. Pics and description are here.
Watson's eBlog

[mk1]

@all

I made a new video YouTube - MK1 Slim TPJT (MOTOR)

Running the motor/generator , detail on the connections ...

From my jt coil.

[kooler]

I put it together on a proto board, but I had a hard time getting anything but a sneeze out of it. Pics and description are here.
Watson's eBlog

sorry i just noticed i put the leds in series .. they need to be in parallel..
see if that don't help..
sorry

robbie

[Watson]

@all

I made a new video YouTube - MK1 Slim TPJT (MOTOR)

Running the motor/generator , detail on the connections ...

From my jt coil.

Your 1k pot needs a 100 ohm resistor between the center wiper pin to the proto board. The resistor protects the transistor from excessive current when the pot is turned to zero ohms, which connects the battery directly across the transistor.

I've tried running a JT with the 3055, thinking that more power meant more current. But it was barely able to light the LED. The power transistors were not made for low voltages of only a volt and a half, so they conduct little current at that low voltage. I started using the 2SD965 transistors from the flash circuits of disposable cameras, then I found the NTE11 at Fry's electronics, but those were too expensive. I ordered a bunch of KSD5041 transistors from Fairchild Semi (or Mouser, I forget), and they work so much better than 3055s that I never use those power transistors any more.

Also, the motors make good generators, just connect a LED up and give it a spin. Have fun!

[Watson]

LEDs In Parallel

sorry i just noticed i put the leds in series .. they need to be in parallel..
see if that don't help..
sorry

robbie

At first I could barely get one LED to light :-P

LEDs in parallel are okay if they're driven from a JT, as long as they're all the same coloer, and preferably from the same maker. So far, you haven't said what color the LEDs were, but I saw white ones in the videos.

I bought a bunch of blue LEDs on ebay, but the seller sent me a bunch of factory rejects with bubbles in every LED. I got them to replace the LEDs with bubble free ones, so I've been trying to use up all those bubbly blue LEDs. I'll try a few in parallel. But I got some 5-chip white 10mm LEDs from Hong Kong on ebay, and they can handle a couple hundred mA, because they're 5 LED chips in each one. Nice and bright too (for a thousand hours or so, then they go dim). I'll have to try one and see.

[Watson]

sorry i just noticed i put the leds in series .. they need to be in parallel..
see if that don't help..
sorry

robbie

I changed the capacitor to a smaller value and it got ten times brighter. It's up to the point where it's putting out some seriously bright light. See the end of my blog below for the details.

Watson's eBlog

[Watson]

tiny orange toroid light

This time I checked to make sure, and it is what I wanted to present.

The reason for this video is because I can see that the toroid is the most limiting factor for making a joule thief. I only realized this when the stock rooms of allelectronics and mouser emptied of the toroids we mentioned in a day.

So, I took the circuit from the 10w cfl that I had removed, and the screw base too.
I could have used the transistor and resistor from the circuit too, but I didn't think about that until later. It was the toroid I was concerned about.

---
yes a solar panel does this, but in the day this light is not needed, so a storage battery of some kind is necessary.
There are videos out there using batteries of aluminum foil and copper, or carbon. I lit a led with 2 of these in series. I didn't try the jtc with that but it would surely work with only 1 instead of 2 in series to some degree.
So, the focus is really on making a bright light with that toroid from the cfl.

And, please keep pushing this envelope, everybody! And, please show others what you have made.

jeanna

edit
let me add the LoA I get from wall mart. I am sure they are in other places. (My hardware store carries some but not all of LoA bulbs.)
They are made to resemble a standard light bulb and have 20 leds in series and a whole lot of protective caps and diodes and resistors and even a fuse, so while I have made and shown(on youtube) how I made a modified LoA bulb into a battery joule thief lamp, I wanted a simpler design.
this eventually turned into the jtpu = joule thief power unit, which is a stand alone joule thief in a switched plug socket which lights one or 2 unmodified LoA bulbs and is entirely portable.

I have nothing but praise for those who try to do their best to improve the world, rather than try to pull electricity out of thin air. It's a noble cause.

One thing that's a big problem with solar cells is a panel that can put out enough to charge a battery for a decent amount of evening light is so expensive that even those in the richest countries can't afford to buy one. I think a much easier way to generate enough to charge a battery is to use a standard dynamo like those found on bicycles to light the headlight and tail light. This dynamo puts out 6 volts at a half amp or so, which is easy for people to power with a crank. Better yet, more than one of these dynamos to develop even more power to charge the battery, and it would be powered by pedals or mounted on a bike wheel which is on a stationary stand.

Magnets and some wire are far cheaper than a solar cell. And it's possible to generate up to a dozen or two dozen watts from leg power. The nice thing about this is that the power is not dependent on the whims of the weather and time of day, which can put solar cells out of commission. Anyone, any time can crank a generator and power the lights and/or charge the batteries.