Big Joule Theif

[Lidmotor]

Lidmotor using wrong 555

@StevanC:
Don`t forget that the strength of this circuit is its energy recovery possibilities.

He could try to run the 555 off a supercap charged up by his peltier or BEMF during operation, then the 555 power consumption does not matter.

I am using 555s for Joule thief circuits that have a very Low Power Consumption
1 mW Typ at VDD = 5 V and Single-Supply Operation From 2 V to 15 V
If anyone cares (maybe Lidmotor), it`s Texas Instruments Model TLC555 (NOT NE 555 !)
We are talking about an IDD Supply current of 250 MICRO-Amps (!).
The peltier might be able to reach that current value alone.

Great finding Lidmotor and thanks to Slayer007 for the basic circuit !

@ Everyone-----I am using a NE555 chip in the video because I had one on hand but it is the wrong one to use here. TLC555 is the correct one and Paul is using that one already. You are right about that power savings being eaten up by the timer circuit if it uses too much to turn the circuit on and off. When I get a TLC555 I will redo the circuit. The charging part of course is diminished by the reduced on time but the hope is that the total energy consumed will be less. Tuning is a real issue to get the most light for the least power. Get it wrong and you defeat the purpose.

@Jesus----Slayer has posted the circuit I am using already a ways back on this tread. The 555 timer part is not refined yet and is not posted. Paul (kubikop) is the person working the hardest on this and you can follow his progress on Youtube.

Lidmotor

[Xenomorph]

@StevanC:

You have found a way to simulate a joule thief in SPICE ???
Then i DO care about it
But maybe that`s not what you have meant.
I think the BEMF would be disregarded by Spice as much as i have experimented with transformers
in spice.
And congrats for your replication!
For some reason my secondary voltage is way too low as it should be with 440 windings, i will
try a different core and different wire now.

[slayer007]

I made it finally!

~19 turns 0.7mm
~20 turns 0.65mm
~370 turns 0.3 mm

on a PSU ferrite transformer (salvaged)

It runs off a D sized 1.2V accu

I used MPSA06 over a TIP35C (gets warm on work)

I shunted the TIP35C's collector to emitter with a small neon bulb

I shunted the MPSA06's emitter to base with a 1N4448 diode

1K resistor to MPS's base

(+) to common primary & triger

(-) to TIP's emitter leg

and 'buz'

the I soldered a neon light (20cm and 30cm long ones) to the tertiary (370 turn) winding

It shines - the smaller is brighter :-)

it works (!) off a 1.2 battery

the spike on the MPS is about 50V (C-E)

over the TIP its about 80V

and over the neon rod it's about 600V

while drawing (RMS) 537 ... 500 mA off said cell

now to see:
1. how many can we use and maintain this glow?
2. is it more efficient to use LEDs instead neon light?
3. Can we make it more bright and still lower the cost?

Slayer007:

nice project

anyone cares for spice model?

Congratulations SteveC

It sounds like your having very good results.

Yes LED's are a LOT more efficient.
You can light quite a few led's off your setup.

When you first try it have your JT turned all the way down.
Then turn it up slowly till your lights get as bright as you want them.

That way you wont burn out your led's.

I had mine lighting ten led's turned all the way down and it ran for three days off one AA.
It was using under 10Ma and the voltage in the battery only went down from 1.27 to 1.23 after running for three days.

The lights were very dimm because it was down all the way.
But it would still light the room up at night enought to see where you was going.

I would allso recommend doing as Lidmotor and Kubikop are doing and use the 555 with it.
That would make it a lot more efficient

[Jesus M NIeves Oliveras]

...
@Jesus----Slayer has posted the circuit I am using already a ways back on this tread. The 555 timer part is not refined yet and is not posted. Paul (kubikop) is the person working the hardest on this and you can follow his progress on Youtube.

Lidmotor

Thank you @lidmotor !
He posted that circuit at overumity forum.

Jesus

[slayer007]

Thank you @lidmotor !
He posted that circuit at overumity forum.

Jesus

There is allso one on page six post #123

[Jesus M NIeves Oliveras]

There is allso one on page six post #123

Thank you @slayer007.

Maybe you meant #154 on the page 6.

http://www.energeticforum.com/renewa...html#post45847

If it is that one, you posted it already on jule thief thread at OU forum.

Jesus

[Mark]

Hi Guys, I just spent about 1 minute putting together a pulse charger that charges a cap to 300 volts in about 30 seconds with a 1.5 volt battery. I'm currently charging a 12 volt car battery with it. I'll let you know what happens. You'll LOL when I tell you how I made it. And it was totally free!

[slayer007]

Hi Guys, I just spent about 1 minute putting together a pulse charger that charges a cap to 300 volts in about 30 seconds with a 1.5 volt battery. I'm currently charging a 12 volt car battery with it. I'll let you know what happens. You'll LOL when I tell you how I made it. And it was totally free!

Sounds very good Mark.
Please keep us posted.

[StevanC]

SPICE it up ;-)

@StevanC:

You have found a way to simulate a joule thief in SPICE ???
Then i DO care about it

Actually i didn't - the key is to compare SPICE wise prediction to what actually get's measured by a scope/DVM/whatever
;-)
It's the difference that's of interest IMHO here:
The SPICE model has T1 [coils here] 1 <--coupling factor is 1 = 100%

but we expect to get more than 1 here: aren't we?

But maybe that`s not what you have meant.
I think the BEMF would be disregarded by Spice as much as i have experimented with transformers
in spice.

It's not at least not in swCADIII i have (shall i upload?) There is a nice back-pop i tend not call "backEMF"

backEMF is INSTANT IMO - like the back pressure of a wall You push Your finger against

follow me here?

The back-pop of a coil is different:

Like, after You let go the wall, he splasher You

*after*
;-)

And congrats for your replication!

For some reason my secondary voltage is way too low as it should be with 440 windings

or is it?

I have trouble to be satisfied with the JT I built:

I understand the back-kick is the power winding harvesting the field:

1. The more copper we have on power winding - the more yield in the back-kick harvested (gain ^)
2. we have simply to little amount of copper there, the way i feel it?

3. The more windings we use to brake and harvest a coil, the more voltage we gain while harvesting: i feel we have to short wire there too.

4. The ratio of copper amount between primary (power) and secondary (receiving HV winding) is simply not good to:
little happens (to little Cu in primary) and then this little is 'stretched thin' in the hi turn secondary

I feel we would have better yield:

A. Really heavy copper and 100 turns over 10cm diameter (1mm dia or so)

B. Same 400 winds of 0.3 ... 0.4 mm wire on same 10 cm diameter (might be aircored too)

C. A decent low Vce(sat) BJT like the MJ15003 is or a FET like the ultra low RDS(on): IRF2204 only one thats 400V rated instead of 40V of the former...

D. If the aircored can't fire, wrap it with soft iron wire (that's unable to magnetize if rubbed by a 'neo')

E. If that coil 'sparks' on work, have it enameled or dipped in oil...

IMHO.

problem:
How does one fire a FET off 1.2V? are there low voltage fets?

?

, i will
try a different core and different wire now.

Try more copper or lesser ratio:

50:50:300
?
I was 'almost there' with 20:20:120 should made the sec. longer ?

[StevanC]

Congratulations SteveC

It sounds like your having very good results.

Kind thanks.

It depends what one is after, isn't it?
:-)

I learn a lot, and seek to improve what i can

Yes LED's are a LOT more efficient.
You can light quite a few led's off your setup.

I wish You could see my fried collection over here

I even looked at one white LED shining orange (burning through), it's with a brown spot now (dead)
...

When you first try it have your JT turned all the way down.
Then turn it up slowly till your lights get as bright as you want them.

That way you wont burn out your led's.

I used a 1k fixed resistors and a dozen of LEDs in parallel

as chop-wood -I realize now :-(
:-)
You should have put a pot in Your thread opening diagram-post :-) instead of a fixed resistor?

I had mine lighting ten LED's turned all the way down and it ran for three days off one AA.
It was using under 10Ma and the voltage in the battery only went down from 1.27 to 1.23 after running for three days.

The lights were very dimm because it was down all the way.
But it would still light the room up at night enought to see where you was going.

I agree:
The LEDs seem a lot more efficient than the neon firing process:

As I get light comparable in intensity (usefulness) i draw by rule 10% for LED compared to 500mA for the 20cm half-bright neon stick.

What puzzles me most is the flash tube i can't lit?

even put a capacitor across the HV secondary - improved the neon a bit (2-5%) but the flash-tube not a 'zilch'
:-(

I would allso recommend doing as Lidmotor and Kubikop are doing and use the 555 with it.
That would make it a lot more efficient

I would like to discuss the use of the 555 here?

We already have a narrow duty cycle here (about 10% or even less) by he way the JT works.

I would suggest improving the harvesting first, then going to spicing-up and adding more energy consumers (controlling or not, they do burn energy).

but what do I know?

8;-)

[StevanC]

SPICE of JOULE THIEF

@StevanC:

You have found a way to simulate a joule thief in SPICE ???
Then i DO care about it
(snip)

could not upload?
copy and paste to file named jtoo1.asc:

You will need:
SwCADIII or better
defs for:
MPSA06
and maybe for:
2n2222

Version 4
SHEET 1 988 752
WIRE 368 0 -160 0
WIRE -160 48 -160 0
WIRE 368 64 368 0
WIRE 432 64 368 64
WIRE 368 80 368 64
WIRE 32 128 0 128
WIRE 176 128 112 128
WIRE 368 128 256 128
WIRE -160 176 -160 128
WIRE -112 176 -160 176
WIRE 0 176 0 128
WIRE 0 176 -32 176
WIRE 112 176 0 176
WIRE 112 192 112 176
WIRE 176 192 176 128
WIRE 256 192 256 128
WIRE -160 208 -160 176
WIRE 432 272 432 128
WIRE 432 272 336 272
WIRE 368 288 368 128
WIRE 432 288 368 288
WIRE 176 304 176 272
WIRE 176 304 0 304
WIRE 256 304 256 272
WIRE 336 304 336 272
WIRE 336 304 256 304
WIRE 352 304 336 304
WIRE 432 304 432 288
WIRE 432 368 336 368
WIRE 560 368 432 368
WIRE 0 416 0 304
WIRE 0 416 -96 416
WIRE -96 528 -96 416
WIRE 16 528 -96 528
WIRE 112 528 112 272
WIRE 16 576 16 528
WIRE 48 576 16 576
WIRE -160 704 -160 272
WIRE 112 704 112 624
WIRE 112 704 -160 704
WIRE 112 720 112 704
WIRE 560 720 560 368
FLAG 112 720 0
FLAG 560 720 0
SYMBOL ind2 96 288 M180
WINDOW 0 -15 80 Right 0
WINDOW 3 -7 42 Right 0
SYMATTR InstName L1
SYMATTR Value 1�
SYMATTR Type ind
SYMBOL npn2 48 528 R0
SYMATTR InstName Q1
SYMATTR Value MPSA06
SYMBOL res 16 144 R270
WINDOW 0 32 56 VTop 0
WINDOW 3 0 56 VBottom 0
SYMATTR InstName R1
SYMATTR Value 1000
SYMATTR SpiceLine tol=1 pwr=0.1
SYMBOL ind2 160 176 R0
SYMATTR InstName L2
SYMATTR Value 1�
SYMATTR Type ind
SYMBOL ind2 240 176 R0
SYMATTR InstName L3
SYMATTR Value 760�
SYMATTR Type ind
SYMBOL Misc\\cell -160 208 R0
WINDOW 123 0 0 Left 0
WINDOW 39 24 84 Left 0
WINDOW 3 24 36 Left 0
SYMATTR SpiceLine Rser=1 Cpar=1u
SYMATTR InstName V2
SYMATTR Value 1.3
SYMBOL schottky 448 128 R180
WINDOW 0 24 72 Left 0
WINDOW 3 24 0 Left 0
SYMATTR InstName D3
SYMATTR Value 1N5819
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL schottky 384 128 R180
WINDOW 0 24 72 Left 0
WINDOW 3 24 0 Left 0
SYMATTR InstName D4
SYMATTR Value 1N5819
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL schottky 352 368 R180
WINDOW 0 24 72 Left 0
WINDOW 3 24 0 Left 0
SYMATTR InstName D5
SYMATTR Value 1N5819
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL schottky 448 368 R180
WINDOW 0 24 72 Left 0
WINDOW 3 24 0 Left 0
SYMATTR InstName D6
SYMATTR Value 1N5819
SYMATTR Description Diode
SYMATTR Type diode
SYMBOL res -16 160 R90
WINDOW 0 0 56 VBottom 0
WINDOW 3 32 56 VTop 0
SYMATTR InstName R2
SYMATTR Value 0.1
SYMBOL res -176 32 R0
SYMATTR InstName R3
SYMATTR Value 1000
TEXT 16 -72 Left 0 !K1 L1 L2 L3 .99
TEXT -168 736 Left 0 !.tran 100u
TEXT -264 -40 Left 0 ;The Joule Thief sim

[slayer007]

Some thing else to try to get more light or power.
If you have two JT's circuits connect the two collectors togeather.

Allso you can make a rectifier out of led's.
That way you can have light and still charge of the AC side of your coil.

I would allso put a rectifier across the emitter and collector of the 2n2222 and the 2n3055 for charging other caps or batterys.

[slayer007]

Here is a short video of two joule thiefs circuits connected to the same light.
There both running off the same battery allso.

When I turned both of them up all the way they would want to cancel each other out.
But just connecting both collectors togeather on the 3055's they work togeather.

It would be nice to see what five of them could do togeather.

YouTube - Two Joule Thiefs Running Same Light

[Xenomorph]

@Slayer007
I could be wrong, but that seems like a good way to get the amperage up also in regards to other applications that need more current.
The way you connected both collectors it results in a parallel connection,
so the voltages of both JTs stay equal.

[Lidmotor]

Pulsed Inverted Joule Thief

I took the big 2" toroid coil and put it in my 'inverted joule thief' circuit. It worked great! Then I took the 555 timer circuit I had and used it to pulse that circuit. That really worked great!!! It lit up a 24" flourescent tube to full brightness on 8 volts and .75 amps. The best part was that things didn't get hot and it even charged on the back end.
Here is the video:

YouTube - Pulsed Inverted Joule Thief

Lidmotor