Big Joule Theif

[Lidmotor]

Foil killing the "Thief"

WOW..Very Nice Lidmotor.

Very bright light I like that.
Did you notice when the foil tutches togeather it will about kill the JT or at least slow it down to nothing.

I thought that was strange.A small gap wouldnt affect it but as soon as both ends tutch it slows it down to nothing.

Yes I did notice that. In fact I thought that maybe I could "tune" the coil by moving the loop around or changing the shape but I it didn't work. It just changed the frequency or decreased the output. You are right-- if the loop came together it killed it completely on my circuit.

Lidmotor

[Xenomorph]

555 Low duty cycle

@Lidmotor: If you are interested in possibly advancing the efficiency of the circuit even more, i can recommend you a 555 timer modification that allows for low duty cycles (<50%).
I have successfully replicated your circuit using a 15% duty cycle at 100 Hz without noticable flickering, just enough to energize the coil.
That should run on a third of the energy needed to run the 50% duty cycle at at specific frequency.
Comparisons in brightness need still to be made between 50% and 15% duty cycle, as i have directly started with the low duty cycle.
Hopefully there is not much of a brightness loss, it really comes down to trying it out.



The values i used:



You can read about how to achieve the low duty cycle here:
Design low-duty-cycle timer circuits - 8/22/2002 - EDN

and there is a handy excel calculator for the resistor values:
http://a330.g.akamai.net/7/330/2540/...ges/di2972.zip

a little vid of my replication: YouTube - PulsedInverterLowDuty

I hope this is of any use for you.

Regards,
Xenomorph

[ABCStore]

Xenomorph,

could you please decipher what value R3 is, what does "Use Diode" mean and where this D1 diode is connected to?

Thanks,
ABC

[Xenomorph]

@ABCStore,
the information is in the article Design low-duty-cycle timer circuits - 8/22/2002 - EDN

The diode is connected in parallel to R2, it is needed to bypass R2 during the charging cycle to achieve the lower duty cycle.
R3 is not specified, i did not use a resistor to connect pin 4 to Vcc. as it is not part of the equation.

[Lidmotor]

Xenomorph 555 timer---short duty cycle

@Lidmotor: If you are interested in possibly advancing the efficiency of the circuit even more, i can recommend you a 555 timer modification that allows for low duty cycles (<50%).
I have successfully replicated your circuit using a 15% duty cycle at 100 Hz without noticable flickering, just enough to energize the coil.
That should run on a third of the energy needed to run the 50% duty cycle at at specific frequency.
Comparisons in brightness need still to be made between 50% and 15% duty cycle, as i have directly started with the low duty cycle.
Hopefully there is not much of a brightness loss, it really comes down to trying it out.



The values i used:



You can read about how to achieve the low duty cycle here:
Design low-duty-cycle timer circuits - 8/22/2002 - EDN

and there is a handy excel calculator for the resistor values:
http://a330.g.akamai.net/7/330/2540/...ges/di2972.zip

a little vid of my replication: YouTube - PulsedInverterLowDuty

I hope this is of any use for you.

Regards,
Xenomorph

@Xenomorph-----I worked all day with the 555 circuit and made a few small changes but basically replicated what you did. I set it up initially exactly like you described and then worked backwards until I got the charging and light output balanced with the amp draw. I had to lower the R1&R2 resistance values. I decided to put a 75K pot with a 100 ohm resistor at R2 so I had some adjustment. R1 went to a 10K resistor. The rest of the circuit stayed the same.
This modification that you made to the "Pulsed Inverted Joule Thief" was a great help in reducing the power consumption. . Here is the video of it:

YouTube - Pulsed Inverted Joule Thief Xenomorph modifications



Lidmotor

[Xenomorph]

@Lidmotor:
I noticed that the CFLs start to fire when the voltage on the charging cap reaches around 37 Volts. After it has fired, the voltage on the cap would fall to something like 29 Volts and the CFL would stay on.
Interesting that you have found a CFL that fires right away.
I was trying to explain that with insufficient voltage for the CFLs to fire, maybe since the cap and the CFL are in a quasi-(one-wire) parallel connection sharing the positive of the supply battery,
but you proove with that, that this cant be the only explanation.
Good to see that you were managing to decrease the amp-draw significantly with the lower duty cycle.
Are you using the low power 555 (Texas Instruments TLC555) in that set-up?

[Lidmotor]

555CN timer

@Lidmotor:
I noticed that the CFLs start to fire when the voltage on the charging cap reaches around 37 Volts. After it has fired, the voltage on the cap would fall to something like 29 Volts and the CFL would stay on.
Interesting that you have found a CFL that fires right away.
I was trying to explain that with insufficient voltage for the CFLs to fire, maybe since the cap and the CFL are in a quasi-(one-wire) parallel connection sharing the positive of the supply battery,
but you proove with that, that this cant be the only explanation.
Good to see that you were managing to decrease the amp-draw significantly with the lower duty cycle.
Are you using the low power 555 (Texas Instruments TLC555) in that set-up?

@Xenomorph-----I am using a plain old Radio Shack 555CN part # 276-1723. Works great!! I really like the way this circiuit came together thanks to your improvement. The right bulb for this circuit is somehow important. I am not quit sure why. I may put another 100 turns on the toroid secondary and see if that makes a difference. It works the way it is but to get the charging to happen and the light to go on the way I want it, I have to use the 13 watt horseshoe bulb..

Lidmotor

[Lidmotor]

CMOS 555 low power timer chip-----Works

@Xenomorph---- UPDATE. I just swapped out the 555CN to a low power CMOS TLC555 chip and it works better. The amp draw went down and the chip is handling this circuit just fine.

Lidmotor

[Mutten]

I was looking for a toroid core and found this one and was wondering what people think of it. Would it be a good start to a JT/555?

Already built a small JT but I can't get enough windings for a good tertiary winding.


829-0080-UNIVERSAL TOROID CORE-Electronic Surplus Inc.Offering obsolete IC's, discretes, motors, relays, switches, pots, and much, much more! Always looking for your excess inventory.

[slayer007]

Mutten it will work better with a ferrite core.

I didn't have much luck with the iron cores with a secondary winding.

Here are the specs of the core Lidmotor and I are using.

OD: 2" (49mm) ID: 1.36" (34mm) Height: 0.64" (16mm)

Material: ASTM P7070 ( TSF7070 ) power ferrite

[Mutten]

Mutten it will work better with a ferrite core.

I didn't have much luck with the iron cores with a secondary winding.

Here are the specs of the core Lidmotor and I are using.

OD: 2" (49mm) ID: 1.36" (34mm) Height: 0.64" (16mm)

Material: ASTM P7070 ( TSF7070 ) power ferrite

Ah right, thanks for pointing out that it was Iron, I missed that.

Was looking for a shortcut since winding a toroid is a lesson in patience (PITA).

Speaking of which, anyone have any advice for winding one ?

[Guruji]

555timer charging

Hi Lidmotor can you please post a schematic for us all?
Thanks

[slayer007]

Ah right, thanks for pointing out that it was Iron, I missed that.

Was looking for a shortcut since winding a toroid is a lesson in patience (PITA).

Speaking of which, anyone have any advice for winding one ?

Mutten When I did mine I wound the wire on a long thin bolt.

Then you can just feed the bolt through the toriod.
Kinda like sewing.
Wind two strands of wire on the bolt for the joule thief.
Then do the same for the pickup coil but just one wire on the bolt.

Wind your JT part first then wind the pickup coil.
DO NOT go over the Jt winding with your pickup coil.
Keep them seperate do not let it overlap onto the JT windings.

[shlodo]

Well done guys!

@ Everyone

This project is really coming along thanks to iput from everyone.
It sure is an exciting circuit, and the numbers lidmotor is getting with the 555 timer pulse is great.. This circuit is much more efficient at lighting CFLs than the radiant oscillator.

Just wanted to thank everyone working on this... great effort guys.

I came across this circuit the other week and thought u might be able to do something similar to the 555 pulse idea. Its a simple led flasher using a 2n2222 as a negistor... NEGATIVE RESISTOR
The circuit is even simpler than the 555, it only uses a 2n2222, resistor and cap.
I build one and it draws less than 10mA to operate.. the circuit i found is only for 12V and im trying to modify it so it will work on lower voltages.

All you have to do to change the frequency is change the value if the Cap. Its not as versatile as the 555 but i thought it might be of some interest.

Simplest_LED_Flasher_Circuit
-shlodo

[Lidmotor]

Round 30 watt bulb on 555/JT

@Guruji------The circuit I am using is posted a page or so back except for Xenomorphs 555 timer modification. I keep changing little things almost on a daily basis -- fine tuning this thing. People are adding great ideas to the circuit. Today I changed the bulb to a round 30 watt and it worked the best of any so far. Slayer, Kubikop, Xenomorph, Shlodo, me and others came up with the ideas and pooled all the circuits all together. I think that it has been alot of fun for everyone.
Here is the video of the round bulb:

YouTube - Round Bulb On Pulsed Thief

Lidmotor