Joule Thief

[TheNOP]

OK, does any one here have the foggiest idea of what my experiments with the Joule Thief are leading up to? I have had 1 nibble, no one here has asked, why am I'm tying 7 torids in a Series Parallel configuration. Did you all miss it?

in series, probably bigger backemf, but less currents.
almost same as resistor in series calculation, except for the resistance variations that must be taken into account.

in parallel, my guess is it won't change much more then if you were to connect 2 transformers one after the other.

those with a scope can test this.
-jt with a pickup, number of turns of the pickup 20 turns
-an other toroid 2 * 20 turns coils

-connect one coils of the 2* 20 toroid's coil to the jt's pickup
-connect a potentiometer to the other 2* 20 toroid's coil, as a load.

vary the load potentiometer and check if it affect the jt's frequency/amps/volts.

if it does, like i think it is, from my own experiments, there are no advantages to do it like this.
only more losses to be get.

[TheNOP]

..., but the thing that makes this work sooo well is the high frequency of the pulses.
I am sure about that now.

i am happy to see someone saying the same thing i said long ago.
probably not from the same understanding of things as me, but that in not important for now.
just know that the frequency is important for more then just one thing.
there are limiting factors in any electronic circuits, no matter its apparent simplicity. 

next step you are begining to, or are already, understanding it.
duty cycle and the _|__|__|__  vs  _/\__/\__/\__  signal, for the current strenght.

[Pirate88179]

@ Jim:

I am using FF version 3.0.10.  It just came out the day before yesterday.  if you are not sure if this is the latest you have, click help on the tool bar at the top, select check for updates, and it will tell you.  Also, unless you turned it off, FF will alert you to new versions every once in a while.  I found the videos on how to speed up FF and I will e-mail them to you.

Bill

[resonanceman]

@stprue

Good question, I havent decided to drive the 7 series toroids primaries with a NPN Transistor, Mosfet or IGBT.
With the last 2 items, I need to watch out for spikes which could destroy them, then again, I need to push several Amps through the primary to get the output I need, so I think I will use something in the 20-30 Amp range device, perhaps an IRF540N MOSFET or equilivent device.

Then again, if I decide to drive each toroid individually, then I will use individual transisters.

I have been testing my JT this morning, and not having much luck with turning on a 2N3055, or should I say, use the pulses of the JT at the collector to pulce the 2N3055 transistor on and off.

I added another 1.2v NiCad in series with the other NiCad, but the bulb weekly glows.

If I bridge out the BC548 transistor, between the Collector and the Emitter, then the 2N3055 switches hard on, the 6v globe shines.

I think I need another medium powered transister to switch the 2N3055 on and off with the JT pulses.

I will draw up a circuit and you can see what I up to so far.

Thanks for asking
jim

Jim

I just checked  the   datasheet for  IRF540N   
I  don't  think it  will work.


Like most MOSFETS  I have seen it has a  diode  between  the  source and  drain  .....   I may be wrong  but I think a  diode there would  waste  half  of  your pulse


gary   

[Artic_Knight]

from what i understand about florescent lights they love high frequencies, if thats true then should a curcuit designed for higher frequencies be better? according to the work posted by jenna putting a base cap on this at 1f i believe should generate a wonderfuly high frequency. also increasing the base resistor will help a bit, yes it lowers voltage so we can utilize the other technique posted in this forum of having a small primary large secondary (correct me if i got that backwards) to raise the voltage back up. if nothing else a secondary transformer could be utilized to step up the voltage once the desired frequency was met.

additionally i am interested to see if adding the cap accross the coils will help. with jennas work and utilizing the caps ability to create a more true AC current i think we could do the following curcuit to potentially maximize the lighting of a florecent.

i noticed the cap accross the coils does dim the light to a led but it should be generating a more natural AC instead of pulse DC that looks like AC. this curcuit i would think would help with the power draw and lighting the light.

i dont know what the optimum uf would be for the cap accross the coils however.

tell me what you think.