Joule Thief 101

[Magluvin]

author=MileHigh link=topic=8341.msg475491#msg475491 date=1456418556]
Bill:



MileHigh

So show us your smart's MH--design a simple circuit that lowers the base resistance as the voltage in the battery drop's. You pride your self on bench time and knowledge,so show us some of that smarts. You draw up the circuit,and i'll spend the time and money putting it together,and testing it. I'll post the results and video's here.


Brad

While im at work, a lot of the time my mind is on this stuff. You and I are on the same plane it seems. I had thought of the 'exact' same thing. To have the base control adjust as the batt voltage goes down.  Not sure of a self adjusting resistor but using another transistor(circuit) configured to do the job.

Good thought bud.

Mags

[Pirate88179]

author=MileHigh link=topic=8341.msg475491#msg475491 date=1456418556]
Bill:



MileHigh

So show us your smart's MH--design a simple circuit that lowers the base resistance as the voltage in the battery drop's. You pride your self on bench time and knowledge,so show us some of that smarts. You draw up the circuit,and i'll spend the time and money putting it together,and testing it. I'll post the results and video's here.


Brad

I had thought about this as well but dismissed it as I "assumed" the resistance adjusting circuit would waste more energy than it was worth.  Maybe I was wrong?


Bill

[sm0ky2]

MileHigh

So show us your smart's MH--design a simple circuit that lowers the base resistance as the voltage in the battery drop's. You pride your self on bench time and knowledge,so show us some of that smarts. You draw up the circuit,and i'll spend the time and money putting it together,and testing it. I'll post the results and video's here.


Brad

it is a difference in thinking. or perspective, point of view, etc.

in one side, you can preserve the AC wave properties of the inductor.

on the other, you can simply switch digitally, and the whole circuit acts similar to a simple boost converter.

https://en.wikipedia.org/wiki/Boost_converter

while, yes you can test this and it will give some results, (ooh, the LED lights up!!)
it will only give insight on one way of thinking.

to understand how a battery voltage can be increased to light up an LED, you can look for those properties that do so.

To understand what goes on inside a Joule Thief,...
throw the LED out the window.....  it causes more problems with analysis than it benefits from the light it produces.
the Joule Thief "does" a lot more things than just boost the voltage from a DC source, to turn on a light emitting diode.

now,- MH may provide us with a circuit to "test".

But let us "test" a JT circuit, with no diode.

and let us test it using a single inductive loop, wrapped around the ferrite (preferably opposite the JT primary coil).
one turn, to the scope.
This gives us a representation of the magnetic waveform through the ferrite.
In this manner, the circuit can be viewed from the perspective of the inductor, while resistance is changed at the base.

This can be expanded by multiple turns on the secondary winding, to further boost the voltage, and impedance in the load-side of the circuit.
This separates any load, to the output side of the transformer. So only the inductive properties affect the operating frequency.

[sm0ky2]

there is only one effective "school of thought".

it is that which is taught in schools

transistors "switching properly"
inductors in the "normal range of operation"

we all have gone over the mathematics of recent days,
we know the circumstances where this "school of thought" explicitly instructs us NOT to go.
Why? What happens when we go there?
capacitors explode, wires burn up, resistors, diodes and transistors fail.
all this happens in circuits that were designed to be well within the tolerances of the components involved.

There are reasons we are taught NOT to do these things.

I will not ignorantly claim these energies to be "overunity" by any definition of the term.
by the contrary, all of the energy involved, we ourselves put into the circuit.

the difference in this "school of thought", is that the normal operation of a circuit,
we throw it all away.

it is our practice to dump everything to circuit ground.

not to loop it back around through itself.

here is one example of a different "school of thought".
this circuit as
proposed by one of our peers, for analysis of the feedback loop.
was presented to me recently.

the effects of this are most prominently demonstrated
as the timing through the feedback loop approaches a phase balance with the input frequency.
the results are quite interesting

[note that this is half-wave AC rectification, and behaves quite differently than its Pulsed DC square wave digital counterpart]

[MileHigh]

author=MileHigh link=topic=8341.msg475491#msg475491 date=1456418556]
MileHigh

So show us your smart's MH--design a simple circuit that lowers the base resistance as the voltage in the battery drop's. You pride your self on bench time and knowledge,so show us some of that smarts. You draw up the circuit,and i'll spend the time and money putting it together,and testing it. I'll post the results and video's here.

Brad

You have to back up first.  Who says that you need a variable resistor for a Joule Thief?  I am not aware of that requirement.  The only thing that I am aware of is that it was done all the time in the old days of this thread, presumably without any serious circuit analysis or scope work done to see what was happening.  It was just an anecdotal observation that the LED changed in intensity when you changed the base resistance.

You are suggesting it so it suggests a question for you:  Why do you want to have a variable base resistor in a standard Joule Thief?

With respect to a voltage-controlled resistor, I did what anybody would do.  I did a Google search.  I saw techniques that used an FET in its linear range which seems reasonable.  I saw more sophisticated techniques using operational amplifiers and others using opto-isolators.  However, there are going to be constraints and limitations with each one of these techniques, and it's undetermined if one of these methods would plug into a standard Joule Thief.  What about power consumption issues?

So for starters, how about backing up.  Why do you want to have a variable base resistor in a standard Joule Thief?

MileHigh