Joule Thief

[xee2]

@ jeanna

xee, do the secondary 100 turns go in the up then down or just around configuration?

They are just wound in one direction. They go almost all the way around the toroid.

[xee2]

deleted - double post happened for some reason

[TheNOP]

parallel RC circuit,  the capacitor current is 90° out of phase with the resistor (and source) current.
it keep the transistor in conducting state for longer.(duty cycle)

the jt also have an inductance( the base coil ) so,
in series with the resistance + cap in parallel to the resistor that give us an equivalent LC circuit and a RL circuit.

LC =frequency(hz) = 1 / 2Pi RootSquareof(LC)

RL = frequency(hz) = R / 2Pi * L  <--frequency that the filter will attenuate to half its original power

the bjt is using an RL circuit.
lowering the current to the base make the transistor conduct for a less longer time.

that is why some of you sometimes get odd measurements values.

calculating the inductance value is possible but a bit complicate.
Maxwell's equations are usually used for that.

for air core coils, inductance is in function of coil geometry and number of turns.
but when there is a core, permeability must also be taken into account.

[xee2]

@ jeanna

Thanks for the help with frequency vs base capacitance. I stopped taking frequency measurements because I thought my frequency counter was not giving correct readings with the unsymmetrical waveforms. But now that you have explained what you were seeing on scope, I think it is giving correct answers (just not the ones I was expecting). The following results match what you are getting. So it seems the base capacitor is narrowing the pulses rather than increasing the spacing between them. In fact, it must be decreasing the spacing between them if the frequency is increasing. Thanks for the info. If you can do any confirmation tests they will be appreciated.

[stprue]

Wow not that I'm home using the mac I cant see any of the pics of xee's I was just viewing at work!  Anybody know what's wrong?