Joule Thief 101

[MileHigh]

@ MH - in your comparative analysis of the two circuits, did you notice the functional difference between the two diagrams?

essentially, they operate the same, minus one important factor.

the R-C component of the tank circuit vs the L of the coil are set to resonate with each other.
This, not taken into consideration in the JT equivalent circuit, reduces performance.

The same could be said about the Armstrong circuit, if one chose to change the value of R or C.

I really don't know what you are saying here when you refer to the resistance in the Armstrong oscillator tank circuit.  For sure it is there but it is not relevant to the powered oscillation of the LC tank.

In a Joule Thief, the operating frequency is dependent on L/R type time constants, one L/R time constant for the energizing of the coil, and another L/R-type time constant for the discharge of the inductor energy through the LED.

Now, seriously, how can you equate the resonant frequency of an Armstrong oscillator based on an LC resonant tank and the operating frequency of a Joule Thief based on a first L/R time constant for the energizing of the main coil and a second L/R-type time constant for the discharging of the coil through the LED?

Armstrong oscillator:  operating frequency determined by LC resonant tank.

Joule Thief:  operating frequency determined by (1/(L/R time constant#1 + L/R time constant#2))

Can you see how completely different these two methods are for determining the operating frequency are and how a Joule Thief's operating frequency has absolutely nothing to do with resonance?

MileHigh

[sm0ky2]

I really don't know what you are saying here when you refer to the resistance in the Armstrong oscillator tank circuit.  For sure it is there but it is not relevant to the powered oscillation of the LC tank.

In a Joule Thief, the operating frequency is dependent on L/R type time constants, one L/R time constant for the energizing of the coil, and another L/R-type time constant for the discharge of the inductor energy through the LED.

Now, seriously, how can you equate the resonant frequency of an Armstrong oscillator based on an LC resonant tank and the operating frequency of a Joule Thief based on a first L/R time constant for the energizing of the main coil and a second L/R-type time constant for the discharging of the coil through the LED?

Armstrong oscillator:  operating frequency determined by LC resonant tank.

Joule Thief:  operating frequency determined by (1/(L/R time constant#1 + L/R time constant#2))

Can you see how completely different these two methods are for determining the operating frequency are and how a Joule Thief's operating frequency has absolutely nothing to do with resonance?

MileHigh

you are basically peeling an apple, taking the seeds out and proclaiming, see, this is not an apple at all....

the Resistance/Impedance of the Armstrong circuit is equally important as the Capacitance and Inductance.
In fact, all 3 must be maintained in perfect balance for the circuit to be resonant at that frequency.

This quality makes the Armstrong Oscillator an RLC circuit, Not simply an LC tank.
Though, under certain analysis, the two can behave similarly.

[I would go even further by stating that an LC tank is technically defined also as an RLC.
because of our wires containing some resistance value, but its effect on resonance frequencies can be negligible]

the Joule Thief, is also an RLC circuit, and its' type is classified as an Armstrong Oscillator.
this is a technical definition written in the stone of electronics theory.
Many RLC circuits exist, and most of them are named according to their Inventor, or a particular aspect of their operation.
When we classify the Joule Thief, this is the category it falls under.
all circuits that fall into this category are considered to be Armstrong Oscillators.

[sm0ky2]

I have done, what I think is (currently) my best attempt to bring this knowledge into the public realm, as it pertains to the JT circuit.

To understand more, from perhaps a more technical aspect than I myself can present.
I would direct you to the works of Edwin Armstrong - who is considered by some to be the GodFather of radio.

http://users.erols.com/oldradio/

[sm0ky2]

Smoky2:

In a regular Joule Thief if you remove the LED it will presumably still operate like you state.  The inductor would have no choice but to discharge through the transistor.  The average power is low so presumably it would not fry the transistor junctions.

You make some big assumptions here, but outside of resonant frequencies, yes your electronics should be safe, given the parameters of the system to be within tolerance.

The voltage output from the coil will spike to a quite high voltage, it all depends on the speed that the transistor switches off.

This "speed" you speak of,. this wouldn't be related to "frequency" would it?
More specifically, the frequency-dependent curve of the transistor switching function?

However, I have "caught" you here with respect to the discharge current.  The current will NOT increase.  Are you sure that you fully understand the complete dynamics of an inductor?

Allow me to clarify, by "increase" upon removal of the LED, it can instead be stated that:
by including an LED, there is a drop in current through the parallel paths, and an associated voltage drop across the diode.
This is important to consider, when analyzing the feedback signal.
It represents a higher impedance, as well as a lower voltage.
Impedance differs from a purely resistance perspective,
because changes in amplitude over time as well as phase come into play.
Note that it does not matter if this impedance is included in series or parallel.
Though its' physical location around the loop does affect certain parameters,
as shown in TK's demonstration above. <- while this makes for great conversation,
I feel that is above the technical level of a basic "101" crash course.

I strongly suggest that you go back and watch the clip about the operation of a Joule Thief that I linked to the other day to review the positive-feedback "snapping" mechanism that switches the transistor ON and OFF and governs the operating frequency of the device.  It is also related to the rate of change of current flow through the main coil which is indeed related to the characteristics of the inductor and battery combination.

I do not particularly agree with the assumptions made by that analysis.
While these factors are related, as I have described in previous posts,
phase angle between the signals must be properly considered to discuss what is being shown.

I agree that you can experiment with the number of turns in the coil that connects to the base resistor.  If you do that then you may want to change the value of the base resistor.  In the context of what you are stating, a transistor does not have an "operating voltage" it has an operating current.

MileHigh

Hmm,.. I've run into this before.  where I come from we use terms like Cut-in/Cut-out, or Cut-On/Cut-Off.
What this refers to is:
the voltage threshold that represents the transition stage between:
the Cut-Off and Active regions of the transistor. Below this voltage, the semiconductor does not allow current to pass.
Above this voltage current can travel.
This function is controlled in part by the base voltage (bias).

The transistor in the Joule thief transitions between Cut-off and forward active operation modes.
(up to the point of saturation) at which point the diode becomes the primary conductor until voltage potential drops below
the cut-off of the LED. at which point it begins to dissipate its' capacitance as light. (discharge)
Saturation only generally occurs in a JT when the LED(s) have a high internal capacitance (long discharge time).
This allows for a unique scenario when the voltage drop across the diode makes the emitter voltage appear lower than the base.

Otherwise, the transistor remains in one of these two states.
The actual timing diagram of the switching function, can display a wide range of characteristics.
Outside of linear mode, and/or resonant operation - this function appears as a sharp spike at cut-on, and a gradual decrease at cut-off. (removing the LED changes the shape of these spikes).
When operated in linear mode, at resonance, it is a pure sine-wave function, with varying amplitudes.

[MileHigh]

you are basically peeling an apple, taking the seeds out and proclaiming, see, this is not an apple at all....

the Resistance/Impedance of the Armstrong circuit is equally important as the Capacitance and Inductance.
In fact, all 3 must be maintained in perfect balance for the circuit to be resonant at that frequency.

This quality makes the Armstrong Oscillator an RLC circuit, Not simply an LC tank.
Though, under certain analysis, the two can behave similarly.

[I would go even further by stating that an LC tank is technically defined also as an RLC.
because of our wires containing some resistance value, but its effect on resonance frequencies can be negligible]

the Joule Thief, is also an RLC circuit, and its' type is classified as an Armstrong Oscillator.
this is a technical definition written in the stone of electronics theory.
Many RLC circuits exist, and most of them are named according to their Inventor, or a particular aspect of their operation.
When we classify the Joule Thief, this is the category it falls under.
all circuits that fall into this category are considered to be Armstrong Oscillators.

No in fact the resistance is not that critical in the RLC resonator because it is an active circuit where an external power source keeps the resonator resonating regardless of the inherent resistance in the resonating components.  There is no special balance with regards to the resistance in what is essentially an LC resonator.

Th Joule Thief is not an RLC circuit as I have clearly shown.  It is an active circuit that charges and then discharges a coil.  It's the charging cycle and the discharging cycle that determine the operating frequency, and there is no RLC resonator in sight.  Instead there are two L/R-type time constants that factor in to determine the operating frequency of the Joule Thief in its standard normal operating mode.

You can try to ignore what I am saying, but facts are facts.  Anybody that is interested in electronics would want to study and learn about both pulse circuits and resonating circuits and the associated need to be able to recognize and make a distinction between pulse circuits and resonating circuits.

Note that I am not talking about a hacked Joule Thief circuit here, just an ordinary plain vanilla Joule Thief that is a basic pulse circuit that switches a transistor on and off.  It's a distant cousin of a 555 timer circuit configured as a free running astable multibrator.  Likewise, a 555 running as an astable multivibrator has nothing to do with resonance.  Its operating frequency is determined by RC time constants whereas for the Joule Thief its operating frequency is determined by L/R time constants.

Like I said, you have a "fan club" and anyone interested in Joule Thieves should build a standard Joule Thief first and understand how it operates and probe it with their scope and observe the positive feedback mechanisms in operation.  Then if they want to hack into it and try to make it resonate then more power to them.  The critical point being that if they are claiming resonance then they need to identify the L and C components that are exchanging energy back and forth and show that in action.

MileHigh