Electronics and Joule Thief Circuits

[astral]

Well I've been messing around with some of these circuits lately and have had some interesting results.

joule thief

make a joule thief


I'm working on a mosfet joule thief that is powered by radio frequency alone, no battery I'll post some pics once I have it built.

Anyway I was wondering if anyone has played around with these circuits and what results you have received.

[QwazyWabbit]

That's not a Joule thief! That's a juice squeezer. You already own the Joules in the spent battery so it can't be called thieving.

It's actually an astable multivibrator that runs on the pulses created in the transistor due to the feedback in the coil, the coil increases the voltage by mutual inductance between the windings. Put a scope on that LED, I'm dying to know what frequency it's running at.

I'll bet if you put a diode across a long enough dipole antenna you could get the thief to run on the ambient RF alone.

[astral]

I'll bet if you put a diode across a long enough dipole antenna you could get the thief to run on the ambient RF alone.

Dude did you read my post.....here:

I'm working on a mosfet joule thief that is powered by radio frequency alone, no battery I'll post some pics once I have it built.

 

[astral]

Plus it takes more than a DIODE across the antenna to transduce RF signal to voltage, your need to rectify the complete signal. So it would take 4  diodes for full wave rectification.  Plus you need dc signal filtered , so that would take some capacitors.  and that alone is just the RF>>> to voltage transformer.

if you have another way to do such things please share...

As for the joule thief of semantics of that circuit, there are many types brands or claims of what they are or do.  Mostly I'm trying to make a cell phone charger I can take around with me anywhere and mosfets are the best in terms of consumption of energy.

Again if you have other ways, share

[QwazyWabbit]

I'll bet if you put a diode across a long enough dipole antenna you could get the thief to run on the ambient RF alone.

Dude did you read my post.....here:

I'm working on a mosfet joule thief that is powered by radio frequency alone, no battery I'll post some pics once I have it built.

 

Yes, I did read your post completely and I agreed.

No, you don't need a full-wave bridge, but it might help. The problem with a bridge will be the lost voltage due to the forward drop of the conducting diodes. You will lose a lot of energy overcoming that. You can get several microamps off a dipole like this:

Diode: 1n4148
Capacitor: .1 uF disc ceramic.

------------------------------------------------------------->|------------------------------------------------------------------
                                                                  |                 |
                                                                   ----) |--------
                                                                  |                 |
                                                                  0                 0
                                                                  (-)              (+)
 

Why no full-wave? Because it's not one frequency, you are getting the full spectrum of RF around, including 60Hz (in US) or 50Hz. The circuit is not very frequency selective except for the resonant frequency of the antenna as a 1/2 wave of some radio frequency. (150MHz for a 2M dipole for instance). This is the basis for simple field strength meters if you put a 50 uA ammeter movement across the terminals. The longer it is the better. The Joule thief won't care about the filtration of the energy it needs since it's already a chopper to start with.

PS: I am a EE, ECET and an active ham radio operator, advanced class since 1974.

[astral]

I'll bet if you put a diode across a long enough dipole antenna you could get the thief to run on the ambient RF alone.

Dude did you read my post.....here:

I'm working on a mosfet joule thief that is powered by radio frequency alone, no battery I'll post some pics once I have it built.

 

Yes, I did read your post completely and I agreed.

No, you don't need a full-wave bridge, but it might help. The problem with a bridge will be the lost voltage due to the forward drop of the conducting diodes. You will lose a lot of energy overcoming that. You can get several microamps off a dipole like this:

Diode: 1n4148
Capacitor: .1 uF disc ceramic.

------------------------------------------------------------->|------------------------------------------------------------------
                                                                  |                 |
                                                                   ----) |--------
                                                                  |                 |
                                                                  0                 0
                                                                  (-)              (+)
 

Why no full-wave? Because it's not one frequency, you are getting the full spectrum of RF around, including 60Hz (in US) or 50Hz. The circuit is not very frequency selective except for the resonant frequency of the antenna as a 1/2 wave of some radio frequency. (150MHz for a 2M dipole for instance). This is the basis for simple field strength meters if you put a 50 uA ammeter movement across the terminals. The longer it is the better. The Joule thief won't care about the filtration of the energy it needs since it's already a chopper to start with.

PS: I am a EE, ECET and an active ham radio operator, advanced class since 1974.

   sweetnees qwazzy rabbit sweetness...  ok so the full wave isnt needed cool I was always in the thought process that I needed to full wave rectify.  Now I am using about a 100ft antenna wound around the ceiling of my apartment think that should suffice?

I've been having issues with the transistors in the joule thief blowing also,  don't know if its because of the step up voltages that  are being put out if I am doing something wrong with the toroid itself.  I have many toriods top play with various sizes and makes from old computer power supplies.  plus I have other inductors to experiment with as well any suggestions?

[QwazyWabbit]

From the winding video it looks like it's 10 turns of each winding around the toroid. (I think the first link said 20 turns but it seems to be 10 turns with the two goofy dudes.

The failure mode of the transistor might tell you what you're doing wrong. A 2N3904 has a Vce max of 40 or 50 volts so you should not be failing there. Don't use a big inductor, a small toroid like they use will be sufficient for powering an LED. I think the white LED's pull about 20 mA forward current at 3V but they don't say what the part number is so it's hard to look up. The 2N3904 can handle Ic max of 200 mA easily in a TO-92 case. If you are blowing the transistor then it's possibly due to too much current through the base-emitter junction. If the transistor is opening between C-E then it's over-voltage collector-emitter. If it's opening B-E then it's over current in the base. Cracking the case and letting the magic smoke out is excess power dissipation.

If you look at the circuit the LED's forward voltage drop clamps the C-E voltage across the transistor, what's happening is the collapsing magnetic field of the inductor is producing the diode forward current when the transistor is OFF. If you hook the diode up backwards or you don't have a diode the inductor can over-volt the transistor. When the transistor is ON it's pulling max current out of the battery through the inductor to build up a nice fat magnetic field, the changing current is what tickles the base-emitter junction into turning the transistor on and off via V_batt. * di/dt through the secondary. You end up with a very high frequency oscillation as the magnetic fields build and collapse.

DON'T use a new battery. It needs to be very depleted before this circuit won't over-produce. The battery is being shorted by the transistor through the inductor so you can pop a transistor pretty easily when that inductor kicks back.

I am not sure an open air coil will have enough inductance to kick the oscillator, trial and error there I think. 2 or 3 turns in a 10' x 10' room maybe? Ring the primary and secondary VERY close together around the ceiling.

You definitely won't get enough voltage out of it from RF to charge a phone battery.

[astral]

From the winding video it looks like it's 10 turns of each winding around the toroid. (I think the first link said 20 turns but it seems to be 10 turns with the two goofy dudes.

The failure mode of the transistor might tell you what you're doing wrong. A 2N3904 has a Vce max of 40 or 50 volts so you should not be failing there.

Ok here is where I am different, my transistor is a 2N4401.  Now I am not an expert here so how much impact does this make?  Other builders suggested the above because of better characteristics.

Next maybe I should have explain that my goal is to make a circuit that will power my cell.  Not that this one was the one... because I am hoping to get certain results, though it might not be possible and so be it then it will prove my hypothesis wrong.

 

Don't use a big inductor, a small toroid like they use will be sufficient for powering an LED. I think the white LED's pull about 20 mA forward current at 3V

I am using a yellow toroid from a recent power supply, with 22g pvc jacket 600v yellow copper about 40 windings iron ferrite I dont have the exact specs I know I need them but off hand the coil is about 3 orders of size  larger than the one they use  so i might be just destroying that little transistor. Plus I would assume this one has a higher q as well.

The 2N3904 can handle Ic max of 200 mA easily in a TO-92 case. If you are blowing the transistor then it's possibly due to too much current through the base-emitter junction. If the transistor is opening between C-E then it's over-voltage collector-emitter. If it's opening B-E then it's over current in the base. Cracking the case and letting the magic smoke out is excess power dissipation.

Since I have now explained what is different what do you think about pure over load of this 2N4401 I am using?  For the pun of is this one hasn't smoked yet .    The 2N3904 was history pretty quickly.  This is why I wanted to try mosfets, since it can be mounted to heatsinks and cooled quiet nicely as you sure know. 

If you look at the circuit the LED's forward voltage drop clamps the C-E voltage across the transistor, what's happening is the collapsing magnetic field of the inductor is producing the diode forward current when the transistor is OFF. If you hook the diode up backwards or you don't have a diode the inductor can over-volt the transistor. When the transistor is ON it's pulling max current out of the battery through the inductor to build up a nice fat magnetic field, the changing current is what tickles the base-emitter junction into turning the transistor on and off via V_batt. * di/dt through the secondary. You end up with a very high frequency oscillation as the magnetic fields build and collapse.

DON'T use a new battery. It needs to be very depleted before this circuit won't over-produce. The battery is being shorted by the transistor through the inductor so you can pop a transistor pretty easily when that inductor kicks back.

They have been new, I just tested them, though residing in the dead bin, they were in fact new and full of life.  So now we have some diagnosis.

I am not sure an open air coil will have enough inductance to kick the oscillator, trial and error there I think. 2 or 3 turns in a 10' x 10' room maybe? Ring the primary and secondary VERY close together around the ceiling.

You definitely won't get enough voltage out of it from RF to charge a phone battery.

Yeah again I am trying to get the hypothesis down to make a schematic.  ahem that works


Now I guess that might help out.

[yahoo]

can i use this to power my house?

[astral]

can i use this to power my house?

that would sure be awesome right yahoo? 

[QwazyWabbit]

A 2n4401 is higher power than a 2n3904 and can handle 600mA vs 200mA. Slightly lower F_t so it might switch slower and have a higher internal power dissipation as a result.

Your bigger coil will store more energy (more turns, bigger ferrite) therefore your transistor needs to be that much bigger. Now your getting into the realm of magnetics and ferrite and how much energy is stored in a magnetic field. Time for some experimentation and some math.

With a MOSFET you won't be able to use a single resistor, you need the current flow through the secondary. You will have to use two resistors, one to replace the 1k and another from the gate to common. You will have to calculate the values based on the gate voltage needed to fire the transistor. I'm not sure off-hand if a MOSFET will work in this application.

The switching frequency is a function of the parasitic capacitance of the toroid and the F_t of the transistor. Essentially, the toroid is a tank circuit (parallel L-C network) and it's going to resonate at some frequency, at resonance the voltage across the inductor gets rather high, depending on the available energy in the magnetic field. This is the controlling factor for the power dissipation in the transistor.

I think you're going to have some fun blowing up transistors until you find your solution. Either that, or do the math to calculate the currents and voltages for the particular toroid you're going to need for charging the phone battery, then work it back from there.

[astral]

Question...

How do you go about attaching this dipole to the joule thief circuit?

[Pr0c3550r]

lol @ this thread

[QwazyWabbit]

The plus terminal of the dipole will be the cathode end (striped end) of the diode in that circuit I drew above. Connect the wire from that junction to where you would have put the plus terminal of the battery. Same thing goes for the negative terminal. You will get microamps of current from RF in your region, probably about 100 microvolts if you have a nice strong AM radio station within a few miles of your location. Note: the voltage from a spent AA battery (0.35 V) is 3500 times stronger than 100 microvolts of received RF (0.0001 V) so you will be very lucky to get the chopper to get excited about it. Changing the dipole to a loop antenna, like about 100 turns of wire around your 10' x 10' room will get you more voltage. One continuous run, with the two ends terminated in a corner or middle of the wall, with the diode and capacitor as show in the dipole drawing will work the same way.

[astral]

Well finally..

The postal service sucks at delivering things, but I finally received my goodies for testing this out.  I think i have abandoned the idea of the ambient module for the time being.  The antenna that works with it is about 175 feet of 26g wire and that isn't feasible for moving around and being mobile.

Anyway I decided to work on a simple nicad charging circuit or charge pump with this idea instead and see what I can accomplish.  Also I am going to use an TS555 astable circuit with a 50% duty cycle to see if that helps with controlling the efficiency of this system as well.  But I ordered resistors and capacitors for other duty cycle ratings just in case more or less is needed.

I ordered various transistors like the 2N4401, 2N4403 and the big dogs like IRF510 and the 2N3055 and variable resistors to tune the circuit.

Since I just got these parts yesterday I haven't had time to assemble nothing yet, but today I am going to get the timers built and get some scope shots of my square wave at least.

Looking forward to a bread-board of a day..muahhahahah!