What's the purpose of a joule thief?

Thread Starter

tjohnson

Joined Dec 23, 2014
611
I tried using a hex nut as a toroid, but my circuit still isn't working. With both the nail inductor and the nut toroid, the voltage reading at the LED is -1.5V, so the voltage doesn't seem to be getting amplified like it should be.
 

bertus

Joined Apr 5, 2008
22,270
Hello,

The wire you use has a to thick isolation.
The coupling between the coils will be to low.
Also increase the number of turns, as Dick used even 2 X 60 turns on his nail (120 centre tap).
You can also have a look here for more info on the joule thief:
http://www.bigclive.com/joule.htm

Bertus
 

ian field

Joined Oct 27, 2012
6,536
Not a JT builder, but I do know it's easy to get the leads from the 'tickler' coil (the secondary) reversed. If it's not working, try reversing those leads.
The way I described winding on the collector winding, then twisting out a tap for Vcc and continuing further turns in the same direction for the base winding, automatically gives the correct phasing.

Possibly, more turns are needed when using a nail - maybe pile winding may be better than single layer too.
 

Thread Starter

tjohnson

Joined Dec 23, 2014
611
What value resistor is that? - I'd put a 15R in series with at least 100R pot and tune for maximum smoke............er - LED brightness.
1kΩ - I tried to replicate @DickCappels' flooring nail inductor as closely as possible, but he needed a capacitor to make it work and I don't have one with a small enough value.
 

Brownout

Joined Jan 10, 2012
2,390
After further review, I determine that your coil is conected with the correct polarity. Check that the other components are correctly connected, espically your transistor and LED. You might also need more loops on your coil, and thinner wire.
 
Last edited:

Brownout

Joined Jan 10, 2012
2,390
One more thing. Your transistor needs to be just above threasold for your JT to work. So, get a variable resistor, say 10K. Connect the ends to VCC and ground, and the wiper to you transistor base, the same point where your coil connects. Power up and slowly turn up the base voltage until your JT magically springs to life.

Course, check all those other things mentioned first.
 

DickCappels

Joined Aug 21, 2008
10,153
If one winding is in the wrong direction, just swap the two wires to correct the phase relationship.

You might need a whole lot of turns to get this thing working -the one I made had 60 turns collector winding and 60 turns base winding. A small capacitor like 1000 pf or more across the base resistor sometimes helps.
 

Thread Starter

tjohnson

Joined Dec 23, 2014
611
If one winding is in the wrong direction, just swap the two wires to correct the phase relationship.

You might need a whole lot of turns to get this thing working -the one I made had 60 turns collector winding and 60 turns base winding. A small capacitor like 1000 pf or more across the base resistor sometimes helps.
I added many more turns and tried using a different transistor and LED, but the output voltage (measured at the LED) is 0.02V less than the input voltage (measured at the battery). What in the world is going on here?
 

DickCappels

Joined Aug 21, 2008
10,153
(From the photograph, I can't tell what the left end of the long inductor is connected to.)

Consider this: The center tap and the collector are connected by a copper wire (the collector winding). As such, the average (DC) voltage at the collector contact should be almost the same as that on tap. The "magic" occurs when the peak voltage on the collector exceeds the DC on the tap (though the average voltage is the same) and the peak voltage is that which can drive the LED. Is you LED lighting up?

Is the base winding at the opposite end of the coil from the collector winding?
 

Thread Starter

tjohnson

Joined Dec 23, 2014
611
(From the photograph, I can't tell what the left end of the long inductor is connected to.)

Consider this: The center tap and the collector are connected by a copper wire (the collector winding). As such, the average (DC) voltage at the collector contact should be almost the same as that on tap. The "magic" occurs when the peak voltage on the collector exceeds the DC on the tap (though the average voltage is the same) and the peak voltage is that which can drive the LED. Is you LED lighting up?

Is the base winding at the opposite end of the coil from the collector winding?
The left end of the inductor is wired to the resistor.

The LED is not lighting up. That's the problem.

I originally wound the base winding next to the collector winding, but I changed to winding one around the other.
 

Brownout

Joined Jan 10, 2012
2,390
(From the photograph, I can't tell what the left end of the long inductor is connected to.)
It appears to connect to the battery "-" terminal, and continues as the secondary, as in an auto-transformer. I've looked at the coil, and it appears to me to be connected correctly. Assuming everything else is connected correctly, the problem is, IMO, the transistor isn't biased into conduction, and collector current is zero, meaning the transformer isn't doing anything. I've suggested a remedy, but the TS has decided to ignore the suggestion.
 

Thread Starter

tjohnson

Joined Dec 23, 2014
611
The left (in the picture) end is connected to a resistor which connects it to the base leg.
The right end is connected to the LED and negative battery terminal.
The center tap (at left end of nail) is connected to the LED, positive battery terminal, and emitter leg.

@Brownout: I appreciate your suggestion, but I'd need to buy a variable resistor so for me it's a last resort.
 
Last edited:

ian field

Joined Oct 27, 2012
6,536
One more thing. Your transistor needs to be just above threasold for your JT to work. So, get a variable resistor, say 10K. Connect the ends to VCC and ground, and the wiper to you transistor base, the same point where your coil connects. Power up and slowly turn up the base voltage until your JT magically springs to life.

Course, check all those other things mentioned first.
The blocking oscillator JT works best with nothing more complicated than a current limiting resistor in series with the base winding. Getting above the threshold gets mathematical if you use a voltage divider.

The ideal resistance depends mostly on the transistor gain, you need a minimum value in series with any pot so turning it full on doesn't dump the 1.5V into the 0.7V B/E Vf. 100R should be sufficient for the minimum value although I've seen a design with just a fixed 15R resistor - that used a BFY51 or similar. The pot might be anything from a few hundred R to a few k.
 

Brownout

Joined Jan 10, 2012
2,390
Getting above the threshold gets mathematical if you use a voltage divider.
Not really. Use a 10k pot and slowly turn up the voltage. No math required. Though your comment about a limiting fixed resistor is good. Don't want to burn up the transistor while turning up the voltage. a 1K fixed resistor should limit base current to a safe value.

Right now, making the oscillator the best it can be is secondary to just getting it to work. Given the components used so far, and the lack of success, quickly implementing a starting circuit would go a long way towards making a working device.
 

Thread Starter

tjohnson

Joined Dec 23, 2014
611
Right now, making the oscillator the best it can be is secondary to just getting it to work. Given the components used so far, and the lack of success, quickly implementing a starting circuit would go a long way towards making a working device.
Sorry, I don't understand what you mean. Could you please explain?

A CFL bulb blew out at my house just a few days ago, so I'm thinking that I might try to salvage a toroid from it. I'm somewhat hesitant to do that, though, because I'm afraid of breaking the glass tube which contains mercury.
 

Brownout

Joined Jan 10, 2012
2,390
You don't have to break the tube, just separate and base from the bulb. I've done it numerous times. But I doubt you'll find a toroid. You'd have a better chance with an old computer power supply.

The explanation of my post can be found in post #50.
 
Top