Joule Thief Is Not Working

Discussion in 'The Projects Forum' started by geratheg, Jul 24, 2014.

  1. geratheg

    Thread Starter Member

    Jul 11, 2014
    107
    3
    Applying 3 V, the LED lights up, but applying 1.5 V, no light.
    Green wire is 22 ga, blue wire is 26 ga.
    Using a 2N3904 Transistor and a Ferrite Toroid.

    Am I doing something wrong?
     
  2. to3metalcan

    Member

    Jul 20, 2014
    228
    23
    Hey, can you post your schematic? I can't see any reason for the transformer windings to be joined together at one point the way you have them here.
     
  3. to3metalcan

    Member

    Jul 20, 2014
    228
    23
    Also, and probably more importantly, the sources I've seen recommend about 12 turns of wire around the toroid...only looks like you have about six!
     
  4. geratheg

    Thread Starter Member

    Jul 11, 2014
    107
    3
    Last edited: Jul 24, 2014
  5. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,363
    Couple of points to note:

    1) The number of turns makes a difference. You can get more turns on the toroid by using enameled copper wire of a smaller diameter, 28 AWG for example, or you can use wire-wrap wire.

    2) The direction of the turns (i.e. phase) is critical. That is the reason for the dots shown in the diagram. If it does not work one way, interchange the connections of one of the windings.
     
  6. geratheg

    Thread Starter Member

    Jul 11, 2014
    107
    3
    1) Is it possible to use uninsulated copper wire? Of course it would cause shorts in the toroid, but hmmm.

    2) Tried it and it didn't work. Since this is a 2-wire toroid doesn't the current flow in both directions, meaning direction of turns does not matter?
     
  7. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    yes change the both blue wires.

    If it does not work, change both greens against the blues, and again, swap one of them.

    I made quite a few actually, which you can see on my dragonsnest blog (as pages), and on my google+. You can find them with "dragonsnest altervista", under "RESEARCH" and "MICROCONTROLLER PROJECTS".

    Actually to me it looks the windings are parallel (which is wrong).

    Number of turns is secondary but ratio plays a role too. And it also works with quite small cores.

    you can also rectify the output + install a small capacitor, and make the 1K resistor adjustable, as well use 2 batteries and more LEDs in series.

    I got 56 volts from the one I made yesterday night (with rectification).
     
  8. geratheg

    Thread Starter Member

    Jul 11, 2014
    107
    3
  9. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    Yes sure.

    The dots on the transformer mean the two coils are anti parallel. Simply, if it does not work, swap the two blue leads with each other.

    Also the transistor could be the wrong way, check the datasheet.
    It could be defective as well.

    And the turns ratio could be wrong, that can prevent it from working too, so the next step would be to make sure the two coils have a different turns ratio.

    But maybe you can get it working like it is.
     
  10. geratheg

    Thread Starter Member

    Jul 11, 2014
    107
    3
    I appreciated the help.
    Those guys at makezine used the same amount of turns for both wires.
    I got it to "barely" work using another toroid found from an old computer motherboard, which had one thick wire around it, so I just wrapped another copper wire around it.

    Used it the toroid with the 1.5 V battery and the LED turned on pretty dimly, and only lit up if I just touched the battery, but wouldn't stay lit. Didn't observe this before. Probably a toroid issue more than a transistor issue.

    LED lit with all 4 different wire connection possibilities, so I don't think it depends on direction?
     
  11. takao21203

    Distinguished Member

    Apr 28, 2012
    3,577
    463
    it does depend very much on the direction of the feedback winding.

    Try a different turns ratio. The wire gauge does not matter beccause the low currents. Yes it could be the ferrite isnt suitable.

    Try for instance a 1:2 ratio, and 2:1
     
  12. to3metalcan

    Member

    Jul 20, 2014
    228
    23
    Your results at 3V are a false positive...they don't mean anything. That's enough voltage to turn the LED on through the wire even if you take everything else out of the circuit! Likewise, 1.5V is almost enough to turn it on, so seeing it light dimly at that voltage doesn't mean much, either. A successful Joule Thief should be able to turn on around .5V.

    This circuit is based on the idea of storing up pulses of charge in an inductor...see the eBook: http://www.allaboutcircuits.com/vol_2/chpt_3/2.html

    Inductors are made from turns of insulated wire...the more turns, the more induction. With only six turns, there's almost no way your circuit has enough induction to "hang on" to a charge long enough to build it up the way the circuit is supposed to. Enameled magnet wire (looks like bare wire but is actually varnished with clear insulator) is a good choice to fit some more turns on the toroid.

    And, as several other posters have pointed out, the phase of the windings DOES matter...the two inductors on the same toroid act as a small transformer, and the transistor and the transformer are supposed to turn one another on and off in a never-ending loop...if they're not in the right phase, this won't happen. They'll both turn off and nothing will happen.

    TL;DR: seriously, more turns on the toroid! And yes, the wire needs to be insulated. :)
     
  13. geratheg

    Thread Starter Member

    Jul 11, 2014
    107
    3
    If this is important, my toroid is 7mm inside diameter and 13mm outside diameter.
    Which size copper wire should I use?
    I could stop by a radioshack and get this:
    http://www.radioshack.com/product/index.jsp?productId=2036277

    The price seems a bit high?

    Hmm maybe I wasn't understanding phase correctly, is phase the direction you turn the wire around the toroid?

    Or is it the connections of the wires that determines phase?

    In my case, the wires appear to be connected such that the currents flow both in opposite directions around the toroid, and in opposite directions relative to each other (one current flows into the center, the other flows out of the center of the toroid) if that makes sense.
     
    Last edited: Jul 25, 2014
  14. Fibonacci

    New Member

    May 23, 2014
    25
    5
    Yes, first you don't need the Toroid because you have a CD power supply. Furthermore, with 1.5 volts or less the LED isn't in the forward mode and it has reverse bias, it can not turn on. When you conect a 3 V power supply, the LED turns on because it has forward bias and can operate like a closed switch. Remeber LEDs are diodes. The transistor isn´t helping because the Vce is about 0.3 V in switching mode. Try conecting the LED in the colector lead in series with the toroid, but you don´t need either.
     
  15. to3metalcan

    Member

    Jul 20, 2014
    228
    23
    Fibonacci, I'm not sure you understand the circuit...correctly built, it outputs more voltage (at the expense of current) than it gets from the battery. You're basically just telling him to hook the LED to the battery, which will light it, for sure, but that's not the point...the point is, this circuit will turn on a ~2V LED from a battery that has sagged to 1V or less. You definitely do need the toroid to do this!
     
  16. to3metalcan

    Member

    Jul 20, 2014
    228
    23
    Geratheg, that would be great wire...as far as seeming expensive, well...that's copper prices + RadioShack markup for you. :( On the plus side, you'll be able to make about a million of these...the price is for all 3 spools. You can also get some finer stranded wire that will work fine at the 'Shack, and if memory serves, they usually have toroids, too...

    As to your question, there's really no such thing as a "direction" of winding...the wire doesn't know where you started! Putting 'em side by side is great...perfect, in fact. After you wind it, take one of the colors of wire and "switch" the leads...that inverts the phase. It's not important which one you change, just that you change one of them.
     
  17. geratheg

    Thread Starter Member

    Jul 11, 2014
    107
    3
    Looking in the picture, do you mean switch the single green wire with the single blue wire?

    Or do you mean connect the single green wire with the single blue wire together, and unconnect the already connected wires and make them single?

    I have a T50-7 Toroid

    I'm thinking of buying 1 size copper wire on ebay, which should I get 24, 26, 28, or 30 awg?
     
    Last edited: Jul 25, 2014
  18. to3metalcan

    Member

    Jul 20, 2014
    228
    23
    28 or 30 is fine.

    As far as phase goes, see the attached image, and try to follow this:

    One wire is A. The other wire is B. One end of the wires is 1. The other end of the wires is 2.

    If A1 is connected to the battery, then B2 should be connected to the battery, as well (or vice versa.)

    But if A1 and B1 are connected together, or A2 and B2 are connected together, nothing will happen when you turn it on.

    Clear?
     
  19. geratheg

    Thread Starter Member

    Jul 11, 2014
    107
    3
    Yes that is very clear, thank you.

    A lot less confusion now. And to answer the phase questions, my toroid was wired with the correct phase. Like so: A1 with B2 together.

    Ever heard of this website? http://kitsandparts.com/toroids.php
    Toroids are pretty cheap there and specs sheets provided.
     
  20. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,363
    I said back in post #5, two things matter:

    1) the number of turns,

    2) the direction (phase) of the windings.

    I'm not sure how much attention you have been paying.

    1) Get some fine diameter enameled copper wire (magnet wire), any gauge from 24 AWG to 30 AWG. Wind about 20 turns each for the primary and secondary. One way to do this is to double the wire (i.e. fold the wire at the mid point and then wind the pair around the toroid 20 times.

    2) Yes the direction matters, whether you go clockwise on both windings or one clockwise and the other counter clockwise. This together with how you label the ends of the windings is critical. Interchanging the ends of the windings will reverse the phase.

    But if you don't know the direction and labeling of the ends there are only two possibilities, either the windings are in-phase or they are out-of-phase.

    So, label the first winding as #1. Label the ends of the windings as A and B. So you have two ends labelled 1A and 1B.

    Do the same for the second winding #2 with ends labelled 2A and 2B.

    Assuming all the other components are wired correctly and in working order, wire the primary and secondary windings as shown in the circuit diagram.

    If it doesn't work, simply interchange 1A with 1B. That's it!

    (Sorry, I have used a different labelling scheme from to3metalcan.)
    If you prefer to stick with to3metalcan's labelling:
    The first winding is A with ends A1 and A2.
    If it doesn't work, exchange A1 and A2.

    (Or B1 and B2. If you exhange A1 with A2 and B1 with B2 you are back where you started, i.e. no phase reversal.)
     
    Last edited: Jul 26, 2014
Loading...