Early on in my JT experiments I experimented with a pot (potentiometer) in place of a resistor, which was used as an adjustable resistor and allowed me to optimize that particular JT. After the optimum setting was determined, the pot could be removed and replaced with a regular resistor. It seemed that the pot was confined to use just for experimenting. But at the same time I realized that using that same pot, a JT could be made to have adjustable brightness. One problem was that the size of the pot was as big or bigger than the rest of the JT circuit, and pots are more expensive than a simple resistor. One other thing I noticed: it seemed that no one else left the pot in for daily use.
Then I just came across this circuit that does use a pot to adjust the brightness. Good to see someone else decided to make use of this pot for every day use. But why did the author contradict the design by adding a 39R resistor and the second BC547 transistor to limit the current? He could have done the same thing with the 330 ohm resistor by using its value to limit the maximum current. The 330 ohms is low, it should be closer to 1000 ohms as used in a typical JT.
The BC337 is a very good choice for the JT transistor, but it has a self-imposed limit on the amount of current it can handle in a JT circuit. Once the 330 ohm resistor has been set to give a maximum LED current of 20mA, then the second transistor is no longer needed. And along with that, the high speed diode, 100u filter cap and 39R resistor can go away. All that’s really needed are the three LEDs across the BC337.
Oh, and another one of my humble opinions. The high speed diode and 100u filter capacitor are not necessary and wasteful of power. The 0.8V or more voltage drop across the diode, multiplied by 20 milliamps, equals 16 milliwatts of wasted power that could be better put to use making light in the LED. When you think of the word diode, remember that the LED is already a light emitting diode. So why would you want to add something you already have?
Last but not least, the three LEDs are connected in series. With each (white) LED having a voltage drop of 3.3 volts, there is about ten volts total across these LEDs when they’re lit up brightly. A normal 1 LED JT has to boost the voltage a bit more than twice (3.3V / 1.5V), but this JT has to boost the voltage 6.7 times. This puts a much higher demand on the coil, which should be low loss and have low wire resistance. The 35 turn primary winding should not be made of thin wire, it should have a resistance 1/4 ohm or less.
Needs More Power Also, you must remember that this JT will not be able to light up the LEDs to full brightness. Let’s do some simple math to see why. The three LEDs require 10V total, at 20 milliamps for full brightness. That is 200 milliwatts, or three times the power of a typical one LED JT. Therefore the input current from the battery must be three times higher – a typical JT uses 80 milliamps, so this JT needs 240 milliamps. The problem is that even as good as the BC337 is, it can’t handle that much current. What this JT needs is either three transistors or a transistor that can handle three times the current. Typical high current transistors are the KSD5041, ZTX651 or NTE11. Or put three BC337s in parallel. Just connect the emitters ans collectors of the three transistors together, and connect each base to its own 1k resistor, with the other ends of these resistors going to the feedback winding.
The coil has a feedback winding with 20 turns, which reduces the negative voltage that the transistor’s base to emitter junction has to handle, which is specified at 5V maximum in the datasheets. But 20/35 times 10V is 5.7V, which is still somewhat over the maximum. The feedback winding should have a few turns less, say about 17 or 16.
Obnoxious I noticed whenever my mouse pointer went over one of the highlighted, underlined words in the text, it shot up a big advert for something totally unrelated to that word. This is really obnoxious and it’s a shame that the blogger/blogspot (owned by Google) has to stoop so low to get people to see their ads. I hope they end up in the same situation as AOL: no one wants to put up with this pestilence.
Back to experimenting…