Joule Thief

[xee2]

@ gadgetmall 

That is Why i put a Vr

A variable resistor is certainly easier to tweek for best performance. But I find fixed values are better to just get an idea of what the circuit is doing.

If i have a 680 pf across a variable resistor and use only wiper and one end  will it adjust the capacitance too ?

I do not thinkd so. But I am not sure..

I have lots of Darlington's from the old power supplies . They Hook up the same Correct . I all ways used two xsistors .  .

Yes. The Darlington has an emitter, base, and collector just like a single transistor. Maybe you can find one that works better than the one I bought. You seem to be doing better with the germaniums than I was able to do.

[xee2]

@ jeanna

Capacitor codes are a pain. It seems that every manufacturer wants to do something different. In my experience a cap marked 680 is a 68 pF cap. Here are two links in case you do not already have them.

http://wiki.xtronics.com/index.php/Capacitor_codes

http://www.csgnetwork.com/capcodeinfo.html

I am still puzzling over your test results. I do not know if you have the parts, but if you do you might try using something around 5K for resistor and a 1000 pF for capacitor on the base as in my circuit. For me this only drained 4 mA from battery and provided 88 volts for an output which should light a lot of LEDs. How much the circuit is dependend on the toroid I do not know yet (future tests). It does seem that the transistor makes a difference in performance.


EDIT: I think your base resistor has too low a value. With a 1K base resistor I was using about 25 mA from the battery. So your results are not that different than mine. If you go to a 5K or higher resistor your battery current should drop.


EDIT 2: Your base cap may be too small also. If the cap is less than 1000 pF then the base resistor has to be even larger. I am not sure what best value for base casp is. !000 pF was just my first try.


EDIT 3: Toroid is critical. With R=10K  I get 3 mA and 54 vots out with my test circuit toroid. With a different toroid with 20 turns and 20 turns I get 11 mA and 30 volts. Test toroid is 2" diam and 3/4" thick of unknown material (maybe laminated steel) and other toroid is 1" diam by 1" long of ferrite. Test toroid is from power supply and is coated with epoxy so it is not possible to see material it is made of.

[Mk1]

@all

Some more .

[jeanna]

@xee2 (mostly),

OK I got the amps down but the lights on the secondary are hard to see, and the light when it is in the jt is pretty dim too.

I did not have a 102 but I had a 103, so I used that and I had a 4.5k R (actual 4.6k) for the base so I used that
Also, I have a 47uF cap at the battery.

So, here it is.

Battery 1.3v
R to see amps at bat 9.5r
Cap at battery 47uF
Transistor N3904
Cap at base of transistor 103=10,000pF
Resistor at base of transistor= 4.6Kr

3 lights on a secondary array with a 2 turn secondary (but I checked rectified voltage with a 1 turn secondary - so I didn't go mad)
or
1 light in the usual joule thief spot
lights are yellow and show really well how dim they are when there is not much to light up with. 

At the battery
Amps grew a bit over time from .000842A to .00126A
Across 47uF cap across whole ckt=1.29v

At the transistor,
across light 1.30v
across the (10,000pF) base cap 1.17v

the secondary wire with a single turn .025v rectified... then I removed that 10r resistor at the battery just to see and the secondary wire with single turn returned .046v rectified.

Also, I checked the transistor itself
base to - bat 0.165v
base to +bat 1.14v

I guess that is all. It is pretty unimpressive. Maybe the big difference is in the transistor?

Do you have a N3904 that you could put into this and see?

Since your toroid is wound 10:40 I actually think that is making a big difference. These numbers may be right in line given the 6 turns bifilar on my toroid. It looked great without the caps, and I got more rectified voltage and it took a lot more amps!  if I can remember back that far! 

thank you,

jeanna

[xee2]

@ jeanna

See my edit 3 above.  I think it takes optimizing of transformer to get best performance. I do not know what works best but I think it is important to have lots of turns on the collector coil. The base coil does not need so many. I am sorry this is not working better for you but now you have a new challenge.