Joule Thief

[innovation_station]

i finally got the knot wound ....

it looks  kinda kool ..

ist

i have a question ...  do you think is it sheer coincenidence that you live in a 60hz grid?  i yi yi

so i have another question ..might it also be sheer coincenidence that 60 hz flicks 60 times per second ...  and it is allso odd there 60 min in an hour ...   ever think bout that ... and we record time .. on 60hz... 

maybe i just lost my mind ..  lol   dont think so tho .. 

[PaulLowrance]

IST,

That looks cool, but scary. 

Hey, I just thought of a very simple way to test the efficiency of my JT.

Paul

[PaulLowrance]

I finished my quick & dirty JT efficiency test, but it's inconclusive. I already know what the skeptics will be thinking before hand. Yes, yes, yes, I know, it's just a quick & dirty method that has flaws, but I could not resist since it's so quick & easy.

Anyway, it showed 1.03% efficiency, which is far higher than I'd expect, but remember there's a problem with my method. I'm not doubting that some people here might have found the right combination for cop>1. Lets find out.

What I did was placed two LED's facing each other, and taped them together. One is the light source, the other is only a sensor. So had the JT light the LED till the sensor output was 50mV. Then I did the control experiment by replacing the JT with a DC voltage source from a power supply, and adjust the voltage till the LED sensor produced the same 50mV.

Here are the numbers,

JT circuit: 0.634V, 15.47mA = 9.81mW

Control: 1.887V, 5.35mA = 10.1mW

Efficiency is 103%

The problem with this method is that the LED efficiency varies with current. The JT circuit produces a pulse, and therefore the LED current is much higher as compared to the control experiment, which is DC current. I have no idea if the LED is more efficient above the control DC current of 5.35mV. It probably is.

Any other ideas how to test this without getting out the scope and messing with spikes?

Anyhow, this is fun testing, and I'll try some more. The JT is pretty cool circuit. Conventionally speaking, one would probably expect this to be 80% to 90% efficient, maybe a bit higher if special care was taken.

Paul

[PaulLowrance]

Ah, I know, the thermistor method. Replace LED with a resistor, and then place thermistor on resistor.

Paul

[broli]

IST it seems you propose to tune the frequency at the mechanical vibration of the core. I don't think this is as easy as you make it out to be without using any timer circuit, also remember you also have years of knowledge in this compared to others. I think if this gets nailed this forum should host tutorials from a to z describing each step with photographs to build the "perfect" joule thief. Tesla's research was all about natural vibration and amplification not only electrical resonance as the hertzians think so tuning a JT to its natural vibration might amplify the power like you say.