Joule Thief

[TinselKoala]

The NE-2s are in _series_ and the probe is across the series stack. The circuit is below, it's a standard JT, more or less, with a special inductor. Sorry, I didn't indicate the phase dots on the inductor diagram.

ETA: the extra components near the battery on the photo of the board are a wireless power receiver, not in the circuit at all for these tests, and of course without its antenna loop installed.

[MarkE]

The NE-2s are in _series_ and the probe is across the series stack. The circuit is below, it's a standard JT, more or less, with a special inductor. Sorry, I didn't indicate the phase dots on the inductor diagram.

ETA: the extra components near the battery on the photo of the board are a wireless power receiver, not in the circuit at all for these tests, and of course without its antenna loop installed.

The 1:10 step-up transformer explains the 10X voltage.

[ramset]

TinselKoala
here you have a few interesting Vids that show some interesting effects
as a builder yourself ,would you explain this acceleration in the second video
to some of our builders here working on acceleration effects and voltage increases


I know the second Vid [with the motor ] has other "snake oil" purposes ,but you also are showing acceleration under load,
it would be good to discuss this here,I know other builders respect your work.

Voltage  https://www.youtube.com/watch?v=FLsOCYPCvGc


acceleration  https://www.youtube.com/watch?v=9x2YfA9LU5s


When you have time and in the appropriate Thread ??


respectfully


Chet

[TinselKoala]

The 1:10 step-up transformer explains the 10X voltage.

And if you ask Bill or other Joule Thief builders for their optimized coil turns ratios in specific applications, you will find that some of them are also in high proportion, thus producing high voltage at the outputs. In fact, only the very most "basic" JTs will use a 1:1 turns ratio, like my testbed's baseline toroid with 13:13 turns.

Here's one of mine, from a magpwr/conradelectro circuit, with 5:80 ratio on a random "bead" toroid, lighting a LED with just over 250 mV input. It still works down below 150 mV but I didn't show that in this video.  With the proper ferrite material as specified by magpwr and conrad, 4 FETs in parallel and the 4:100 ratio it should work down to 50 mV:

http://www.youtube.com/watch?v=35IMWNp2akg

[MarkE]

And if you ask Bill or other Joule Thief builders for their optimized coil turns ratios in specific applications, you will find that some of them are also in high proportion, thus producing high voltage at the outputs. In fact, only the very most "basic" JTs will use a 1:1 turns ratio, like my testbed's baseline toroid with 13:13 turns.

Here's one of mine, from a magpwr/conradelectro circuit, with 5:80 ratio on a random "bead" toroid, lighting a LED with just over 250 mV input. It still works down below 150 mV but I didn't show that in this video.  With the proper ferrite material as specified by magpwr and conrad, 4 FETs in parallel and the 4:100 ratio it should work down to 50 mV:

http://www.youtube.com/watch?v=35IMWNp2akg

The lowest voltage harvester that I have seen was a 25mV input from a thermopile disclosed in a thermostat patent years ago.  That oscillator also used JFETs and a a pretty tall step-up ratio.  IIRC it was about 200:1.