Joule Thief

[xee2]

@ jeanna

It is a big clue to what Tesla discovered. 

I agree. This is a basic part of much of what he was doing. That is why I usually call it "one of Tesla's magic tricks".

[jeanna]

... I do not want to put an led on the primary side, just like I did in my video "Earth Battery Lights 7 LEDS".

It is a good idea to put a led in the usual led spot (BJTL) to be sure everything is working and there are no surprises. Then take it out. that is fine.

I am wondering if, after I put the wires that would normally go to the led into the breadboard power rail (like I did in the video) if I can then also put the secondary wires there too? 

The secondary is producing AC, so you won't get it all unless you rectify it.
So, my question always is, why bother with making the ac if you are just going to rectify it.

AC has special properties which we in this modern world (except for theNOP of course) don't know much about.

It is this secondary that will light 30 to 100 lights for little more than the amps cost of 1 and often the amps draw actually goes down when you do this.

OH Bill, I am so glad you are going to do this!

It seems to me that if I get any higher voltage from the secondary side, I should be able to place this in parallel with the primary power output on the rails and, it should be "in sync" as it is running of the same 3904 transistor.  I am trying to get all of the volts out I can for another earth battery experiment.  Has anyone done this that we know of and/or can it be done?

I am not sure, but it sounds a lot like my 2 tier circuit.

I know I need thinner wire for a real secondary but, I figure if I have enough winds, it should step up a little anyway.

You will be fine. The thinner wire allows for more turns but 20 or 30 is ok to see the concept. Remember it was only 1 turn that lit a led in january. I figured for good brightness it was best to plan about 3 to 5 turns per led when making a secondary.

jeanna

I am trying to post my results from today and the site won't take my post. Hopefully soon.

[jeanna]

@xee2,
I made another chart and added your chart to it and the other without the change for comparison.
I cannot seem to get the site to do anything with it. It says done, but nothing happens. I will try later.

Basically the results were not anything special.
I killed a bunch of lights. I guess 41mA is too high.
I was drawing 41mA from the battery when I had the led across the secondary wires. The volts were 80v or 64v and freq 33KHz or 55KHz so in a range that has not killed lights recently. .

The only place with the low mA draw was with the 104 and the led at the bjtl spot.
The waveform on that one with the 104 cap is weird and I could see it today with the 2TB,5TC92T setup along with that 104 cap.

It is a camel hump.

Both peaks of a hump pair go to the same height 91-96 volts, but the low of the wave goes to about the middle and returns up to the top again. This is probably why one of my number sets looked off yesterday. The scope was having trouble with it, I guess. I will continue to try to post this this evening.

jeanna

Please note: My two charts were made on different days with different battery voltage or strength. I put this here for reference, but it is only helpful, but too bad I didn't do it all at the same time..

[electricme]

@Jeanna,

I killed a bunch of lights. I guess 41mA is too high.
I was drawing 41mA from the battery when I had the led across the secondary wires. The volts were 80v or 64v and freq 33KHz or 55KHz so in a range that has not killed lights recently. .
jeanna

I think you might have to begin to include a "protection" resistor in your chain of LED's.
I have'nt said anything about this before, but have been waiting for the above results to occur.

You are getting blown LED's because you have been able to increase the current by refining the design of the toroid, with the high voltages.
Before hand, you have been running the LED's with high volts, and hardly any mA's

The protecting resistor is a current limiting device, it will stop the LED's from being over driven and being destroyed, what size/value to use I wouldn't know as the higher voltages and frequencies your LED's are being driven at is a fluxating quantity, so if you want to see bright LED's use a lower value, if you want very good current protection, use a higher value.
I suggest you put a adjustable pot (say 5k ohms) and a (separate DMM set to mA) in series with the LED string, that way you could alter the LED brightness and monitor the current drawn.

The same principle is used to protect Neons from destructing from higher voltages, as was posted a few pages back.

hope this helps

jim

Ps set up a single LED with a DMM set on mA, add a 1K ohm pot, (press the Hold feature to capture the highest result) and take the LED to destruction, note down the reading.
This will give you a basemark on what to keep an eye out for.
I know it will destroy the LED, but better to fail 1 LED than a whole string of them.

 

[xee2]

@ jeanna

  "I will call this one xee's LED killer"   

Yes it is dangerous. As you get higher output voltages this circuit will become more dangerous. I use neons instead of LEDs when voltage is over 100 volts (they can survive several thousand volts for a short time). I try not to use LEDs if voltage is over 40 volts. Be careful not the fry the scope or DMM. But if you are going to light a fluorescent tube you are going to need about 400 volts. BE VERY CAREFUL. Always use your /10 probe.

Your 2TB,5TC,92T2 with 104 cap and 220 resistor yields 91-96 volts @ 41 mA. My 2TB,5TC,100TP with 104 cap and 250 resistor yields 161 volts @ 36 mA. To get over 400 volts I needed to use a 100 ohm resistor and 247 mA with the 104 cap and 288 mA without the cap. I do not think your 2N3904 transistor will survive that.

I suggest you wait until you have some NE2 neons and then try using the 2N3055 transistor - if you want to continue to try to make a circuit to light a fluorescent tube. At this point it looks like your toroid will work but it may need extra turns.

What ever you do - good luck and be careful.


EDIT: I think that more pickup coil turns will reduce the current required, but this is yet to be proven.