Question on toroids and Joule Thief

[Voltage Spike]

Yes, thats how some computing actually works.

Single cell (1.2 or 1.5V) LED inverter (Joule thief)

danyk.cz

That is like what I am going for. Just going to use a toroid for the ferrite. Thare are numerous designs on the internet but I am just wanting mine to be more efficient. I will report back once I get testing but it won't be today.

 

[ChrisP58]

The Joule Thief is a variation of a blocking oscillator. Here are wiki's for each of those.

Joule Thief https://en.wikipedia.org/wiki/Joule_thief

Blocking oscillator https://en.wikipedia.org/wiki/Blocking_oscillator

Also, you can find more discussions about the subject that have been had here on this forum through the search function at the top of this page, as well as the Similar threads at the bottom of this page.

 

[Voltage Spike]

The Joule Thief is a variation of a blocking oscillator. Here are wiki's for each of those.

Joule Thief https://en.wikipedia.org/wiki/Joule_thief

Blocking oscillator https://en.wikipedia.org/wiki/Blocking_oscillator

Also, you can find more discussions about the subject that have been had here on this forum through the search function at the top of this page, as well as the Similar threads at the bottom of this page.

Thanks Chris. I already knew about and read the wiki article on the joule thief as I had seen that BigClive gave it that name, and it stuck. But I will read the other link about the blocking oscillator for extra credit. I think the problem is I have read too much but until I start making one or more to compare I won't find the answer to efficient core type, winding wire size and turns, and size of toroid. I want it small but efficient as for now it's all about LED and so called dead batteries and length of on time. Doing is going to answer a lot!
VS

 

[Buk]

Successive approximation methods have there place in computing maxima and minima in complex differential equations, but this is electronics not computing.


The link you gave provides a parts list for a known working solution. You can obtain the pieces and put it together like lego or paint by numbers.

It doesn't answer the OP questions:

1. Best toroid type to use? Many different options out there when it comes to colors and specs.
2. High or low permeability?
3. Wire size smaller or larger with more or less wraps? What range in uH values are best for efficiency?
4. Single layer wraps only or can I wrap over existing wires when making my toroid?

You gave a list of equations earlier in the thread; but you gave no definitions of what those equations were, or how they might be used to answer the questions. Nor even any indication that you know how to use them.

 

[MacIntoshCZ]

That is like what I am going for. Just going to use a toroid for the ferrite. Thare are numerous designs on the internet but I am just wanting mine to be more efficient. I will report back once I get testing but it won't be today.

Core losses increase with peak value of magnetic induction and by higher frequency. Different cores has diferent losses. Number of turns affects inductance and inductance affects resonance frequency. Losess in resistance can be calculated as R*i*i ...
Number of turns should be large enough to stay in linear region ( stay away from Bsat/2 point)
Btw you need to use formulas for sine waves, since you are using resonants topology.

 

[MacIntoshCZ]

Successive approximation methods have there place in computing maxima and minima in complex differential equations, but this is electronics not computing.


The link you gave provides a parts list for a known working solution. You can obtain the pieces and put it together like lego or paint by numbers.

It doesn't answer the OP questions:


You gave a list of equations earlier in the thread; but you gave no definitions of what those equations were, or how they might be used to answer the questions. Nor even any indication that you know how to use them.

Sorry i am not able to provide solution. And i dont even want to. he will not learn anything and i can not lead his every step remotely. But i can try to provide a direction.
Since yours english is way better than mine i bet you will fill gaps here...

 

[Buk]

until I start making one or more to compare I won't find the answer to efficient core type, winding wire size and turns, and size of toroid. I want it small but efficient as for now it's all about LED and so called dead batteries and length of on time.

Try playing with this: https://tinyurl.com/yzb7b5f7

Try varying the base inductance; the base resistor value; and the turns ratio of the transformer and note the effect the changes have on the frequency & duty cycle of the LED. Then try reducing the voltage level until the led no longer lights.

You can learn far more, far more quickly about the effects the components have using a simulation, than you can building half a dozen actual circuits.

Then you can try relating your observations to the formulae MacIntoshCZ posted earlier.

 

[Voltage Spike]

Try playing with this: https://tinyurl.com/yzb7b5f7

Try varying the base inductance; the base resistor value; and the turns ratio of the transformer and note the effect the changes have on the frequency & duty cycle of the LED. Then try reducing the voltage level until the led no longer lights.

You can learn far more, far more quickly about the effects the components have using a simulation, than you can building half a dozen actual circuits.

Then you can try relating your observations to the formulae MacIntoshCZ posted earlier.

Buk,
Thanks for the link to the simulation. That is really cool. I have heard of LTSpice but never used it yet but I think it is free. I also read a lot where LTSpice was not able to properly simulate some circuits but that could be from people that don't know how to use it, or how to make a proper circuit. I have been playing around with it and already find it very useful. I changed the base inductance, resistance, the LED type, the current, but didn't know I could change the turns ratio of the transformer. I will go back and look at that. Question: What does the "u" mean as in 100u in the simulator? Is that the same as uH?
Thanks

p.s. I know the online simulator is not LTSpice but I brought that up as another simulator that I could use if I could figure it out.

Did you create that circuit? I was looking in the menu and there are all kinds of circuit simulations including a joule thief (under the transistor section). But once I looked around at the other circuits I couldn't find (in the menu) the circuit you sent me in the first place. Was it one already in the menu??

 

[MacIntoshCZ]

Buk,
Thanks for the link to the simulation. That is really cool. I have heard of LTSpice but never used it yet but I think it is free. I also read a lot where LTSpice was not able to properly simulate some circuits but that could be from people that don't know how to use it, or how to make a proper circuit. I have been playing around with it and already find it very useful. I changed the base inductance, resistance, the LED type, the current, but didn't know I could change the turns ratio of the transformer. I will go back and look at that. Question: What does the "u" mean as in 100u in the simulator? Is that the same as uH?
Thanks

p.s. I know the online simulator is not LTSpice but I brought that up as another simulator that I could use if I could figure it out.

Did you create that circuit? I was looking in the menu and there are all kinds of circuit simulations including a joule thief (under the transistor section). But once I looked around at the other circuits I couldn't find (in the menu) the circuit you sent me in the first place. Was it one already in the menu??

if its used for coil, than yes it will be msot likely uH.

 

[Buk]

Thanks for the link to the simulation. That is really cool. I have heard of LTSpice but never used it yet but I think it is free. I also read a lot where LTSpice was not able to properly simulate some circuits but that could be from people that don't know how to use it, or how to make a proper circuit. I have been playing around with it and already find it very useful. I changed the base inductance, resistance, the LED type, the current,

YW.

but didn't know I could change the turns ratio of the transformer. I will go back and look at that.

If you select Draw->Passive components->Custom transformer; draw one, and then double click to edit it, there is a label "description" in the dialog that is a link to a help page that explains the nomenclature for defining any kind of transformer you might think of.

Question: What does the "u" mean as in 100u in the simulator? Is that the same as uH?

The u 'suffix' is actually an SI units prefix, where the unit has been omitted.

'u' generally means e-6 or 'micro'. In the context of an inductor it means 100uH or 100e-6H or 1e-4H; for a capacitor: 100uF or 100e-6F or 1e-4F; a resistor 100uΩ ....

An m (milli) means X.Xe-3; n (nano) is X.Xe-9; p (pico) is X.Xe-12.

p.s. I know the online simulator is not LTSpice but I brought that up as another simulator that I could use if I could figure it out.

I'm quite certain that LTSpice is a much more accomplished simulator. If I ever get my Master's in Arcane UI design and work out how to use it, i'll confirm that.

Meanwhile Falstad serves as a way to explore simple circuits and also share them.

Did you create that circuit? I was looking in the menu and there are all kinds of circuit simulations including a joule thief (under the transistor section). But once I looked around at the other circuits I couldn't find (in the menu) the circuit you sent me in the first place. Was it one already in the menu??

The circuit I linked is basically the same as the one MacIntoshCZ linked above -- minus the capacitor that the author notes "can be omitted", that actually has no effect --regardless of value -- on the circuit at all that I could find.

Perhaps it would in a more faithful LTSpice simulation.

 

[Voltage Spike]

YW.


If you select Draw->Passive components->Custom transformer; draw one, and then double click to edit it, there is a label "description" in the dialog that is a link to a help page that explains the nomenclature for defining any kind of transformer you might think of.



The u 'suffix' is actually an SI units prefix, where the unit has been omitted.

'u' generally means e-6 or 'micro'. In the context of an inductor it means 100uH or 100e-6H or 1e-4H; for a capacitor: 100uF or 100e-6F or 1e-4F; a resistor 100uΩ ....

An m (milli) means X.Xe-3; n (nano) is X.Xe-9; p (pico) is X.Xe-12.



I'm quite certain that LTSpice is a much more accomplished simulator. If I ever get my Master's in Arcane UI design and work out how to use it, i'll confirm that.

Meanwhile Falstad serves as a way to explore simple circuits and also share them.



The circuit I linked is basically the same as the one MacIntoshCZ linked above -- minus the capacitor that the author notes "can be omitted", that actually has no effect --regardless of value -- on the circuit at all that I could find.

Perhaps it would in a more faithful LTSpice simulation.

Thanks Buk for all the good info and that simulator. I was playing around with it quite a bit and will do some more tomorrow. Some of that math hurts my head though but I can generally get through it will looking at an online converter etc. When I was a kid I never liked math. Then when I got into computer programming I wished I would have liked it more when I was young. But I did like I always do and bought some math books and figured out whatever I needed at the time. Now electronics has even more interesting math.

 

[Buk]

Some of that math hurts my head though but I can generally get through it will looking at an online converter etc. When I was a kid I never liked math. Then when I got into computer programming I wished I would have liked it more when I was young. But I did like I always do and bought some math books and figured out whatever I needed at the time. Now electronics has even more interesting math.

I'm in much the same boat. I did the math in college but then barely used it during my career. I implemented a mass of code that did very complex math. Complex in both senses of the word -- I even implemented a couple of limited complex math libraries in assembler and C -- but for the most part, the equations I was implementing were either devised by others or taken from reference books, so I rarely had occasion to consider how they did what they did. Just implement, verify with some test input and move on.

The same with electronics. What little formal education I had on the subject was assimilated enough to allow me to pass the courses and then mostly never re-visited. Lately I've had a bee in my bonnet about an idea or two that require electronics that cannot be bought off the shelf, so when I discovered Falstad a few years ago, it became my go to place for trying out circuits I found on line and exploring themm until I understood them enough to modify them for my needs.

I'm currently playing with an idea that requires the use of a switched voltage upscalar and RC resonant circuits. Hence my interest in this thread. I'm far from an expert -- just barely ahead of you I think -- but things are beginning to gel slowly.

 

[MacIntoshCZ]

but things are beginning to gel slowly.

tell me about it... I would need several hands and minds to complete everything i am interested in.

 

[Dick Cappels]

People often have a lot of worry about the core and make it harder than it really is for this non-critical circuit.

A nail does not make the best core for a switching power supply -that is why the 2700 pf cap is needed, but see how easy it can be?

This person wound his tapped inductor on a piece of wood!

This fellow wound his tapped coil around the lamp enclosure.

The above lights used a circuit that is electrically equivalent to the Joule Thief but does not use separate windings for base and collector circuits.

 

[MacIntoshCZ]

Thats briliant idea, no more buying of nanodust cores. I will use nail, becouse losses are not what i am aiming for =D.

 

[Buk]

People often have a lot of worry about the core and make it harder than it really is for this non-critical circuit.

But this non-critical circuit can be optimised to waste less mW.Ys.

And the same process of optimising the windings and core for this purpose, can then be applied to more critical circuits.

 

[Voltage Spike]

I'm in much the same boat. I did the math in college but then barely used it during my career. I implemented a mass of code that did very complex math. Complex in both senses of the word -- I even implemented a couple of limited complex math libraries in assembler and C -- but for the most part, the equations I was implementing were either devised by others or taken from reference books, so I rarely had occasion to consider how they did what they did. Just implement, verify with some test input and move on.

The same with electronics. What little formal education I had on the subject was assimilated enough to allow me to pass the courses and then mostly never re-visited. Lately I've had a bee in my bonnet about an idea or two that require electronics that cannot be bought off the shelf, so when I discovered Falstad a few years ago, it became my go to place for trying out circuits I found on line and exploring themm until I understood them enough to modify them for my needs.

I'm currently playing with an idea that requires the use of a switched voltage upscalar and RC resonant circuits. Hence my interest in this thread. I'm far from an expert -- just barely ahead of you I think -- but things are beginning to gel slowly.

I never went to college so you got me there. My dad had nine kids and I was one of them. Two oldest girls got the chance to go to college went into nursing. Ran out of money for the rest of us. But they paid back their loans so all was good at least in that regard. I was always the kid that fixed everything. Took things apart to see how they worked etc. Took peoples trash (like a broken bicycle) and fixed the problems and got a free bike. I had a TI994A for my first computer/game console but couldn't afford the add-ons to make it really useful as a computer so I just played games with it. Then I got a Commodore 64. Then I started programming is BASIC. Eventually I ran a BBS using the C64 and had it connected to LTK SCSI hard drive(s). Too bad only one phone line but still great fun. Then life took over and got in the way of all my spare time. I have a large workshop now with many machines and stuff. I built a powder coat booth, I do welding (TIG), have a 9x49 mill, a lathe, a shear, sandblaster, and fix just about anything. But with all that (and a few 3D printers) I still have that strong desire to mess with electronics. Somebody once told me that my "hobby" was "collecting hobbies". I wrote some shareware software for hand crafted soap making too. My main problem now is the workshop is stuffed full of everything and has no air conditioning as it is too large to install and pay for that. And my spare room (converted single car garage) is full of our last business (soap making stuff) and I want to set up my electronics area (work benches etc) so I have to make room. The stuff in there has to be climate controlled so I need to add on to the house (bad time for that) or buy a cargo container and modify it with air conditioning so I can move all that stuff out of my way. And I think it's a bad time to buy a cargo container too. So I end up working at the kitchen table or countertops and have to move all my stuff all the time. I sure would like to set up my high quality microscope so I can see surface mount (SMD, SMT) stuff. https://tinyurl.com/u37re8rd And I have a 3/4 finished reflow oven done https://www.whizoo.com/controleo3
So if I seem a little slow at getting this done, that's why. But I am waiting on a few things from ebay and know I can slap one of these together in less than an hour, but I am looking for efficiency. The buzzer just went off, my time is up.

 

[Voltage Spike]

People often have a lot of worry about the core and make it harder than it really is for this non-critical circuit.

A nail does not make the best core for a switching power supply -that is why the 2700 pf cap is needed, but see how easy it can be?
View attachment 134261

This person wound his tapped inductor on a piece of wood!
View attachment 134262

This fellow wound his tapped coil around the lamp enclosure.
View attachment 134263
The above lights used a circuit that is electrically equivalent to the Joule Thief but does not use separate windings for base and collector circuits.

View attachment 134264

Hi Dick. I have a boat load of links and schematics etc but I appreciate your input as the more the better. My main concern is length of run time for my experiment. I also wanted that to by with full on and not blinking but I see your example are all full on, just not efficient. I have always been confused with inductors as I never used them but it finally sunk in how valuable they really are. Now I just have all the noob questions on toroid type, size, turns and wire size, permeability, and resistance. And I see its gonna take some hard work until it all makes sense as there are so many joule thief designs and they are all different mainly with the toroid used and thereby the efficiency obtained. Do you know any good tips on toroid making?
Thanks

 

[Voltage Spike]

But this non-critical circuit can be optimised to waste less mW.Ys.

And the same process of optimising the windings and core for this purpose, can then be applied to more critical circuits.

That one is very efficient. Too bad its a blinker. But nonetheless it is a keeper!

 

[Buk]

Somebody once told me that my "hobby" was "collecting hobbies".

I think my wife might hold a similar opinion rearding me. E-bikes, leather craft, 3D printing, knife making, carving, wood turning re-building my 30 year old car; not to mention programming, metalwork...

But she's tolorant because in this last month I've repaired the lid for her out of production but better than anything now available spice grinder. Repaired the boiler that would have cost £3000 to replace. Made a new battery clip for the watch her mother gave her. And this last week, done a 3D jigsaw around the existing water, gas and drain pipes to replace the kitchen cupboard under the sink that got severely water damages by a leaking tap.