The Cool Joule is back !

[Slider2732]

TinMan's Cool Joule was something from 2013 that really worked out well.
I built a couple of them, using the 500ohm coils from Dollar Tree waving toys.

Well, 1 build had its coils reused, the other simply sat around in the spare room.
Just before Christmas last year, that one fired up, for some unknown reason. As I walked in and would switch the light on, it would be seen to be running.
It's now 25th March and it's still going !
I haven't dared to move it or take a reading...which will be next.

Neither battery has been changed out or swapped or anything else. It's the very same circuit, in the very same position as back in 2013 !
Another circuit sits to its side, a 'Forever Light', which charges a polyacene 'supercap' by solar during the day and allows it to run 24/7. Could there be some influence there ?
The yellow AAA's from $1 solar garden lights are normally past their prime in any case by now, suffering readily from rust and internal leak damage.
And just to clarify...the original method was to swap over the 'run' battery and the 'charge' battery, when the run battery ran down below the other. These haven't been touched for the last 3 months, since it self kicked back in to running.

Here's my original video from April 2013 of this circuit:
https://www.youtube.com/watch?v=J7Fxr3OP1sU

Attached is a pic from last night of it running (is the one on the left).

[Chet K]

well seems your wee beast was awakened

There is now a moderated [TinMan and Magluvin]section at OU.Com where this will be discussed in great detail

TinMan quote from today
Snip"
I will also be putting the cool joule circuit back together,and having another look at that,and the effects of miller capacitance,where the circuit !can! operate using this effect.This will be done in the thread linked below.

http://overunity.com/16486/resonance-circuits-and-resonance-systems/new/#new

thanks

Chet

[Slider2732]

Good timing then...
Thanks for the heads up.

It's also been realised, as of 5 minutes ago, about which weekend this is ! and I truly hope noone gets offended by the phrase 'back to life' 

Am just saving out a public YouTube vid of this, which i'll post here.
Readings are 1.147V on the left batt, 211mV on the right side one.

[Slider2732]

Here we go:
https://www.youtube.com/watch?v=60eF3XyWfew

It quit while measuring the voltages and might well relate to the Miller capacitance ?
These used to be very finicky about any metals being touched at all while running.
Got it running again though, in the old way of touching 2 transistor legs and...onward it goes.
They might fire up using a BBQ piezo lighter, will have to try that at some point. If that works, then the theory of a close lightning strike starting this one up might be right.

[TinMan]

Well i have managed to put together a cool joule that dose not mind having scope probes placed all around the circuit.

As most of us that have built the cool joule,it is very touchy-as Slider said. But it seems that the two larger coils i have used in this build,has eliminated it's sensitivity,and now i can scope all of the circuit without it stopping.

Below is the circuit diagram,along with scope probe placement for the attached scope shot.
Anyone see anything odd here

Will do up a quick video of the build,and then start looking at some mod's.

If ION or Poynt get some time,i would like to solidify the operation of this circuit.
I know that it has always been thought to be operating due to the miller capacitance effect,as thought by MarkE as well. But now that i have my digital scope,and a test platform that will accept my scope probes all over it,i would like to confirm the operation of the cool joule. It seems to me that it may in fact be a combination of the transistors junction capacitance,and also(if not- more so)the coils capacitance.

The reason i say this-looking at the scope shot below,and where i have the probes placed,it would seem to me that the only way we can have a 0v level(even slightly into the minus)across the emitter/collector junction before the transistor actually switches on(going by base voltage),would be due to capacitance in the L1 coil(primary coil) being discharged along with the field collapse of the L1 magnetic field.

Anyway,video coming soon of the new rig.


Brad

P.S--i have both scope probes set to 10x,and also have the scope channels set to 10x as well--so as the voltage values on the scope read correctly.

[ion]

There is not much to say about this circuit that we have not already covered. From the scope shots it seems the base drive is rather soft, due either to a minimum of coupling between the coils or Miller effect just not having enough current drive to effectively pump the base from the resultant L-C oscillator.

Can you completely define what task you expect the circuit to perform, and why.

Some time ago I built a similar inductively decoupled circuit using an FET, but it snapped on a lot harder than your transistor version. It required a higher voltage for startup as expected and a gate biasing arrangement.

Personally, I never liked the name Joule Thief, preferring the original term which is "blocking oscillator". From watching the videos by the guy that invented the term, with all due respect to him, he never really addresses the real engineering aspects of designing such a circuit, such as optimizing base drive, and how and why the turns ratio and core material type is important.

Many variants of the humble transistor blocking oscillator have been used in many millions of consumer products over the last 50 years but it's origin is from the vacuum tube era. You will find them in small fluorescent tube emergency or camping lights, camera flash assemblies, HV boost converters etc. It is ubiquitous in it's many incarnations, all tailored specifically to the task at hand.

The  desired end application of the blocking oscillator will determine it's ultimate design and no single version is best. No one type fits all applications.


If the intention is merely to steal a little of what is left from a nearly dead single cell to  light an LED, a well designed blocking oscillator will work just fine, but you could also just stack a few of your nearly dead cells in series to light the led (the no component approach). By the time the cells are that dead, the internal resistance is fairly high and will nicely do the current limiting.

Well designed blocking oscillators have been used in Tektronix vintage scopes, the circuits of which are a good education in the art.

So best to define what you want to do (the job at hand) then design the circuit to fit this and still operate as efficiently and economically as possible.

Good engineering considers all this and balances it against the cost of perhaps some extra components, which spread over many millions of produced units can be considerable, or may not be necessary, depending on the market you are competing in.

If you are shooting for the most efficient design for a given application, and the cost is not an issue, you have a wide choice of topologies and components.

Regards
ION

[Slider2732]

Back to this with an update -
The circuit is still running !

Voltages now, are 1.119V on the left and 0.149V on the right.
On the 25th March they were 1.147V on the left batt, 0.211V on the right.
So several mV have been used, but also several mV will have been naturally lost due to the inherent losses of Nickel Cadmium from dollar store garden light batteries.
I imagine that ferrous oxides will start to become factors with moving toward the 100% humidity months. That's kinda tongue in cheek, but might well be true because of the low quality of these cells.
28mV + 62mV. So what's that, under 5mV per day combined on these 100mAh 1.2V's ?
I can't imagine that the right side one is doing anything more than offering some internal resistance nowadays. It could likely be replaced with a resistor.

Yes, to agree about what the aim of the circuit is, which will dictate the design.
Here (for me) it's to have a 'Forever Light' running, just trundling along always going, not degrading and to keep on keeping on for years.
An electronic canary is one idea for a practical use, while the light output isn't good enough for even a night light on this build. Something where too much electrosmog snubs out the circuit, sitting there, no maintenance needed.
It does form a great simple circuit for adding atmospheric collectors to, or other tiny ambient energies. A harvester base. As such, the variable resistor of the original design could be reduced, for more light with the extra coming in.