General JT Topic

[stprue]

Here is the video in regards to the data I have been posting. 

I apologise ahead of time that you cant see the blinking. 

I will make a quick follow up video showing the LED blinking!

[szaxx]

Hi ST,

Not too sure if this would work but, could the LED be series wired to charge the main cap as it flashes?

simple but effective regeneration, maybe incorporate that inductor too somewhere to perhaps increase the voltage on the charge/flash cct.

Its your baby and getting better.

Cheers Steve.

[stprue]

Thank you for the comment Steve! 

I will try this test and time it!

[stprue]

Ok I'm have not had any inspiration in a while and I'm bored so I have been further tweaking my BJT

Cap charged 1500uf @1.5v=1.688mJ

Test Finished at .573v which is .246247mJ

so 1.688mJ-.246247mJ=1.441753 over 611sec

2.3596uW

Draw=1.573uA

The difference is that it will run a little longer and and down to a lower voltage.

[stprue]

Ok I have tuned my novel device/toy to where I want it!!!!

I have an oscillator BJT circuit that will power an LED indefin.....well with normal circumstances 

The BJT is a Kooler mutation but will run on a 1.5-2.3v mini calculator solar panel as long as the components last.

The solar charges an already charged 2.3v@10F max cap and will light an led (not bright) all day and through the night.

The OSC range is around 20hzs-7Khz output

If you choose to build this please try Koolers easier BJT's as pre-rec because this is very tough to tune!!!!!.....but I built this off his most successful version so use his schematic 

[poynt99]

Joule Thief enthusiasts,

For those seeking further challenges and improvements with this project, here is a circuit that claims to allow battery operation down to 250mV.

Credits to the author: Louis Vlemincq of EDN Magazine. (I put it all together in a convenient PDF.)

.99

[BEP]

.99,

Kudos for the compilation.

This looks like a great little circuit. I seem to recall posts about folks being satisfied about perpetual motion if only they could light an LED with no batteries (probably on another forum).
This could do it. Power it off of a few series connected thermocouples at room temperature or wind a large air-core coil with a detector diode. Either should suffice as a power supply for this circuit.

Of course, today this won't be OU because now we know how thermocouples and EMI detection works.

[ion]

.99,

Kudos for the compilation.

This looks like a great little circuit. I seem to recall posts about folks being satisfied about perpetual motion if only they could light an LED with no batteries (probably on another forum).
This could do it. Power it off of a few series connected thermocouples at room temperature or wind a large air-core coil with a detector diode. Either should suffice as a power supply for this circuit.

Of course, today this won't be OU because now we know how thermocouples and EMI detection works.

Thermocouples are inherently differential measuring or emf generating devices. The "hot" junction is subtracted from the "reference junction". As such you need a source of heat and sink to generate emf.

[BEP]

ION,

What you say is true but thermocouples will generate an EMF when heat is applied simply because the act of drawing current from the thermocouple cools one side of the differential contact point.
I wasn't shooting for accuracy just enough EMF to feed the circuit.

[ion]

ION,

What you say is true but thermocouples will generate an EMF when heat is applied simply because the act of drawing current from the thermocouple cools one side of the differential contact point.
I wasn't shooting for accuracy just enough EMF to feed the circuit.

You said "Power it off of a few series connected thermocouples at room temperature"

In a series connected thermocouple circuit there is no current flow until there is a difference in temperature. You cannot draw any power from a TC circuit unless there is a temperature difference between the various junctions. You cannot have current flow in the closed loop at equilibrium temperature of the junctions, nor can you induce current to flow in a closed thermocouple circuit i.e. create a heat engine if there is no temperature difference. It cannot bootstrap itself into a temperature difference.

You can, however,  inject current into the loop and create a temperature difference between the differential  junctions.

If you can make it work otherwise, you will have a free energy self bootstrapping system, and I will throw out  all my reference material and life work in thermocouple thermometry.

http://www.allaboutcircuits.com/vol_1/chpt_9/5.html

read all the way to the bottom of the page.

[BEP]

My thought was the use of a thermopile by means of connecting thermocouples in series. Since I have never built one, don't throw out your reference material. I will concede to your greater knowledge of the device.

You learn something every day here.

I take it the series connected thermocouple connections must have common cold and hot areas separated. i.e. heat sink on one side and heat source on the other.

Then I must correct my suggestion from a bunch of thermocouples in series to using a thermopile, instead. Since they can produce voltage into the hundreds of millivolts, that should be enough to power the circuit.

>>Edit

It is good you made me think about my earlier post. Obviously, series connections of thermocouples, to boost voltage, would mean creating additional t-couple junctions with each series connection. Since t-couple wire IS the same metal as that side of the joint, this would create a reverse polarity canceling the effect of the manufactrured junction.

I don't know why I didn't just try it before blathering. I have a case of E & K types, with the correct wire, in the basement. DUH 

[poynt99]

A thermo-pile is a string of connected thermo-couples.

There won't be any potential difference between the terminals if there isn't a temperature gradient on either side of the junction(s).

In order to produce a voltage potential, you first have to create a temperature gradient within the device. One method is to place one side of it against a surface that is hotter or colder than the ambient temperature.

.99

[BEP]

Understood..... NOW 

Thanks to ION !

ION,

I'm embarrassed. I've been using such devices (not thermopiles) since the late seventies. I know all about the requirements for wire used and the importance of running it all the way to the instrument (so the circuit can compensate for the unintended junction at the terminals).

Lately, I have cringed at the sight of speaker wire being used to extend process monitoring sensors.

BUT --- I still made the mistake because I never looked into exactly how a thermopile works. I believe E types can put out around 50mV/deg C? Maybe it is time to build a thermopile 

I've been on a rampage to 5S (don't ask) my benches and junk piles. I was going to scrap the S wire for the platinum. I'll have to think about it now.

[ion]

I've been on a rampage to 5S (don't ask) my benches and junk piles. I was going to scrap the S wire for the platinum. I'll have to think about it now.

I made a lot of mistakes early in my career that taught me the hard way. Luckily I had a good mentor.

As for the platinum rhodium S wire, be sure you have the real McCoy not extension grade wire or it may not be worth that much.

Extension wire

Extension grade wires made of the same metals as a higher-grade thermocouple are used to connect it to a measuring instrument some distance away without introducing additional junctions between dissimilar materials which would generate unwanted voltages; the connections to the extension wires, being of like metals, do not generate a voltage. In the case of platinum thermocouples, extension wire is a copper alloy, since it would be prohibitively expensive to use platinum for extension wires. The extension wire is specified to have a very similar thermal coefficient of EMF to the thermocouple, but only over a narrow range of temperatures; this reduces the cost significantly.

[BEP]

Oh it is the real McCoy. It isn't extension wire.

One spool is 100%, the other is the alloy. If memory serves, the expensive one cost me almost $5k back in the late 80's.

We used them to feed a machine that made thermocouples for a government job.

Probably the best investment I've made, even though it was accidental (Uncle Sam canceled the contract mid-stream).

[stprue]

Hello all and happy holidays!!

I thought that I would post my ultra low power JT circuit.  It will run off a copper and magnesium ring but I have not fully tested it's capabilities.  The battery that I mostly use is a rechargeable that was at 1.01v, which will pull between .4-.8ua, at the time of the tests.

http://www.youtube.com/watch?v=WlXKvnj-TuQ