Henry L Imelmann - Method and apparatus for converting heat directly to electric

[Peterae]

I just found an interesting patent on improvements of thermocouples.

[ion]

I just found an interesting patent on improvements of thermocouples.

Oddly the patent is short on claims unlike others, unless I am missing pages.

Interesting that he claims to use Thomson effect, rather than Seebeck. Note that the Strachan/Aspen device also claims to use this effect in a novel arrangement.

Background reading: https://en.wikipedia.org/wiki/Thermoelectric_effect.

This has long been a pet study of mine and I hope to get something done in this area, as I have some novel approaches similar in spirit to the authors work, but different in execution.

I will post as time permits, and when drawings are close enough to convey the ideas.

Regards, ION

[BEP]

Very interesting.

I must say such a device shouldn't be called a 'thermocouple' because there is no 'couple' or junction. Perhaps that is what he means by 'reverse Seebeck'?

I would like to test the idea but I don't have constantan wire but do have some platinum wire made for the junctions.

I noticed he mentions heating one end of a very short length of constantan wire to a very high temperature with the result of the other end of that short wire remaining at room temperature() I wonder how that was possible.

[Peterae]

To me he has 2 brilliant ideas here, he takes a piece of constantan wire 1.6mm and puts a groove in it as a heat break, because the width of this groove is small and because the diameter of the wire is very small, the heat does not travel across to the cold junction but keeps the electrical resistance low because it's not long, see fig 1.

Also in the disc shape the heat transfer is small because the heated area of the junction is small and the resistance goes lower as the diameter of the disc increases, too very simple methods of increasing efficiency.

getting a flat sheet of constantan wont be easy i think.

[gyula]

See here constantan tape choices

http://www.wires.co.uk/cgi-bin/sh000001.pl?WD=constantan&PN=tape_cn%2ehtml#aTAPE_2dCN02031270_2d001M   

Choices of constantan wire:
http://www.wires.co.uk/cgi-bin/ss000001.pl?page=search&SS=constantan&PR=-1&TB=O&ACTION=Go%21   

Gyula

[ion]

See here constantan tape choices

http://www.wires.co.uk/cgi-bin/sh000001.pl?WD=constantan&PN=tape_cn%2ehtml#aTAPE_2dCN02031270_2d001M   

Choices of constantan wire:
http://www.wires.co.uk/cgi-bin/ss000001.pl?page=search&SS=constantan&PR=-1&TB=O&ACTION=Go%21   

Gyula

Constantan is one of many choices for thermomaterials. You could use iron and copper as thormoelements, which has fairly high output.

https://www.youtube.com/watch?v=JQUY_bs59a4

There is a lot of experimentation in this area visible on youtube

[conradelektro]

Constantan is one of many choices for thermomaterials. You could use iron and copper as thormoelements, which has fairly high output.

https://www.youtube.com/watch?v=JQUY_bs59a4

There is a lot of experimentation in this area visible on youtube

Very nice video, and instructive too.

Concerning the patent US2,857,446, it has three claims at the end (column 15, line 58 to column 16, line 37). But modern Peltier Elements are much better, in the attached photo I show a 12 Watt thermocouple (at 200 Kelvin temperature difference between cold and hot side, about 4 Volt and 3 Ampere, 50 mm x 50 mm, about 50.-- Euros).

I attached a Joule Thief circuit to the Peltier Element which allows to show a bright LED with a very low Voltage and Amperage, hence a very low temperature difference. In the photo my range is almost cool and the little tin plate on top has room temperature. To light the LED dimly one can place the cold side on the table and a hand on the hot side (temperature difference about 15 degrees Centigrade). If placed on the hot range and with some ice in the top tin plate I could light a CFL with a different Joule Thief circuit (3 V to 220 V conversion with the Joule Thief circuit, a few Watts).

The efficiency of a modern Peltier Elements is about 4% if a paste is used to close the gap between heat source and cold sink.

I once saw a Russian wood stove with Peltier Elements producing about 200 Watt, but the cold side of the Peltier Elements (cooling fins) was supposed to be exposed to arctic temperatures.

Note: you can have a 10 Watt solar panel for less than 50.-- Euros, making the Peltier Elements a bit obsolete.

Greetings, Conrad

[ion]

Thermoelectric conversion won't become really efficient until the thermoelements can be made into highly electrically conductive or even supeconductive ceramics or other thermal insulating but highly electrically conductive materials.

One creative approach was the Strachen/Aspden device which got a writeup in Wireless World many years back:

http://www.overunityresearch.com/index.php?topic=48.msg355#msg355

http://www.overunityresearch.com/index.php?topic=156.msg1285#msg1285

I agree with Conrad, there are way more efficient means of conversion. Hopefully the new materials are around the corner.

[Peterae]

I'm wondering if it would be possible to make these discs of constantan by electroplating, last year i uncovered a method of plating most metals using easy to get hold of chemicals which i have, the great thing with this method was that it could plate alloys by using 2 anodes of different metals, once the disc is made then plate copper electrodes on the outside and inside circumferences.

Thanks for the link gyula.

Indeed ION
Iron and Copper sounds good as well and easier to get hold of as well.

[Smudge]

While on the subject of thermal gradients creating free energy, it is also possible that magnetic gradients can do something similar.  Perhaps this requires a new bench but here is the gist of it.  Spin polarized conduction electrons act like tiny aligned bar magnets and as such they can be dragged along by a magnetic gradient.  If this is done in an isolated ferromagnetic conductor there will be a concentration of electron charge at one end and a deficit at the other.  In some respects this is related to the magneto-coulomb effect discovered in magnetic tunnel junctions and to the magneto-Seebeck effect also discovered in such small devices.  Clearly it could be possible to connect non-magnetic leads to the ferromagnet where the conduction electrons lose their spin polarization and can then be transported back the other way through the magnetic field gradient back to the other end of the magnet.  The attached paper suggests just this possibility.  You can buy sample sheets of mumetal in different thicknesses, anyone here interested in trying this out?

Smudge

[Smudge]

I see I did start a thread on my magnetic battery idea some time ago and didn't get a single reply.  Maybe because the paper is suggesting a deposited array of miniature elements which is beyond most experimenters.  But the principle could be examined using a small number of flux concentrators in series with copper wires connected between them, and the whole lot placed in a magnetic field provided by a pair of PM's.

Smudge 

[Smudge]

For what it's worth here is a suggestion for an experiment with a fabricated array of triangular flux concentrators cut from sheet, interconnected with copper wires.  The whole lot placed between two slab magnets (not shown).  This could produce measurable DC voltage or current.

Smudge

[Grumage]

Dear Smudge.

Your last post struck a chord.

http://www.rexresearch.com/meyers/meyers.htm

Any similarities?

Cheers Graham.

[ion]

Dear Smudge.

Your last post struck a chord.

http://www.rexresearch.com/meyers/meyers.htm

Any similarities?

Cheers Graham.

I was highly interested in the Myers device some time back and studied the patent intently.

I did a few crude experiments, but by no means adequate to prove or disprove the device.

I think it is worthy of it's own thread, and there may even be one here?

The story in itself is intriguing.

Smudge:

Thanks for your latest offering. I have some thin sheet mu metal salvaged from CRT monitors, and a spool of some ribbon stuff of high quality used in one of your earlier offerings. Will build as time permits. Certainly should have it's own thread.

Regards, ION

[Chet K]

I did start a Mirror Topic over here
http://overunity.com/16974/-smudges-musings-a-magnetic-battery-consideration-with-a-simple-experiment/msg495515/#new

respectfully
Chet K

[Smudge]

Dear Smudge.

Your last post struck a chord.

http://www.rexresearch.com/meyers/meyers.htm

Any similarities?

Cheers Graham.

Interesting!!  I think there is a possible explanation so here is my attempt.

Smudge

[conradelektro]

While on the subject of thermal gradients creating free energy, it is also possible that magnetic gradients can do something similar.  Perhaps this requires a new bench but here is the gist of it.  Spin polarized conduction electrons act like tiny aligned bar magnets and as such they can be dragged along by a magnetic gradient.  If this is done in an isolated ferromagnetic conductor there will be a concentration of electron charge at one end and a deficit at the other.  In some respects this is related to the magneto-coulomb effect discovered in magnetic tunnel junctions and to the magneto-Seebeck effect also discovered in such small devices.  Clearly it could be possible to connect non-magnetic leads to the ferromagnet where the conduction electrons lose their spin polarization and can then be transported back the other way through the magnetic field gradient back to the other end of the magnet.  The attached paper suggests just this possibility.  You can buy sample sheets of mumetal in different thicknesses, anyone here interested in trying this out?

Smudge

Hi Smudge, are you aware of this fact stated in the paper about the magnetic battery:

However a word of warning, there is a characteristic of the electrochemical
potential across junctions between dissimilar metals known as the magnetoSeebeck
effect. The Seebeck coefficient is affected by a magnetic field, and it is
possible (likely?) that the Cu-Fe junction at the low B field end will have a different
voltage to the Fe-Cu junction at the high B field point which would nullify the
magnetic battery effect.

So, the Fe-Cu junction (there is always copper in the drawings of the paper and in your triangle-drawing)  could nullify the effect.

What is your idea to counteract the Seebeck-effect (Fe-Cu junction)?

Greetings, Conrad

[Smudge]

I did start a Mirror Topic over here
http://overunity.com/16974/-smudges-musings-a-magnetic-battery-consideration-with-a-simple-experiment/msg495515/#new

respectfully
Chet K

Chet,

Could be worth putting my pdf paper "Magnetic Battery Considerations" over there to give the background to the idea.

Smudge

[Smudge]

Hi Smudge, are you aware of this fact stated in the paper about the magnetic battery:

However a word of warning, there is a characteristic of the electrochemical
potential across junctions between dissimilar metals known as the magnetoSeebeck
effect. The Seebeck coefficient is affected by a magnetic field, and it is
possible (likely?) that the Cu-Fe junction at the low B field end will have a different
voltage to the Fe-Cu junction at the high B field point which would nullify the
magnetic battery effect.

So, the Fe-Cu junction (there is always copper in the drawings of the paper and in your triangle-drawing)  could nullify the effect.

What is your idea to counteract the Seebeck-effect (Fe-Cu junction)?

Greetings, Conrad

Of course I am aware, I wrote the paper!  Whenever you have two junctions in series, as you must to connect at each end of a material, the Seebeck Effect is only present when you have a temperature difference between each end.  Here we have no temperature difference.    The two galvanic voltages cancel out.  However the magneto-Seebeck (MS) effect is different.  Since writing the paper in 2006 I have learned a lot more about the MS effect and now I am of the opinion that, far from cancelling the effect I describe, the the MS effect is the effect I describe.

Smudge

[Smudge]

Interesting!!  I think there is a possible explanation so here is my attempt.

Smudge

Or more likely the iron-zinc connections along with atmospheric water vapour create unstable Fe-Zn batteries.

Smudge

[Peterae]

Hi Smudge
Let me get this thermocouple out the way and i will look at your flux device.

I tried plating copper onto the mild steel edge but it did not work well.

I am now waiting for a quote to get a mild steel and copper part laser cut, i will then braze the 2 together.

With this setup i can clip a current clamp meter over the copper lip.

[Peterae]

Looks like i need to learn how to braze
I had 2 of each made out of 1mm copper & mild steel sheet, 1 set with copper in centre and outer and mild steel in middle and then i had then use the offcuts to reverse this.
There's only 0.2mm clearance between them so not sure how easy this is going to be to braze, but will give it a go.

I am wondering how well a fet would switch mili volts but high current, if this is possible then i could try a 1 turn primary step up transformer to gain voltage at the expense of the current.

Anyway who knows it probable wont work at all and that's if i even manage to successfully braze them together.

[Orthofield]

Hi Peterae, all,

I was happy to find this thread. I've been interested in the many methods for increasing efficiency of thermoelectric generators for quite a while.

There are a whole host of patents like Imelmann that describe 'thermal gaps' that have high electrical conductivity and low heat flow. It seems to have been pretty well known in the art, before the advent of ceramic thermoelectric materials. Imelmann is exceptional in using the Thomson effect.

The attached patent by Winkler shows the heat restriction idea used to improve the efficiency of a standard Seebeck generator. 
Imelmann is right when he says there are many inventors with claims of very high efficiency thermoelectric generators-- at least Carnot efficiency, maybe beyond that. I think this is quite possible.

The Aspden-Strachan device was mentioned. The lit suggests that any way you reduce the thermal conductivity while maintaining the electrical conductivity will lead to very high efficiency (40%+)   The A-S device uses an AC thermocouple of a sort, so that power can be transferred through the dielectric layers while heat is more or less blocked. I've seen other designs where the thermocouple elements are essentially two layers of a capacitor that is cyclically dumped into the load.  Spark gaps have also been used, so that thermal conduction, aside from the small radiative component, is completely cut off. The same effect has been achieved by making the TEG elements without any 'thermal gaps' but a part of an AC circuit with capacitive reactance.
I can fill in the data on these if desired-- there are some truly extraordinary technologies in this neck of the woods!

After surveying the literature on this stuff, I came to the conclusion that broadly considered, thermoelectric technology was truly the wave of the future...
Fred

[ion]

Dear Fred :

You can post info on the Aspden-Strachan device here:

https://www.overunityresearch.com/index.php?topic=48.0

Thanks for the Winckler pdf, will look it over.

Regards

[Orthofield]

OK, will do!
F