Manufacuring Thermionic Power Chips

[MadMorlock]

Hello,

I'm new to this board, but I'd like to get some verification on an idea I'm working on.

In 2001, Eneco, a company owned and operated by Professors at MIT, delivered their "Power Chip", a semiconductor structure that delivers to the solid state realm the powers of a vacuum tube thermal diode.

However, their packaging technique has limited their ability to market the device. In the following 5 years, they still have not managed to make progress in this field.

I believe I may have found a solution to their manufacturing trouble, but I need verification from outside sources.

My idea is fairly basic. I want to rotate the diode interconnect assembly 90 degrees to bring it parallel to the substrate. This should make the process much easier to develop. Right now the device requires a laminating technique where the junctions are perpendicular to the substrate, which is very difficult to manufacture on such as small scale. I've included a couple pictures of broken open TECs (which use a perpendicular diode array manufacturing technique) as well as some lego designs of the new arrangement for visualization purposes. Please note that the final device would operate perpendicular to the heat source.

Please check out the images at

http://morlock.newdor.com/

Thank you for your time.

 

[IRstuff]

Frankly, I don't see that you invented anything that works. Heat flow runs parallel to current flow for a TEC. Your layout makes no physical sense.

TTFN

 

[nbucska]

Is this supposed to be a thermoelectric power source?
Why not use In-Sb Peltier elements?


Plesae read FAQ240-1032
My WEB: <

http://geocities.com/nbucska/>

 

[IRstuff]

It's supposed to be a better TE cooler/heater than bulk material devices because the tunnelling vacuum junction blocks thermal conduction. Since the bulk material can also be made thinner and more conductive, the net heat flow can be greater and more efficient.

However, the devices operate on the same general principles as the standard TE device, i.e., the current flow itself moves the heat. That makes putting the hot and cold surfaces parallel to the current flow a non-starter

TTFN

 

[MadMorlock]

You wouldn't see a temperature gradient in a powered version of this device?

I understood that the evolution of heat took place at the junction interconnects, not really inside the junction itself.

As for the hot and cold surfaces, they're still there, just dealing with hot and cold edges now instead.

 

[MadMorlock]

Sorry, I meant heat evolution takes place within the interconnect, not inside the junction or the ceramic substrate.

The internal design of each diode prevents return flow, but the heat evolves into a hot carrier electron or hot carrier hole at the interconnect as it begins to traverse the semiconductor.

Which means that in this design heat flow would still be running in parallel to the current flow, just perpendicular to the substrate.

 

[IRstuff]

No heat is generated; it's moved by the current. The purpose of the ceramic substrate is to remove heat from the diodes and to spread the heat for the heat sink. Your substrates are shown in the pictures to be paralleling the current flow, thereby shorting out and reversing the junctions' heat transfer process.

TTFN

 

[MadMorlock]

It only runs down to ambient.

It's designed to cool to room temperature.

Until that point, the thermal flow is always downhill.