For those going off the grid....

[WQ9E]

Old tech Russian radio/lamp/stove for the power challenged:

[Ed/KB1HYS]

Or for real QRO operations try some of these...

http://www.dself.dsl.pipex.com/MUSEUM/POWER/thermoelectric/thermoelectric.htm

Lay in the supply of coal now while it's cheap...

Thermopiles are a neat concept, if you have plenty of heat...

[k4kyv]

The voltage produced is proportional to the temperature difference between the two junctions. The proportionality constant a is called the Seebeck coefficient.

Something I have never quite figured out about thermocouples and the fact that the voltage is generated due to a difference in temperature between each of the dissimilar metals, is in regards to the thermocouple rf ammeter.  The thermo power generators have heat radiating fins to maintain a temperature gradient between the flame and the  heat sink, which serves  to keep the two metals at different temperatures.

How does a thermocouple rf ammeter work?  Since this is a precise measuring instrument, wouldn't the reading be highly dependent on the ambient background temperature at the thermocouple, versus the heating effect of the rf current?  I would seem that the heating effect of the rf current would eventually reach equilibrium with its surroundings, and everything in the thermocouple would tend towards the same temperature, so that the meter reading would slowly drift towards zero even with a steady current flowing through the meter.  The laws of thermodynamics say that a heat source and its surroundings will eventually all reach the same temperature, unless there is a sink to constantly drain the thermal energy away.

Even with an effective heat sink surrounding the heating element, why doesn't ambient room temperature affect the reading of the meter?  These meters seem to be totally unaffected by room temperature.

[K4TLJ]

Guessing... RF flows on the outside of conductors. If the inside were made of a different metal than the outside then heating caused by the RF flowing on the outer conductor would cause a difference of potential and a flow of current between the two metals which could be measured by connecting a ammeter to the two conductors...

[W1EUJ]

Arent the thermocouples made of thin wires in evacuated ampules, surrounded by insulation? Also most of the RF current is shunted, right? Heated area is small, and loss is only through conduction and radiation. It seems like outside influences on that single heated point are weak.

[Bill, KD0HG]

EUJ is on the right track.

The thermocouple is very fine wire and runs at a high temperature, so the ambient temperature of the meter doesn't really affect the delta-T much. Plus there is a feature called cold junction simulation (?) that compensates for changes in ambient.

[K3ZS]

The Peltier coolers/heaters that are commonly found in auto 12V electric coolers are basically large scale thermopiles.     I wonder if these elements would generate an equal amount of 12V power if they were heated?

[Bill, KD0HG]

Thought about this on the morning drive.

Don, IIRC, they attach a thermistor to the cold side of the thermocouple as a temperature compensation element. When the cold side of the thermocouple warms up some, the thermistor's resistance drops, which allows more current to flow into the attached DC meter, which compensates for the reduction of delta-T.

[W1EUJ]

Only one way to figure this out - send me one for analysis.

Good in the callbook.

[k4kyv]

Thought about this on the morning drive.

Don, IIRC, they attach a thermistor to the cold side of the thermocouple as a temperature compensation element. When the cold side of the thermocouple warms up some, the thermistor's resistance drops, which allows more current to flow into the attached DC meter, which compensates for the reduction of delta-T.

How long has that technology existed?  I don't recall ever hearing of thermistors in anything made before WW2.  I have some thermocouple rf ammeters that were probably made in the late 20's.  I used one of those for a couple of years until it became intermittent.

Last time I blew an rf ammeter, the whole thing opened up.  I didn't remove the meter, but had to short the terminals together so the circuit would work without it.  It was in the Gates 160m transmitter, which is almost as finicky about taking a load as a ricebox.  I use an outboard L-network to match the 50-70 ohm output  to the transmission  line.  One day I forgot to re-connect the feedline after a t-storm, and the circulating current between transmitter and unloaded L-network was enough to instantly blow the meter.  I decided not to replace the meter because that is something that will inevitably happen again.

[Bill, KD0HG]

The technology has apparently been around a long. According to Wikipedia, the thermistor was patented in 1930. I do know that Globar resistors work similarly, resistance inverse to temperature, and they've been around since the 1930s, maybe the 1920s (?)

Before thermocouple ammeters for AC, there was such a thing as a hot wire ammeter where a pointer was attached to a piece of wire. As the wire heated, it got longer due to expansion, which moved the pointer. But that device was probably 1800s, Westinghouse era. Antiques by the 1930s.

[Opcom]

I was told by an expert that the lamp there in the Russian picture would not have put out enough juice to run that big radio. But then the guy was not an old radio expert and the Russians were well known for building a big wooden cabinet around a very small chassis.

[WQ9E]

They do appear to work and there is a bit more information about this and some bigger thermo-electric generators:

  For example: http://blog.modernmechanix.com/2007/10/05/kerosene-lamp-powers-radio/

and:  http://www.dself.dsl.pipex.com/MUSEUM/POWER/thermoelectric/thermoelectric.htm

[Ed/KB1HYS]

The Thermopile was one of the earler electric current producing systems.  They were used quite a bit in electroplating.  Their primary downfall is that they can only produce a low voltage value, but with a high current, although large size and inefficiency don't help either.  Their advantages are their incredible simplicity, and the fact that they are not fussy about fuel types at all (coal, gas, oil, you name it!).

Common Junctions were made of German Silver, Zinc, antimony, and copper alloys,  and Iron.

Theoretically they will generate power as long as heat is applied, and the cold junction cooled, though some types had corrosion issues that increased internal resistance which reduced performance.

[Mike/W8BAC]

The Collins 20V-3 has an RF ammeter with an external thermocouple. Mine was vaporized before I got it. To fix it I had to understand it. Stu AB2EZ gave me a hand when he sent me this patent application from 1936. Thanks again Stu.

[WB3JOK]

The Thermopile was one of the earler electric current producing systems.  They were used quite a bit in electroplating. 

Also were used for powering some satellites, I think. Although not very efficient, a Pu-239 decay heat source (with a 24,000 year half-life) would surely outlast the satellite!

[Bill, KD0HG]

Voyager 1, now 30 years running on three Plutonium-powered thermopiles.

Started at 470 watts at 30 VDC, now at 315 watts, NASA has shut down much of the science
due to needing to conserve power. Still, they expect the power packs and communications to hold out until
2025. 14.6 light-hours from Earth right now.

From Wikipedia:

"Electrical power is supplied by three radioisotope thermoelectric generators (RTGs). They are powered by plutonium-238 (distinct from the Pu-239 isotope used in nuclear weapons) and provided approximately 470 W at 30 volts DC when the spacecraft was launched. Plutonium-238 decays with a half-life of 87.74 years,[3] so RTGs using Pu-238 will lose a factor of 1 − 0.51 / 87.74 = 0.78% of their power output per year. In 2006, 29 years after launch, such an RTG would produce only 470 W × 2-(29/87.74) ~= 373 W — or about 79.5% — of its initial power. However, the bi-metallic thermocouples that convert heat into electricity also degrade, so the actual power will be even lower. As of August 11 2006, the power generated by Voyager 1 and Voyager 2 had dropped to 290 W and 291 W respectively, about 60% of the power at launch. This is better than the pre-launch predictions based on a conservative thermocouple degradation model. As the electrical power decreases, spacecraft loads must be turned off, eliminating some spacecraft capabilities."

http://en.wikipedia.org/wiki/Voyager_program

[WB3JOK]

I stand corrected - Pu-238, not 239.
ONLY an 87-year half-life 
Thanks for the update!