Re: invention help

Your invention to prevent problems from candles sounds very interesting.



In answer to your question, there are several ways you can have an 

electrical measurement for temperatures.



For example, thermistors change their resistance with temperature.



Thermocouples are made up of two different conductors, for example two different
 metals like iron and copper.  The two different conductors are connected at
 two junctions. You could detect the temperature change from the candle by 
putting one junction where you want to measure the temperature, and the other
 further away  where it won't get much of the heat from the candle. The 
voltage between the two junctions can be used to measure temperature.



However, a more interesting way may be to use a diode.  Diode are one way 
valves for electricity - current will flow if either a battery or a current 
source is hooked up one way, but not the other way.



If you hook up a diode in the "forward bias" direction (the direction in which 
electricity will flow), and force a current through it, then measure the 
voltage between the two sides of the diode - this voltage will be directly 
related to the temperature.  



For example, with a silicon diode, if you force about 1 mA (1 milliampere), 
you might measure a voltage of about 650 mV (650 millivolts) at room 
temperature. Room temperature is about 298 degrees Kelvin (the Kelvin scale 
stars with zero degrees Kelvin being "Absolute Zero" - the lowest possible 
temperature, a temperature at which all motion of atoms and so forth stops).



If the flame gets close, the diode might get to the boiling point of water, 
which is

about  373 degrees Kelvin.  At this temperature, still forcing the same 
current, the voltage will be   (650 mV)  x  (373 / 298) ,  or about  815 mV.  



If you tried this with an L.E.D. (Light Emitting Diode) instead of a silicon 
diode, you might get several benefits.  If you forced enough current through 
the diode to get it to light up, you might get a voltage that is somewhat 
higher - between 1 and 2 volts. If the forward voltage was 1.5 volt, then 
heating it up to the boiling point of water would generate (1.5) x (373/298), 
or about 1.9 volts - this larger voltage difference will be easier for you to 
measure.



Other benefits are that you could visually tell what is going on - the LED 
lights up when you forward bias it.  Also, if you have an LED that turns on at 
just under 1.5 Volts, then you could bias it with a D-cell battery for 
example, and the L.E.D. light would go out when the temperature got so high 
that  the forward voltage is more than the 1.5 volts that the D-cell battery 
can deliver.



If you want to bias the diode with a current source, you might want to look up 
in a basic electronics book how to make a current source - the electronics 
book may refer to it as a "current mirror".



If you use the diode approach,  try to get a diode in a metal can package 
rather than a plastic package - otherwise, since plastic burns, you might want 
to find some way to make sure either that the flame never gets near the diode, 
or that your invention turns off the fire before the flame hits the diode.



Another approach that a friend at work mentioned is to simply hook up a 
circuit with some solder wire.  When the flame gets too close to the solder, 
the solder will melt, and break the conduction path for the circuit.  You 
could use this instead of using a diode at all - as a kind of "digital" (Yes 
or No) electrical detector of how close the flame is.   



Good luck with your invention!