new invention?

[BrainyCapacitance]

so, I've realized that all the fuses out there, are all amperage-dependent fuses. there are some very very rare voltage-dependent "fuses," but otherwise, I don't know any fuse that is voltage-dependent. I'm not exactly a genius, so... I need a little help. this could very well be a stupid, dumb, and expensive idea, but I'm just a little curious.

let's say I have a Tungsten wire. it is 0.8mm thick, and 10mm long. The ambient temp is 45°F.
if the wire is conducting a voltage of say, 24V at 2.5A, it should generate a temp of at least 90°F on the wires.
here is my plan:

1)

   
I will put two tungsten wires close together, but not touching.

2)

   
next, I will drip about 0.025mL of Gallium, joining the two wires.

3)

   
now, when a voltage of like 24V flows through, the high resistance of the tungsten heats the wires up. since Gallium melts at just 87°F, when the wires heat, the Gallium will melt, thus breaking the connection



What do you guys think? is it kinda dumb? or is it worth a try?

[EDIT] So, I guess the conclusion is that this was a stupid idea

[wasedadoc]

Don't quit your day job. 

[wraper]

[brucehoult]

I believe the normal technique is to use a Zener diode to turn an over-voltage into over-current and blow the fuse. Alternatively, use the Zener to trigger an SCR for larger current ability.

[wraper]

let's say I have a Tungsten wire. it is 0.8mm thick, and 10mm long. The ambient temp is 45°F.
if the wire is conducting a voltage of say, 24V at 2.5A, it should generate a temp of at least 90°F on the wires.

Now how much power do you think it would be 24V*2.5A? Then think about tungsten resistance, and how much current you'd need to actually push through 10 mm *2 of 0.8mm diameter tungsten wires to get 24V drop (2 * 1.1 mOhm at room temperature). Also any fuse is voltage dependent, since power=heat is current multiplied by voltage. Voltage drop over the fuse generally is an unwanted side effect that is intentionally minimized but cannot be avoided. Against what your invention is supposed to protect?
BTW what you described as construction basically is a readily available thermal fuse.

[Muxr]

Teaching my 10 year old nephew electronics. I'm stealing this! Thank you!

[BrainyCapacitance]

ok! thanks!

[BrainyCapacitance]

ok

[BrainyCapacitance]

Thanks! I'm a bit new to electronics...

[BrainyCapacitance]

wait... so if I don't want a voltage higher than say, 50v in my 1A circuit, a 50v thermal fuse would break if the voltage reaches 55v 1A?

[brucehoult]

Teaching my 10 year old nephew electronics. I'm stealing this! Thank you!

When I was about 10 I was home from school, sick, one day. My father came in and said he was planning water distribution for a newly developed part of the farm. He needed litres per hour of water at the most far-flung trough so that it could keep up with cows circled around the trough (10 of them maybe, I don't remember) drinking at the same time, and periodically replacing each other. He gave me some books with flow rate in pipes of different materials (iron, copper, alkathene, PVC...) and diameter and how it varies with the "head" (pressure). And other tables of how much loss of head there is in each metre of pipe, how much loss of head in fittings of various kinds: bends, inline valves, change of diameter, simple joins, T-junctions ...

The question for me: what was the cheapest way to get the necessary flow rate at the end of the system, given the available head at the end of the current system? Hint: use a large diameter, maybe more expensive material, pipe at first, proabably along the metalled track, but maybe not all the way, and then cheaper material side-pipes to the troughs.

[Zero999]

The voltage drop across a fuse is tiny whilst it's carrying current. It behaves like a low value resistor. The total voltage applied to the whole circuit is irrelevant, as far as the fuse is concerned, until it blows.

In other words, there's no way, your fuse knows what the voltage across the entire circuit is, until it goes open circuit. There's no way to make a voltage dependant fuse, without another connection to the other side of the voltage source i.e. 0V.

[brucehoult]

In other words, there's no way, your fuse knows what the voltage across the entire circuit is, until it goes open circuit.

Until the fuse goes open circuit, or the stuff after the fuse goes short-circuit (which a Zener diode connection to the other rail can do for you, as it sees the voltage)

[wraper]

wait... so if I don't want a voltage higher than say, 50v in my 1A circuit, a 50v thermal fuse would break if the voltage reaches 55v 1A?

How do you think it could possibly protect your circuit from overvoltage? If it really was a voltage controlled fuse, it needs to "see" that voltage on itself. If it's in series with the circuit, it does not "know" actual voltage on the circuit, it only sees current and voltage drop on itself, the same as regular fuse. If it's in parallel to the circuit, it sees the same voltage as the circuit and becomes open over certain threshold, so protects nothing and circuit will still be powered.

[TimFox]

Remember that voltage is "across" two points, but current is "through" a path.

[BrainyCapacitance]

oh, I get what you mean. you're saying that the circuit will break before the fuse "blows"

[BrainyCapacitance]

ok

[BrainyCapacitance]

but, what about like, a power surge scenario?

[coppercone2]

ok well usually what happens is it does not drip but it vaporizes. a fuse does the same thing but it uses a different alloy. what you are proposing is a fuse made with different materials.

The only difference might be theoretically if this drip mechanism can be controlled, it would be extremely slow blow fuse. look at how fuses are made. you made a fuse that has a gigantic thermal mass.

it would probobly have a different response then a real fuse. ok there is a infinite number of responses possible. is this a useful response? nothing stands out too much.

now you can end up with a hybrid between a fuse and a thermal fuse with this.



despite everything, this can be fun to play with I think. make it and hook it up to a scope and see what it does when you overload it

[BrainyCapacitance]

ok!

[coppercone2]

s not a  stupid idea, its called experimenting with fusable links. if you are interested then you might as well study the response. chances are you will learn something playing with gallium, tungsten and electronics equipment


read littlefuse literature about fuses to get an idea how  to design the thing finely

but the main take away is that fuses are built to be fast, because thermal is slow compared to electrical, and electrical damage occurs instatanously almost. so the fuse is there to prevent grevious damage. but slow fuses do have their uses.

[coppercone2]

for a voltage dependant fuse, that is in every appliance. you make it by combining a MOV with a fuse. the MOV is voltage sensative and it shorts out the fuse if voltage is too high to blow the fuse. you combine semiconductor and fuse


no one did a mechanical one that I know of, there is no effects to work with really that make it possible (or maybe I should say practical). you could make a mechanism that deflection of capacitor severs a link if voltage is too high based on electromechanical principles. it would be like the over sized swiss watch of crazy

[BrainyCapacitance]

nice to know!

[BrainyCapacitance]

can I use a high-voltage FET with a fuse instead of an MOV?

[coppercone2]

its probably less robust but its a method. this is a field of engineering. try threads by t3slaco1l