Ideas for low voltage sound activated switch

I'm looking for ideas how to solve a problem of designing a very small sound-activated switch.

The basic requirements are:

0) Sound triggers a mechanical action. A piece of nylon stops the mechanical action, so if the nylon is melted, the action takes place. Hence sound must melt nylon.

1) Circuit works from two button cells. Size is a real limit. The ultimate (and distant) aim is to build a circuit that is small enough that a person could swallow it.

2) Sound at a specific frequency triggers the switch.

3) When the switch is triggered, the battery voltage is applied to a 0.2 Ohm resistor, which heats up and melts the nylon.

Are there any switching devices that will have an on-resistance of well under 0.1 Ohms whilst still being very small? A large power FET is out of the question due to size.

A high on-resistance will mean the switching device will dissipate more power than the load resistor, which is not a good idea. At that point I might as well use the switch as the heater! That might be possible in fact, although a concern is the semiconductor will be destroyed before the nylon melts.

The load, which acts as the heater is (or at least should be) matched to the internal resistance of the cells. So the on-load voltage will drop to half the off-load voltage. So whatever switch mechanism is used, it will have to keep triggered even if the battery voltage drops to about

1.4V (0.7 V/cell), although triggering will start with full battery voltage.

Using a DC-DC step up converter and storing some charge in a cap is one idea I am considering. That way, I will have a source of reasonable voltage even if the battery voltage drops to 0.7 V/cell on load. But could anything be built that is smaller than a couple of button cells? If not, more batteries would be easier. But the potential client does not want the power source being any larger than two small button cells.

I've been asked to look at designing this, but I think the size constraints are quite serious given the circuit will need to supply about 7W of power for a short time period.

Any ideas?

OK, so you're trying to convert sound to mechanical action, and the electrical/thermal business is an assumption? How about avoiding the middleman and going for direct conversion?

Do you have any control over the input sound frequency? I'm thinking make the control element resonant at the input frequency so that it trips a mechanical latch allowing the mechanical action to occur. Think of a mousetrap; no cheese, instead a string is attached to the trigger. When it vibrates in response to the input sound it gets shorter and WHACK!

FTM can you tell us what kind of mechanical action is to be triggered; i.e. spring-driven motor? Also how much power will it dissipate mechanically, and what the actual mechanism is? It's easier to figure out how to upset something if we know how it's set, if you see what I mean.

Yeah. Dissipating 7W even briefly will make this thing HOT. You really want that to happen in somebody's gut?

I can't imagine that the desired mechanical action is gonna dissipate any 7W; the whole idea of cybernetics is to control a large force with a smaller one, not the other way around.

Mark L. Fergerson

A medical application.

I think that will be too big. And it is still going to be a very low voltage. Not even sufficient to make an electet microphone work.

Such as? I can't find anything small that offers such a low on resistnace.

That is one aspect I need to discuss. I am not sure exactly how the mechanism is triggered and so how it might be done in another way.

I was not thinking of trying to power the load from the cap, but perhaps use charge in the cap to keep the electronics (perhaps gate drive, or base current) flowing whilst the battery dropped to 50% of its value.

I'm told they are using 0.2 Ohms. I believe this matches the internal resistance of the cells used, but I have not verified this myself. I am taking their word for it, but will check more carefully if I take on the job.

No. I am open to ideas.

I understand they are using two hearing aid cells and don't wish to use anything bigger. But this may just be impractical to use.

I'm not sure of exact type of battery. Of course, they are getting the circuit to heat up the resistor (which I belive is 0.2 Ohms, but may be wrong) wehn they connect two cells to the load. That is not the same as doing it electronically of course, in response to sound.

I can't belive 7W is needed, but if the cell's impedance is matched to that of the load resistor, and the load resistor is 0.2 Ohms (both may be incorrect, but I belive are true), then it would imply 7W.

Potentially yes. Sound was just one idea. Radio starts getting big. Light it not practical as the cirucit will be hidden.

Yes.

Yes, that does not sound a bad idea.

Yes. It does not even have to be sound, but that seemed one of the more logical to me.

Get you, although mechanically that might be quite difficult to do.

I'm not sure exactly what it is, but there is some sort of spring loaded device which will fire once. It should take very little time to occur. A bit like a mousetrap that is held down with nylon, then the nylon melts and the mousetrap goes bang.

Temperature rise will be proportional to energy, and the energy will be of short duration.

Perhaps I am wrong about the 7W. I am told they are using a 0.2 Ohm resistor as the heating element and two 1.4 V cells, giving a total of

2.8V. Assuming the resistor is conjugately matched to the battery, that means the batteries internal resistance will be 0.2 Ohms (0.1 Ohm/cell). So the current will be 2.8/0.4 = 7A, and the power in the load 7^2*0.2=9.8W. OK, so now I make it 9.8W. Either way, if the load really is 0.2 Ohms, and matched for maximum power transfer, it is a lot of power.

If I am lucky, the 0.2 Ohm resistor is far too low which would mean I can get better power dissipated in the resistor whilst taking less current from the battery. That is quite possible. I don't know how much optimisation (if any) has been done to date.

As you might guess, someone else has been playing with this and got it working with a battery heating a resistor. Now they want it to be remote controlled, so they have asked me if I was interested in doing it.