How to design the worst RF transmitter ever?

So, first off, why do I need it. My brother has taken up fencing. Not the sale of illegal goods, but the sword-fighting kind. Turns out that wired systems are sill widely used. There is wireless fencing equipment, but apparently they are hella expensive. SO, I thought I'd make a wireless fencing system.

What do I need the worst RF transmitter ever for? Some time ago I got an RC car for fixing. It simply didn't work. Well, the only way the signal wen't through is when the controller and cars antenna touched. Apparently the car was set to work on 27 MHz and the remote was transmitting on 40 MHz, the signal only got through when the antenna touched without much degradation from the filters on the cars end. From this I got the idea of using radio mismatched/bad radio transmitters and receivers to register a touch between the sword and the suit, both acting as an antenna. The sword would be the transmitter and the suit would be the receiver.

The only way to make wireless fencing work ( I think ) is to use radio signals, because direct current doesn't travel from one battery to another and a wireless system cannot have a common ground.

Anyway, for the worst transmitter/receiver I have following ideas:

Would it be better to use a square wave instead of a sine wave as the radio signal? I need the exact opposite of range, so if square waves transmit worse that sine waves, it could be a great idea.

Would it be better to use some sort of a simple binary data on the radio signal or for starters try and use an IC designed for RC usage?

There has to be some sort of data on the signal, so that different objects can be identified ( your and enemy sword would have a different signal, so stabbing yourself can be registered too... Not that it's a real problem in the fencing world)

Hello, I've worked on wireless fencing scoring systems before and it's definitely not as simple as you're thinking. At RF frequencies, the human body is a good conductor, and it becomes very difficult to distinguish contact with a lame with contact with a sweaty jacket, or a bellguard in contact with a sweaty glove. Even the systems used at the highest levels of competition do not work reliably (especially in sabre, where most fencers object to using the system).

It would take some work to get it working reliably, but if the signal is degraded enough, then shouldn't it only be readable when the epee touches the lame?

Ofcourse it's always possible to make a MEMS suit that has a MEMS chip every couple of cm which would create a 3D representation of the person and the epee on a PC which then could be used for 3D collision detection?`:D

I agree with mwieg, this isn't really about a "worst" transmitter, it is really about how to reliably detect two objects touching without wired connection. Certainly using a square wave is a bad idea, from a signal aspect a square wave is just a sine wave with lots of harmonics so all you would do is muddy the radio spectrum.

As a suggestion, and I've never tried this, how about tranmitting a low power 'clean' signal from each side and applying a hefty audio tone to the contacts. I'm thinking that the conduction of modulation, especially when it can easily be filtered to eliminate interference, is easier than conduction of the radio signal. A nearby receiver would listen for the tone on an established link rather than the presence of the link itself.

A different approach might be to use capacitive touch technology.

Brian.

If the jacket is moist with sweat (it will be) and the frequency is high enough, then yes the lame will pick up signal from the body, and vice versa. I've seen this firsthand.

The optimal frequency to use is a tricky question. You want it to be high enough so that the capacitance between the fencers' bodies is a low enough impedance to form a decent return path for the high frequency signals. But at the same time if the frequency is too high the human body starts to conduct very well, and suddenly it becomes very difficult to determine whether you're hitting a certain conductive surface or not. So basically it's a tradeoff between sensitivity and specificity. IMO the optimal frequency is likely in the range of 100kHz to 1MHz. If I were to try doing it again, I would likely try using a more wideband pulse, and actually calculate the full impedance matrix of the system over that bandwidth. So you're looking at a hefty amount of real time DSP and matrix inversion. It's a legitimately difficult problem, and I'm not even convinced a robust solution exists without extensive modifications to the equipment.

Can you not use the same system as a GDO (grid dip oscillator)? i.e. use a very weedy oscillator on each fencer which will stop if it it is loaded by the impedance of the other fencer? So you have two receivers to monitor the field strengths.
Frank