inasity_rules's Journal: Gambling with my degree :)

I am doing something rather stupid, but nevertheless interesting. I am not sure what possessed me to do it, but I chose to do a rather strange honours project. I am implementing a wireless (CM)FSK mesh network. Sane so far? Just wait, it gets a bit crazy. I sat and thought, we don't have Oscilloscopes at my university that can cope with frequencies above 20MHz (heck some can't cope above 2MHz), so I was thought, "Hey, there is an ISM band at ~13.5MHz. Why not use that?" After all debugging a system is easier if you can look at the signal anywhere on the circuit...

Data rate? I'll be lucky to get 300bps. Which is fine. Its not a commercial project in any way, its just something I thought would make a pretty cool project to hack around with. Plus I implement lots of blinking lights and some embedded nodes which can be remote controled, which'll make everybody think "wow..."

So, I thought let me implement the whole thing using Phase Locked Loops. Cool so far - one chip and minimal external circuitry becomes a modulator and demodulator. Nice and easy(I hope), especially at such a low frequency and data rate. Now the fun begins; what the heck am I gonna use for an antenna? 1/4 wavelength is going to be 5.55m long! It would be fine, but to implement 4 nodes, I need ~23m of cable... At the distances I'm likely to run, I may as well just implement a wired network... Simple truth - my budget does not support 23m of wire, unless I cannibalize a transformer or something. Besides, its not elegant - its an ugly hack. And I don't have a spectrum analyser to measure the resonant frequency of any loop antennas. Its trial and error and CROs all the way. Here I am worried about ugly hacks when I got very limited time to complete this thing... Anyway, Shorter antennas will complicate the radiation pattern somewhat (I'd like an ideal sphere and to win the lottery) and probably degrade my signal to noise ratio.

I'm selling it to my supervisor as a control/sensor array idea, a use for which it is completely unsuited, since in the ISM band you can never guarantee data transfer.

This is all only interesting if you're a hardware geek/ham I guess. In any case now I've gotta start looking for a challenging, fun, and commercial masters project. Assuming these ideas of mine work and I actually graduate.