|
|
|
network pseudopod
Monday, December 13 2004
This evening I ran a temporary pseudopod of ethernet from my house's computer network out across the lawn, up into the trees, over the access road to the uphill neighbor's farm, and into another neighbor's swampy field, the one I'd been walking around in last night with my laptop. A little beyond where I'd first noted the mysterious new access point, I set up one of my nifty little wireless bridges, a DWL-800AP+ reflashed with DWL-810AP+ firmware (and also hacked to produce more transmit power).
I had no available electricity to power the bridge out on the remote end of that 300 foot long section of ethernet, so I decided to use this as an opportunity for my first ever experiment of using power-over-ethernet (PoE), in which unused wires in the ethernet cable are used to carry electricity to devices that it plugs into. In the existing ethernet specification, only half of the eight wires are used, leaving four for carrying electricity. There was nothing fancy about the way I did my PoE. In crimping my wire, I just slotted the Cat 5 jacket to give me an exit for the unused brown and blue wires, which I didn't include in the connector as I crimped it. I then soldered a power connector to these wires on remote end and left the other ends loose for trying a variety of power sources. This was necessary because I didn't really know how much voltage I would need.
Since the wires in my ethernet cable are thin (24 gauge), and since the voltages used to power the wireless bridge is low (five volts), while its power requirement is high (more than an amp), I had to expect a steep loss of voltage in my 300 feet of wire. Since I had four wires to work with, I used two each for ground and positive voltage, but still the voltage loss in the wire was extreme. Here's the ohm's law calculation, assuming a resistance of 0.015 ohms per foot.
600 feet of wire (300 there, 300 back) X 0.015 ohms/feet= 9 ohms
I X R=V (Ohm's Law)
1 Ampere X 9 Ohms = 9 Volts
This means that if I put ten volts into the circuit on one end, I'd have to expect to lose all but one of them to the wire's resistance, and that's only when the wireless bridge is drawing one amp. I didn't actually do any of these calculations when I was testing various power sources, I just looked to see if the light on the hub lit up when I hooked up different voltages. For obvious reasons, the usual five volts did nothing at all. But it came alive at 7.5 volts, indicating that the device must use much less than one amp in normal operation. For reliable wireless communications from my remote bridge, I eventually settled on a beefy 12 volt supply applied to the near end of its makeshift power-over-ethernet hookup.
In the end, I was able to reliably connect to that mysterious new access point, but I couldn't get to anything on its network, and nothing on its network wanted to assign me an IP address. I had no idea what IP range it was in; all I had to work with was the first three bytes of this new access point's MAC address, which is assigned based on its manufacturer. According to an online MAC address lookup service, the access point was made by a hitherto-unknown company called Abocom Systems, Inc.
Stepping back for a moment and examining my unusual interest in locating and exploring local wireless networks, I have to ask myself why this form of traditional "hacking" appeals to me so much, whereas I've never been especially interested in getting into or exploring, say, remote servers. It seems that in order to be of interest to me, there has to be a strong real-world component to my hackerly explorations. Sure, the internet exists and it's full of possibilities, but I've always been especially fascinated by how it attaches to the real world. This fascination is particularly strong in places such as my neighborhood, which are so rural that you wouldn't expect broadband to be here at all (indeed; it's been here for less than a year). To a large extent my effort to flush out local access points and somehow develop a method for using their combined bandwidth and aggregate wireless footprint is just my part of the grand effort to foster a completely networked world. Such a world is neither good nor bad; it just is what it is and I have applications I want to build for it once my little part of it comes into being.
For linking purposes this article's URL is:
http://asecular.com/blog.php?041213 feedback
previous | next |
