Magnolia tree

[Virchanza]

You should never run CAT5 cable parallel to anything carrying 60hz power. It will cause an inductive interference into the cable. CAT5 cable should always cross power lines at right angles to minimize the possibility of inductive interference.

I know there's a lot of guidelines like this, but I tend to be the skeptic and want to try them out first. Right now, I have my wireless aerial mounted on top of an electricity pile-on, it's less than two feet or so away from thick cables running 230 V at 50 Hz, and I'd say there's a pretty decent current flowing through them as well because the pile-on is supplying electricity to other parts of the city. Anyway, before I put the aerial up there, I had considered that maybe the power cables would fudge up the signal, but I gave it a go anyway. Well anyway, it works. Fair enough it might work better if the power cables weren't there, but it works.

Also, no one but the power company or authorized leaser's of the power company are string anything on the poles.

Where I am at the moment, some of the electricity pile-on's are on the land of private citizens, the electricity company actually pays the private citizens a monthly rent charge. My next door neighbour has a pile-one behind his house on private land, so I climbed to the top of it and stuck my aerial on it.

Plus, the maximum distance you can run CAT5 is 100M or 328F. Chances are they're further away than that, plus the cost of getting exterior CAT5 cable would make it not so cost effective.

This is another thing I like to try out. For instance, you're not supposed to have a USB extension lead greater than 5 metres (unless it has a repeater)... but right now I'm running my Alfa off a 6-metre USB lead (without repeater) and I haven't had a single hitch. (That's an extra 20%)

[streaker69]

Where I am at the moment, some of the electricity pile-on's are on the land of private citizens, the electricity company actually pays the private citizens a monthly rent charge. My next door neighbour has a pile-one behind his house on private land, so I climbed to the top of it and stuck my aerial on it.

That may be in your particular situation. But you advised someone that may not realize that they can't just go and string something along a pole that they don't own. Granted, it would seem that the OP is smart enough not to do that, but others may not be. I believe, technically, people aren't even allowed to staple/nail signs to power poles as those things are left in the poles and can cause severe injury to workers if they need to climb the pole. I recently saw a pole that easily has 500 nails in it the entire way around the pole. Just because it's in your yard, really the right of way, does not make it yours to do with as you please.

Specifications are written for a reason, you may not be noticing issues, but in practice, especially in a professional environment, those spec's should be followed.

I cannot tell you the number of times I've seen someone do something outside of spec that worked for a time being and then degraded drastically over time. About the worst case I saw was TokenRing cable spliced together with flat satin using masking tape. Yep, it worked for a time, and then it started to cause so many errors on the line, it caused the server to abend.

[Virchanza]

I cannot tell you the number of times I've seen someone do something outside of spec that worked for a time being and then degraded drastically over time. About the worst case I saw was TokenRing cable spliced together with flat satin using masking tape. Yep, it worked for a time, and then it started to cause so many errors on the line, it caused the server to abend.

***goes to dictionary.com to look up "abend"***

PIC microcontrollers are notorious for having extra safe limits in their spec sheets. They say not to draw more than 25 mA from any particular pin... but you can get away with drawing many times that.

In the first draft of the Connect4 project I did for college, I had an error on my schematic that I had to correct manually by going around my electronics board with a soldering iron and some wire. I needed somewhere to solder some wire, so I ended up using holes that were originally destined for current-limiting resistors that led to LED's.

I figured if I let the higher current flow for a smaller amount of time, the LED's wouldn't heat up too much, no damage done.. Anyway, my board worked perfectly. I even left it on overnight for three nights in a row and it didn't fail. However after about two weeks, the PIC chip finally burns out.

Boy did I learn my lesson on abiding my spec sheets. But in the case of having cables too long, or too close to sources of interference, well the worse case scenario is that you have to cop on and get a repeater (or find a different place to run the wire). It's a different story when you've to redesign a project picking different components with different tolerances... *shudder*

[streaker69]

Boy did I learn my lesson on abiding my spec sheets. But in the case of having cables too long, or too close to sources of interference, well the worse case scenario is that you have to cop on and get a repeater (or find a different place to run the wire). It's a different story when you've to redesign a project picking different components with different tolerances... *shudder*

Did you find 'abend'? If you did, you probably found how far back it goes, and what sort of system I was working with at the time.

One thing I thought of after posting that lastnight was, how are you testing to see if your connection is actually good? You're passing traffic through it, great, but are you actually polling SNMP registers to see if there's errors and retries on the line? Are you actually getting an efficient connection or is it just connected.