We had a site with a 1000 watt AM Hotstick about 100 feet from another 
tower.  We ran the cable (non-shielded outdoor rated) and crimped the 
end at the top of the tower.  The installer all of a sudden, felt a 
burning sensation on his thumb with that wonderful smell of burning 
skin.  Guess what, he was holding the end of the cat5, not plugged into 
anything, just ran up the tower!  Sure enough, nice little burn marks 
right where the copper pins on the cat5 was! 

Put in a few ferrite beads, never looked back.  fun stuff to say the 
least.  ON a side note, we are at 400 foot on a 1400 foot FM tower 
transmitting at 100,000 watts at the top.  We ran Power up, but have a 
440 foot non-poe cat5 down, and don't have any issues! lol

* Dennis Burgess, CCNA, A+, Mikrotik Certified Trainer
WISPA Board Member - wispa.org <http://www.wispa.org/>
Link Technologies, Inc -- Mikrotik & WISP Support Services*
*Office*: 314-735-0270 *Website*: http://www.linktechs.net 

*/ Link Technologies, Inc is offering LIVE Mikrotik On-Line Training 

Marlon K. Schafer wrote:
> Hi All,
> I think we finally have this all figured out.  Now I just have to figure out 
> how to fix it.
> We've been up there for over 6 years now.  It's certainly been a problematic 
> site though.  Constant channel changes (we have 3 competitors a mile away 
> and pick up hundreds of ap's from in town) are the norm.
> This fall (a month or two ago) one of the tenants left the building.  This 
> cleared out most of the hardware that was in there.  A little bit before 
> that I replaced an Inscape Data and a smartBridges combo with a single MT 
> access point, using one of the cables that had been working for one of the 
> other two.
> About a week ago things started to really act up.  Multiple devices were 
> having trouble.  I was able to catch it in the act finally.  This time the 
> problem wasn't a wireless issue, the devices were constantly disconnecting 
> and reconnecting at the switch level.
> I pulled the Cisco switch out and dropped in a Netgear unit.  That didn't 
> fix it.  Next I put in a Digital Loggers rack mount reboot device.  That 
> wouldn't connect right either.
> I finally had to pull all of the hardware off of the shelf and set most of 
> it on the floor (or just let it hang there) to get it working at all well. 
> Still not perfect but better.
> I had by now hiked up there through sometimes knee deep snow 3 or 4 times. 
> Next I took a motorbike with studded snow tires up and got permission to 
> turn down the power to the radio station.  That didn't fix the problem 
> either.
> Next I borrowed a snowmobile and hauled some help and my spectrum analyzer 
> up.  I was unable to see any signals that didn't belong.
> Next day, another hike up the hill.
> OK, maybe a cat 5 cable went bad and I'm getting backfeed through the 
> switch.  DC current or something.  So I started testing the cables that run 
> to the most problematic units.  Well now, look at that.  Bad cable.  In fact 
> there are three of them.  Hmmm, kinda strange though.  All three have the 
> exact same fault!  Oh well, better change them out anyway.  I ran three new 
> cable runs and just for kicks I tested one of them.  What the heck?????  The 
> new cable has the EXACT same fault as the old one!  Even though it didn't 
> follow the exact same path as the old cables.
> Man, this is sure looking like a problem caused by the radio station.  I was 
> using indoor cat5 and didn't run lightning protection or ground anything. 
> Yeah I know, but remember that this has been there for a very long time like 
> this.  And as a guy with an electrical background I know that there are 
> actually two ways to deal with stray electrical.  Grounding is one. 
> Insulating is another!!!!  Anyway, I know it wasn't built to specs.
> I added some grounding and that didn't help at all.
> Yesterday I finally had one of the local wireless companies (Day Wireless) 
> that mainly does VHF radios, backhaul etc.  They also checked things with 
> the spectrum analyzer but couldn't find anything amiss.  I was able to 
> duplicate the wiring fault for them (with my Ideal tester).  But suddenly 
> everything cleared right up!  Stuff was looking good, no cable fault etc. 
> Pings were looking good, devices were finally negotiating the connections 
> right etc.
> I called the radio station to ask if I could try turning the power down 
> again to see if we see any change on the spectrum analyzer.  They said they 
> thought that I'd already done that because the showed the power was way 
> down.  Turns out someone in the building had bumped a breaker and shut down 
> part of the transmitter!  Well, we got all of that figured out and guess 
> what.  All of the problems came right back!  I then turned the power back 
> down and they cleared up.
> Tip for you guys, dropping an 18,000 watt system down by even 60% of it's 
> normal output isn't always enough.  We had to drop down to 10 to 20% to get 
> the problems to clear up.  The guys from Day Wireless had some small ferrite 
> beads with them so we stuck them onto the cables.  Put the beads on and the 
> radios would negotiate at 100full.  Take them off and they'd drop right back 
> to 100 half.  Duplicatable all day long.
> Soooo, current theory is that the radio station is screwing up my cat5 
> connections.  The fact that the building has less hardware in it and we have 
> more snow up there than normal has probably caused some different eddy 
> currents or multipath.  Or some other such strangeness.
> I have some shielded cable and connectors on the way.  I have permission to 
> move my gear from one side of the building to the other side.  I've got more 
> high end ferrite beads on the way (one that is made for cat 5 and is big 
> enough for three wraps to go through it).  Not the little $.50 ones, these 
> are nearly $6.00 each.
> I also have a Fluke DTX on the way. 
> http://www.flukenetworks.com/fnet/en-us/products/DTX+CableAnalyzer+Series/ 
> They rent for $650 per month from http://www.trs-rentelco.com/  List is over 
> $7000.  If I like it I'll probably get one and add it to my two spectrum 
> analyzers and my bucket truck.  grin
> Never ever forget that wireless is 50% science and 62% black magic!
> I am quite hopeful that the new cable, beads on bother ends of the cat5, 
> better grounding, and an equipment move will fix this site up nicely.  If 
> not, who wants to buy it?  lol  Seriously, this site and one of the ones 
> that it feeds cause nearly 80% of my trouble calls.  With any luck it wasn't 
> wireless trouble that I've been having all along.  Shrug.
> Here's the really strange part to all of this.  The hardware that cost the 
> most has the most trouble!  grrrr
> Also, the Day Wireless guy had some kind of wide band sensor.  He could move 
> it around and see where overall signal levels were unsafe for people.  Right 
> beside a 4" or so copper pipe that the radio station used as a wave guide 
> was pretty bad.  A couple of inches from it was ok.  The next highest 
> readings at the whole site?  RIGHT a the BASE of the 100' wooden pole that 
> everything is mounted on!  And what did I do when I installed all this new 
> gear a year ago?  Yeah, I put the antennas up top and the radios right down 
> at the base of the tower so I could get to them without having to rent a 
> lift.  Dang the luck!
> Hope that this helps someone else out.
> marlon
> --------------------------------------------------------------------------------
> WISPA Wants You! Join today!
> http://signup.wispa.org/
> --------------------------------------------------------------------------------
> WISPA Wireless List: wireless@wispa.org
> Subscribe/Unsubscribe:
> http://lists.wispa.org/mailman/listinfo/wireless
> Archives: http://lists.wispa.org/pipermail/wireless/

WISPA Wants You! Join today!
WISPA Wireless List: wireless@wispa.org


Archives: http://lists.wispa.org/pipermail/wireless/

Reply via email to