We shared this tank top with Sprint for many years. They installed an excellent HALO grounding system which we tapped off of. We had never lost a radio on this tank in well over 10 years until this past year.
This past summer Sprint left the tank and hired a salvage company to remove the equipment and wire/coax.
They were supposed to leave the grounding system intact but obviously the 200' of #1 Copper ground wire that ran from the lower ground bus to the upper ground bus was just too tempting for them to take for salvage.
Shortly after they left the tower we lost our first radio in 4 years which I made a post about on this forums last spring/summer but we did not understand completely what was going on at that time.
When a storm came through we lost an AF24 and it also fried the current sensors on Port #1 and Port #2 feeding the (2) AF24 radios which are located right next to the original ground bus bar. The switch continued to function just the current sensors were blown.
At first we were completely ignorant that the #1 ground wire had been removed so we replaced the AF24 and the WS-24-400A.
A couple months later we lost the current sensors again on Port #1 and #2 of the switch so we took a closer look and this is when we discovered the #1 ground wire was missing so my crew quickly ran a new #2 green up the inside of the tank as that is where our wires/conduit runs whereas Sprint ran their wires up the outside of the tank in a cable trough.
A couple months later a heavy rain event occurred and we once again lost the current sensors on Ports #1 and #2 so again we replaced the switch and decided that in the spring time we would redo the grounding thinking the heavy rain was gone until spring.
On Friday night the BIG snow storm and high winds rolled into our area and the AF24 units were going down losing Ethernet connectivity and forcing our service to roll over to our backup MIMOSA link that comes into another tower. The first thing I noticed was the current sensors once again were blown on Ports #1 and #2 on the switch and the AF24 Ethernet links had down graded to 10 Mbps and were getting tons of errors forcing OSPF to fail over to our backup link.
I was able to restore 1G connections by powering off ports #1 and #2 and then power them back up and 1G connectivity was restored but when the HIGH winds were blowing the links would start getting errors and eventually step down to 10 Mbps forcing the traffic to once again fail over to the backup MIMOSA link on another tower.
Eventually I got tired of fighting the wind and restarting the links and simply shut those links off until the winds stopped the next morning (Sunday/Today).
My assumption is that the high winds and snow was causing a static charge that saw a shorter/better path of less resistance to ground via the Ethernet cable instead of the intended ground wires. This charge was causing Ethernet errors and fried the current sensors on the switch ports #1 and #2.
So tomorrow (Monday) we are starting the process of adding a new #2 green that goes from the original lower ground bus to the original upper ground bus next to the AF24 radios. We will also make sure there is a sufficient Ethernet service loop on the AF24 radios to insure that the Ethernet path (cable length) from the radio to the grounding system is longer than the intended grounding path (cable length).
It is imperative that the intended grounding path is shorter and of much lower resistance than the Ethernet cable provides. This was a perfect example how high winds can cause static charges to build up and interfere with Ethernet communications and also damage the current sensors or could even damage an Ethernet port in your devices and or switches and this type of damage is not covered under warranty.
I also noticed last night during the storm that we were experiencing Ethernet errors but no damage on other ports and on other tower switches which would indicate that the grounding system was working to prevent damage but probably not good enough to prevent Ethernet errors from showing up on interfaces in the high winds. Since the storm and winds have passed I have cleared the errors and have not seen any return on any of the towers.
People should also review this post: viewtopic.php?f=17&t=1786&start=30#p13447
IMPORTANT FACTORS:
Ground wire runs should be shorter from antennas to ground rods then Ethernet cable run from radio to ground rods
Ground wire should go through as few of connections/splices as possible because these splices/connections insert resistance and you want the intended ground wire to be of as least resistance as possible.
Use Ethernet service loops at radios to increase Ethernet cable length/resistance to insure it is a longer path than intended ground wire path.
THIS WAS THE ORIGINAL GOOD GROUNDING
THIS WAS WHAT SPRINT LEFT US WHEN THEY LEFT
THIS WAS OUR FIRST HURRIED FIX WHICH FAILED
THIS IS OUR NEXT ATTEMPT - SHOULD WORK
THIS IS THE DAMAGE TO THE WS-24-400A SUFFERED LAST NIGHT - NOT COVERED UNDER WARRANTY
THIS IS THE TOWER LAST 4th OF JULY
