Grounding WS-12-250-DC

[rkelly1]

I hired a grounding expert to help us with some summer lightning issues.

He thinks we should try to strip the cat5 cables before the enter the switch and drain the shielding and drain wire straight to ground...so it doesn't make it to the switch.

I'm not sure this is viable since the drain wire is so fragile and the foil is iffy.

In theory, would the above setup be beneficial since ethernet surges wouldn't make it to the switch? Seems like it could still come in via the data lines.

Is the above more effective than draining via shielded RJ45/ethernet cable drain wire going through the switch?

Thanks

[lligetfa]

Since the PoE ground cannot be isolated from the switch isolating the foil drain won't help in any way.

[sirhc]

If your suffering damage from ground current this is because service ground rods are NOT sufficient to handle the ground spike of current that occurs and or the service ground rods are not bonded to the tower grounds.

But even if your service grounds are bonded to tower grounds if the service grounds are insufficient to handle the ground current spike and if the bond wire is not sufficient to pass the current without the resistance spiking and causing ground current to seek out other paths to the tower ground rods usually via the Ethernet cables this is where ground current damage occurs.

I "ALWAYS" add 2 new ground rods to a new site's electrical service and clean them up (corrosion). If I lose any equipment I add additional service ground rods until the issue completely stops.

I have a couple towers where we ended up with 6+ electrical service ground rods all bonded together. If your site is rocky and has poor soil conditions when rain events occur and that water wicks down the ground rods it raises the resistances to ground and at that moment ground current sees the tower ground rods as a better path to ground.

So we are trying something differnt with our "AC" and "IDC" but not "DC" switches we are assembling now where we went back to attempting to provide better isolation between AC Earth Ground and DC negative.

We have been able to provide 1500V of isolation but at 60Hz, the problem is ground current can go much higher via harmonics. Right now we are able to keep the isolation up to 500Hz whereas it is only passing 0.1mA at 120V @ 500Hz but 0.1mA can do a lot of damage depending on the voltage.

We will see which works best but ground current has always been an issue in this industry and even in IT which is why so many white papers have been written on ground current or ground loops.

In the IT industry they run copper between service panels with differnt ground potentials and when something happens - POOF

In the WISP industry it is more often occurs because WISPs "think" ground is ground, IT IS NOT. They do not improve the electrical service ground rods and often ground the shit out of towers and a lot never BOND the 2 grounding systems together and when they run the Ethernet cable to a radio on the tower that becomes the BOND or GROUND LOOP. Under normal conditions nothing goes POOF but when there is rain, or say a failing electrical motor that is sending current to ground often at a much higher frequency from harmonics it cuts through isolation CAPs and wrecks havoc on DC components.

Other causes can be you get a ground current surge from the electric company as they hand off 2 hot, 1 neutral, and a ground. Well all the grounds in the immediate area are bonded and all the Neutrals that are on that pole transformer. So your ground current can actually come from a bad electrical motor on a neighbors property.

Or another scenario is a ground strike nearby enters some other site service ground rods and is carried up to poles and across wires to your location.

Rules of thumb:
ALWAYS add 2 new service rods to a new sites existing service rods.
If a NEW site do 4 service ground rods instead of 2.
If at all possible BOND service ground rods to tower ground rods with #2
If you lose equipment add 2 more service ground rods.

I have been a WISP now for 20+ years and ground current damage was always my big problem before I figured out what it was.
Now I only lose equipment on towers when the site owner messes with my grounding.

[rkelly1]

This was another discussion I had with the consultant. If the DC is powered by batteries and the only A/C at the site powers the Samlex battery charger, then the DC negative is the ground reference and tying it back to the A/C ground would create a ground loop? (Obviously a different situation than an A/C switch).

The consultant said he rather see the switch earth ground lug disconnected to avoid the loop. My question was, where does the esd go then?

So in this situation is it better to not to tie the tower ground to the a/c ground (charger may blow but batteries isolate the tower electronics....then tie the switch earth ground to the tower ground (tower ground not connected to A/C ground). But is there a ground loop here because the electronics ground reference is still DC negative?

[sirhc]

So a lot of DC chargers internally bond DC negative to AC Earth Ground or they have filter CAPs that block normal AC current below 60 Hz but if the ground surge is some sort of harmonic it can go much higher then the ground current passes right through the CAP.

Most manufacturers specifically allow anything above 500 Hz to get shunted to AC Earth Ground so that noise such as AM or FM is shunted to ground.

Here inlies the rub, if the manufacturer does not put in filter caps to shunt to ground at higher frequencies noise can then cause issues with electronics but if you do put CAPs on then stray AC ground current can fry electronics.


So we are back at doing overkill on your service grounds to insure that your electrical service ground will always have enough capacity and low enough resistance that stray ground current will not seek out tower ground rods during these spikes.

Next monkey wrench to throw in is economics or price, anything can be built but will people lay down the dough for something that can handle anything when 90% pf the time it is not needed?

[rkelly1]

Thanks Chris. So this addresses minimizing stray ground current getting to the charger, and we are adding three levels of surge protection. With switch running off DC batteries , assuming the ground reference is dc- and the charger a/c not bonded to dc-, does connecting the switch earth ground create a ground loop? I.e. second ground reference?

do you think its best to isolate tower grounds from a/c ground and bond switch to tower ground, or bond them in case of current getting through caps?

Is your thinking to ensure the service ground is low enouh resistance to ensure stray current doesnt make it to it the tower rods and the switch is bonded to the tower rods? Do you think its lower risk to bond a/c and tower grounds or isolate them? (In this situation, this question seems unclear as pooposed to an a/c powere switch where tower and a/c must be bonded) Our goal is to have all ground points at 5ohm/minimum 25 ohms either way.

What would you do?

[sirhc]

Thanks Chris. So this addresses minimizing stray ground current getting to the charger, and we are adding three levels of surge protection. With switch running off DC batteries , assuming the ground reference is dc- and the charger a/c not bonded to dc-, does connecting the switch earth ground create a ground loop? I.e. second ground reference?

Electrical or Ethernet surge suppressors do NOTHING to stop ground current. How surge suppressors work is to clamp excess voltage to ground but if the current is on the ground wire then they can not affect it.

do you think its best to isolate tower grounds from a/c ground and bond switch to tower ground, or bond them in case of current getting through caps?

NO, THIS IS EXACTLY OPPOSITE OF WHAT I KEEP SAYING.
Electrical service grounds should always be bonded to tower ground if at all possible

*read above and all my grounding posts and all my RMA emails*

If you do not bond service grounds to tower ground possibly because they are so far apart then the Ethernet Cable becomes that bond. If you can not bond them then make sure you have EXTRA EXTRA service ground rods. If you lose equipment add more service ground rods. Think about it if you only have 2 service ground rods but 6 tower ground rods then AC ground current can possibly see tower ground rods as a less resistant path to ground.

Is your thinking to ensure the service ground is low enough resistance to ensure stray current doesnt make it to it the tower rods and the switch is bonded to the tower rods? Do you think its lower risk to bond a/c and tower grounds or isolate them? (In this situation, this question seems unclear as pooposed to an a/c power switch where tower and a/c must be bonded) Our goal is to have all ground points at 5ohm/minimum 25 ohms either way.

What would you do?

As said above always BOND service ground rods to tower ground rods when you can with a HEAVY short and direct as possible wire. If you can not bond then make sure there are more service ground rods.

Ground current comes from the service and tries to get to tower ground rods. Ground current can be caused by many different things such as AC compressor motors or any type of electrical motor that can kick off stray voltage which with harmonics can be way above normal Hz which is 50 Hz EU and 60 Hz US being clamped to ground. Higher frequency current will go right through filter CAPs that are in electrical devices designed to clamp NOISE to ground to eliminate it and that noise is VERY low current but current coming from bad motors and such can be a lot more current yet still in the HIGH Hz range same as FM broadcast noise.

Ground current can also come from the pole as all nearby service all share a common ground that the electrical company hands you and other buildings and service close by that share the same pole transformer also share a common Neutral.

[rkelly1]

Thanks, I get it. I have read all the posts tonight understand what you're saying about the grounding and bonding.

In this configuration my question really revolves around whether we're creating a ground loop from the switch Earth ground lug ground reference vs. DC negative reference that the switch is running on. Is there a way to resolve that conflict? Or issue?

[sirhc]

If all the grounds on a site are BONDED there is no loop as it is all one ground.

However if the service grounds SUCK and the tower grounds ROCK then even if they are bonded there will constantly be current going from the service to the tower and in some events that current can stray and try to go through the switch.

The chassis ground lug should be hooked to the earth ground which SHOULD be all ONE ground as they are all bonded.