Industry Forums > Wireless Service Providers > [Equipment] Interesting grounding situation

[Equipment] Interesting grounding situation

Be aware that shielded CAT5 wire will still pick up induced voltage from strong electrical field via the shield. Similarly the shield will also see any ground voltage ripple or ground voltage rise during lightning storms.

To protect the radio long-term, you need to install a PoE surge protector which has following specs:

1. low-voltage (ideally less than 6V) clamping on the 2 data pairs. There are specialized LAN protector devices from Semtech, Tyco, Bourne, Littlefuse, etc.

2. medium-voltage clamping (ideally less than 150% of PoE volt) between PoE positive and negative pair. This can be done using either silicon device or low-voltage GDT.

3. Finally high voltage clamping (using GDT or another robust method) between PoE positive and negative pairs to real ground - this is the main surge protection for the radio hardware. [ Note: some commercial products connect the PoE negative pair (7,8) directly to safety ground which is a serious mistake - there must be a protector in between! ]

reply to superdog
Interesting. I bet there is an underlying issue with the ground system there. Your customer needs to have it looked at, as you are now in violation of NEC when it comes to your grounding system and someone could be hurt.

reply to superdog
have you tried plugging your equipment in an isolation transformer?

reply to superdog
1st place to look is verify the grounding of the building and the electrical grounding. 10 to 1 odds that arc unit is either not properly grounded, or isn't functioning properly. (Or they're doing something massivly stupid like allowing the ground lead of the arc welder (Whatever) to come in contact with the main building ground somehow)

I've seen this where bonehead welders just attach a welding lead to their metal desk, which is attached to the building ground because they have a plug mounted on the desk which is grounded to the building ground.

I've seen this in a number of machine shops where grounds aren't done right or people are just doing downright stupid things with it. And then they wonder why breakers pop like crazy and computer and other electronics jam or act weird.

Your first bet is examining this arc unit and chances have it you'll find an improper ground or someone doing something they're not supposed to.
--
Justin - DSLR resident grouch and Mr Negativity
TSI Fanboy - "Dontchya wish your 'net was hot like mine! Ohhh Dontchya!"
Have a nice day!

reply to superdog
Oh lordy... That's not sounding good for sure.
I don't know much about those things, but they look like they could act like a very powerful spark gap transmitter in a way? Damn that thing must piss off neighbors

reply to kewlkeed
We have not had that problem at one welding shop we serve. And they have an outlet on thier table The building is a metal one... without additional grounding to the radio. I would suspect a bad welder.

But there is a reason you pull the negative battery line of your car prior to welding on it.. bye bye ECM most of the time.

reply to superdog
I'm surprised nobody has filed a complaint. I know if I was on that transformer I'd be ripping mad, especially if it broke anything of mine. The unit itself probably isn't operating properly for whatever reason. The more I look this up, the more I'm seeing articles of people building their own. That's a scary thought of itself lol.

reply to superdog
An arc stabilizer is a modern implementation of a Tesla coil. It produces a very high electric field at high voltages. It is used to start an arc for DC and to maintain a stable arc in AC welding.

Just like any Tesla coil, it is a powerful, broadband transmitter, and makes a monsterous amount of RF energy. Enough to light a flourescent lamp held in your hand, for example. Nearby sensitive elecronic equipment often does not have a chance in the presence of such a strong electric field, as you have experienced yourself. Especially if any of the conductors or cables happens to be of a length close to the output wavelength of the stabilizer.

Try this: Measure your existing CAT5 cable and make a replacement that is 1 third longer. That might help avoid being a resonant length. And NO loops in the cable.

Just a thought...

Jerry
WOOF!
--
Wireless from ISP 11 miles away via Motorola Canopy 900MHz, then over 1200ft. wireless bridge of 2 WRT54G/dd-wrt with 15db panel antennas|HP pavilion 2.1Ghz P4 XP-Home-|LAN:Allied Telesys switch|Clients:2 ea. HP Pavilion XP/2.2GHz Acer laptop

There are established EMC procedures for arc welding setups and millions of machines shops who follow them do enjoy broadcast radio, TV and even WiFi. The compliance burden must be on them when there are any EMI problems!

Said by EMCguy:
from US Gov site explains why "Fluorescent Tube Glow in High Potential Field" ... it is due to high voltage electrical field (at DC or VLF) rather than high RF energy.

Yes, the electrical field is what lights the light, but a Tesla coil does not only create an electric field but also a magnetic field = RF or "electromagnetic waves".

Jerry
WOOF!

Wireless from ISP 11 miles away via Motorola Canopy 900MHz, then over 1200ft. wireless bridge of 2 WRT54G/dd-wrt with 15db panel antennas|HP pavilion 2.1Ghz P4 XP-Home-|LAN:Allied Telesys switch|Clients:2 ea. HP Pavilion XP/2.2GHz Acer laptop

For the concerned parties above, I found reasonable advice on reducing HF emission from home-made arc-stabilizers.
»www.hobartwelders.com/weldtalk/s···?t=13859

Just an FYI, I ran the cable outside, across the peak of the roof, down the side of the building where the computer is at inside metal conduit, through the wall in metal conduit right to the computer. I used shielded CAT5 and hooked the drain wire at both ends to the building ground. So far?, so good.
--
»www.wavecrazy.net

Am I wrong to believe that the drain wire should only be connected to ground at one end? I was thinking a loop was a bad thing, but I've been wrong before.

reply to superdog
The drain needs to be bonded to the ground system at one end only. Typically this means only at the base, before the wiring enters the building, or inside at the LPU/demarc location.

Bonding at both ends creates a loop condition, and possible other issues due to both ends not being at the same potential.

reply to jkrusel