2
$\begingroup$

I'm installing a transceiver in an office on the ground floor of a warehouse. The antenna will be 45 feet above the office on the roof of the warehouse. The ceiling of the office is a drop ceiling, and above that is empty space until the roof of the warehouse.

My questions is about placement of the polyphasor lighting suppressor.

I know that usually it is desirable to put the lightning suppression as close to the antenna cable entry point as possible, and then to ground it, but I was wondering if it would be acceptable to install the lightning suppressor about 25 feet from the entry point into the building, right before it enters the office building where the radio is, and then ground it with grounding cable clamped to a grounded pipe.

My rationale is that it will be physically difficult to install it right on the inside under where the coaxial cable penetrated the roof, because it will be difficult to access. If it does take a lighting strike, and needs replacement, it will be very difficult to access if its near the topmost roof.

On the other hand, if it's right above the office, and it needs servicing, it's just a matter of using a standard ladder, and popping a ceiling tile. Keep in mind there are no offices above the first floor, only open space above the drop ceiling.

enter image description here

Improve this question
$\endgroup$
4
  • $\begingroup$ Hello Frank, I just did a search and found some other great information and answers. :-) $\endgroup$
    – Mike Waters
    Commented Sep 14, 2020 at 16:27
  • $\begingroup$ I doubt that it matters, but what is "MPOE"? $\endgroup$
    – Mike Waters
    Commented Sep 14, 2020 at 16:29
  • $\begingroup$ What band, what kind of antenna, and how far above the roof will it be? $\endgroup$
    – Mike Waters
    Commented Sep 14, 2020 at 16:32
  • 2
    $\begingroup$ @MikeWaters minimum point of entry, it's where the telco's responsibility ends and the customer's equipment begins. $\endgroup$
    – Phil Frost - W8II
    Commented Sep 14, 2020 at 16:41

2 Answers 2

Reset to default
6
$\begingroup$

The picture you have drawn suggests a lightning suppressor installed in this way will do little to nothing to protect your station.

The suppressor does only one thing: it limits the maximum voltage between the center conductor and the shield. It doesn't do anything to limit strike current on the shield, which is where most of the energy is anyhow.

The protected equipment inside the office is almost certainly grounded either directly or indirectly through AC mains, Ethernet cables, etc. Remember that at the voltages lightning can generate, anything anywhere near something even a little bit conductive (like wood) is potentially a path to ground.

As such, a more complete picture of your proposal is:

enter image description here

Note that you have multiple ground paths, and the one you've added in green has a high impedance because it is so long. As such, most of the strike current will go through your equipment.

As such, most of lightning protection is not about the suppressor at all, but rather:

  1. ensuring strike current has no path to ground that goes through protected equipment, and
  2. establishing a single-point ground so all protected equipment floats at the same voltage.

To be effective, the connection labeled "to grounding" you've drawn:

  1. must be a short, direct path to ground. Not a meandering, horizontal connection as you've drawn
  2. must be the only path to ground. If any of your equipment is mains powered, it's likely that connection provides an additional path to ground, and thus much of the strike current will flow through your equipment to get to the mains ground.

Something like this:

enter image description here

Importantly, all wires must go to a single point ground. At this ground point, everything is bolted to a conductive panel so they remain at the same voltage. Anything that can't be directly grounded (like the coax center conductor, and the non-ground conductors of the mains wiring) go through some protective device like a gas discharge tube or an MOV. And that grounding panel is near the ground, and grounded with a short, low impedance connection.

Ideally, the antenna feedline doesn't run through the plenum, since there's likely all kinds of conductive stuff in there (other wiring, HVAC ducting, ceiling support structure...) which would also provide a nice path the strike current will use, potentially breaking electronics and starting fires as it does so.

For additional detail, see How can I protect equipment against a lightning strike?

Improve this answer
$\endgroup$
7
  • $\begingroup$ Sorry the diagram was't clear. The gronding cable will run to about 50' to a grounding bar in the MPOE or maybe attached to a pipe in the ceiling. $\endgroup$
    – Frank
    Commented Sep 14, 2020 at 15:58
  • 1
    $\begingroup$ @Frank that's far too long. To be effective, that grounding path must have significantly lower impedance to than any other path. That includes paths that involve arcing to building structure or wiring, of which there are certainly many along a 50 foot path through the ceiling. $\endgroup$
    – Phil Frost - W8II
    Commented Sep 14, 2020 at 16:06
  • $\begingroup$ @Frank It looks clear enough. As Phil stated, your grounding cable is too long. Assuming that the ground is sufficient, it would be better to locate the Polyphaser there. The coax would run first to that point on the outside of the building, and from there to your transceiver. $\endgroup$
    – Mike Waters
    Commented Sep 14, 2020 at 16:12
  • 2
    $\begingroup$ ...and importantly, all other grounds (AC mains, ethernet, control wires, ...) do the same. If you don't make a single-point ground, then you're deliberately offering lightning a path to ground through your equipment, and you've gone through a lot of expense for no benefit, perhaps even made things worse. $\endgroup$
    – Phil Frost - W8II
    Commented Sep 14, 2020 at 16:14
  • $\begingroup$ I made changes to my diagram. Would that improve things? If I run the antenna cable to the existing ground I'll be adding at least another 100 - 150 feet. $\endgroup$
    – Frank
    Commented Sep 16, 2020 at 18:26
1
$\begingroup$

Ground wire is very inductive and thus high impedance to sub-nanosecond arc rise time when the leader path finally conducts.

Therefore you want the shortest path to the air gap arc suppressor which is sub-picofarad higher impedance then the current will have a greater chance of following the ground wire.

The voltage induced may still be high but hopefully suppressed by a 2nd stage <~1kV suppressed with lower current in its path in the unit. This depends on coax cable quality, skin effects and mutual induction from the ground current 3kA to 75kA worst case for 1us. (<~1MHz mostly)

• The lower the coax is grounded to the tower, the less shield current there will be on coax line.
• The lower the inductance path to ground from the MGB or bulkhead, the less shield current will enter the building.

The best lightning protector blocks DC to 1MHz which is the bulk of the RF energy , so the equivalent circuit has 0.1 dB insertion loss in the RF band selected and shunts the antenna with an LCL or Pi filter for high pass or low block . An air gap or properly designed rugged proper gas tube is in parallel with the incoming shunt L then HPF with series C and output shunt L.

Caveat

If there is nothing in the roof and below to divert the coax current and rising voltage to another electrical path with flammable materials or other electrical equipment, you may consider your original plan.

But if inside, there will be ground shift pulse with ringing, which now can radiate the diverted lightning energy.

Consult with supplier and show wire type and lengths of all from antenna to Unit and earth gnd.

Gold centre pin contacts are recommended.

Improve this answer
$\endgroup$
1
  • 2
    $\begingroup$ This reminds me of NRC’s development of the world’s 1st portable 7 Tesla MRI in Winnipeg, the Eng designer told me when even in the basement inside a Faraday cage, all the CRT’s in the 9? Floors above imploded visually during tests. $\endgroup$
    – Tony Stewart EE75
    Commented Sep 14, 2020 at 14:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .