part 15.219 low pwr.am rooftop metal roof instalation

I had an FM transmitter (in a weatherproof box) mounted about 6 feet above a metal roof.  When I moved it and mounted it a few inches above that roof, the signal strength improved substantially.  So even with FM mounting your transmitter above a metal roof will make a difference.

If you mount the Procaster just above the roof I don't believe that you will need to physically ground it to the roof - there's a capacitive effect, and it should work well.  If you do ground it, the metal roof (particularly if it's sloped) will likely radiate, and the FCC takes a dim view of radiating grounds.

As for range, that depends on a large number of factors, including (and this has just been discussed recently) sensitivity and antenna at the receiving end, obstructions between the receiver's antenna and your transmitting antenna, background electrical noise, etc.  Plus, there is a listenability factor - what sounds good to one person might sound like crap to another. Noise gets introduced into AM signals fairly quickly as you go any sort of distance from your transmitting antenna, and the signal to noise ratio plus that subjective factor really defines your range.  Good ground mounted installations described here have reported a mile+ range listenable (at least to them) so with all things being equal, you would probably see range in that vicinity, maybe a bit more if there is line of sight to a sensitive car radio.  But the physics of Part 15 radio does limit that range.

There has never been an opportunity for me to try any kind of AM rooftop antenna, but some things I know just from learning about how some things work... The ground plane underneath an AM transmitter has got to be flat; it must be horizontal on all sides. If the roof slopes in two directions, Artisan is right, the two sides will not be perfectly out-of-phase and will add to the radiation, which violates Part 15.219. I would also mention that the AM vertical antenna must be centered on the flat metal roof if you want even coverage in a full circle, but if you want more signal in a particular direction, the AM vertical antenna will put more energy in the direction of the larger portion of flat metal rooftop. I am very curious about the connection between the transmitter and flat metal roof. I know Artisan is right when he says there will be capacitive coupling without an electric-wire connection, but is one better than the other? Or, are both methods equal. Love to try it.

Ground planes need not be directly connected to be effective...it's generally categorized as counterpoise in such a case, and there is truckloads of information about it...but predictabilty on what will happen in a given case is another story.

However, the antenna system and TX still need to be well-grounded for safety. Doing that from a rooftop presents problems because if you ground it to a metal roof, then the roof needs to be grounded, which in turn makes it an above ground ground plane, and it will indeed radiate.

I don't think an uneven non-flat roof, especially if used as unconnected counterpoise, will affect directionality much. Tighter focus of the pattern such as a dish or slotted waveguide would, but I don't see that it would increase much range at these puny power levels in midband frequencies.

Below is a NEC comparison of an elevated Part 15 AM system using a counterpoise with no ground wire to that same system when using a ground wire.

The groundwave field intensity (signal strength) at 1 km is shown below each system, at the bottom of the graphic.

The field improvement from the system using the ground wire is due to the added radiation from that wire, itself, and not from the counterpoise.

NEC shows that the radiation pattern from both of these configurations is perfectly circular in the horizontal plane.  This is as expected, because in both cases the two horizontal wires used as a counterpoise produce no far-field radiation, themselves.  All of the far-field radiation from these systems is generated by their vertical conductor(s), only.

______________

I experimented with a Rangemaster mounted on a flat non-metalic roof using 4 radials about 10 feet long at a height of about 15 feet. The performance was terrible.

What do you mean by terrible?  1/4 mile listenable signal? 1/2 mile?  ??

<1/4 mile

Interesting.  I was able to get approx. a mile in all directions (in some, a bit more) with a ground mounted Rangemaster, no radials (the transmitter grounded to the metal mast only, which was pounded into the soil about 6 feet).  I used a Symetrix mono outboard compressor, and an Inovonics 222.

The soil where I was testing was part of a river valley (so probably ideal for grounding, very moist, no clay).  I suspect that if you used more, longer radials, your results would be a bit better.  But I guess it goes to show that elevated radials won't beat out a ground mounted transmitter (at least when the ground is ideal).

yeah I am going to ground mount the Rangemaster. The soil here in Florida pretty much sucks so radials here I come...

yeah I am going to ground mount the Rangemaster. The soil here in Florida pretty much sucks so radials here I come...

There is a night-and-day difference between ground-mounted radials and an elevated ground counterpoise.

Elevated radials beyond a reasonable distance from ground must be RESONANT for optimum performance. Ground level radials do not need to be resonant.

The commonly recognized Ground Plane antennas that we see so often for CB or VHF operation have resonant radials just below the antenna. These antennas are probably the source of the myths surrounding the commonly discussed concept of Part15 AM elevated radials.

A random non-resonant elevated counterpoise will produce widely varying results depending on how close or how far away from resonance it is at the operating frequency.

To take a simplistic view, a random non-resonant metal roof ground will cause ground current to be divided between the roof and the wire to earth ground. If the elevated ground is perfectly resonant, all the current will flow there and none to earth ground. Otherwise, RF current will also flow to earth ground.

It is well known that an almost perfect elevated ground can be created with just 2 radial wires. The kicker is that they must each be 1/4 wavelength long (exactly resonant). That's really long in the MW band. You could add inductors in series with each radial to resonate them to frequency, but is that practical?

In reality, elevated radials aren't an option for Part 15. Sure, you can do it, but it is far from being practical. It only becomes practical at higher SW and VHF frequencies where the length gets reasonable.

Good point Phil. 4, 146 foot radials not to practical. 🙂