Marconi trans-atlantic wire antenna in 1902

[floobydust]

I was looking at the early trans-atlantic radio experiments and wondered about the antenna Marconi used in Glace Bay and Poldhu. It looks like the freq. is unknown.

'The station constructed in 1902 consisted of four wooden towers 64 metres (210') high, set in a square-shaped 64 metre square piece of land, which together supported an antenna of copper wires in the shape of an inverted pyramid.'

To me the antenna design looks upside down; very Frankenstein looking station (12kV battery bank  ).
I am wondering if anyone knows the theory for this antenna.

http://www.newscotland1398.net/nfld1901/marconi-nfld.html
http://www.krausehouse.ca/krause/FortressOfLouisbourgResearchWeb/marconi/marconi1.htm
Wire Antennas for Ham Radio

Glace Bay Antenna pic taken from HMS Collingwood’s Historic Collection {CHC} of Naval Radar and Radio equipments

[CopperCone]

is it a giant desicone

[FlyingHacker]

Doesn't the disc-cone go the other way (inverted vertically), and have a disc at the top?

The top image is a drawing on top of a photo, not a plain photograph.

[cdev]

It actually makes more sense this way than the traditional way.

[T3sl4co1l]

Earth serves as the ground plane in the discone.

It's not symmetrical (because pyramid corners), so the exact pattern will be a little odd.  It will be mostly omnidirectional, which seems less desirable for a transatlantic link, but oh well.  It won't be radiating into the atmosphere, at least (except at the horizon, where you then get ionospheric skip, which is actually the point).  It will also be wideband, which was probably more helpful than not, in those heady days of spark transmitters.

Tim

[Howardlong]

I'm not an HF guy, but have been to Poldhu where there are still a very few concrete remnants of what once was, and I assume the small Marconi Centre is still going.

Did they actually know what they were doing with those antennas, or were they the results of largely empirical investigation?

[rfeecs]

Did they actually know what they were doing with those antennas, or were they the results of largely empirical investigation?

I'm curious about that, too.  I found this article:
"Guglielmo Marconi's antenna design" by A.G.P. Boswell of GEC-Marconi Res. Centre, UK
http://ieeexplore.ieee.org/document/608674/

Abstract:
The article reviews the antenna design used for the earliest experiments employing untuned systems, culminating in the December, 1901 transatlantic experiment from Poldhu, UK to St Johns, Newfoundland. To understand the system it is necessary to consider the method by which signals were detected. The filings coherer was a device on which Marconi (1967) himself was probably the leading practical expert. This uses an evacuated glass tube partially filled with filings (the material was actually various combinations of nickel and silver), and containing two electrodes immersed in the filings. To compare the performance of various possible antennas we have considered four possibilities which are: a monopole, a Tee antenna (i.e. a monopole with a horizontal top-loading wire), a fan antenna and a cone antenna. All were 43 m high, and the last three had a horizontal extent of 45 m. Any supporting structures were not included in the simulation.
Published in: Antennas and Propagation, Tenth International Conference on (Conf. Publ. No. 436)

Unfortunately, it's something like $33 for the article.  I'm not that curious.

I have a feeling the author simulated different antenna designs and drew some conclusion.

[floobydust]

Thanks Coppercone, that was enough for me to dig into it more. 
I was wrong thinking it was a well-tuned system, and the upside-down cone also had me confused, as I thought radiation pattern went downwards.
snippets from http://glendash.com/Dash_of_EMC/Modeling_Antennas/Modeling_Antennas.htm:

"Marconi’s antenna was known as a discone, a cone shaped structure placed vertically over a ground plane.  Combining two such cones back to back produces a bi-conical antenna.  Though the general characteristics of conical antennas have been known for almost a century, precise mathematical models predicting their behavior proved illusive.  Prior to the development of computational techniques, researchers had to rely mostly on experimental data to design their discone or biconical antennas. 

"The reason that conical antennas proved superior was that Marconi’s spark transmitter was a broadband source.  When combined it with a broadband antenna, Marconi was able to transmit far more energy than he could of have, had he been limited to the resonant frequency of a single wire."

Antennas by Kraus:
"Sir Oliver Lodge constructed a biconical in 1897, while the single cone working against ground was popularized by Marconi and others"


It looks like the antenna was (kinda empirical?) design that was broadband and had reasonably tame SWR above LF cutoff.

I'm amazed at the energy transmitted 2,100 miles, as the Coherer detector does not look very sensitive.

[CD4007UB]

Here too are some short extracts from the paper about Marconi's Poldhu transmitter:

"The transmitted pulse was produced by sparking a previously charged antenna to ground, thus radiating the stored energy. It was known that the stored energy could be increased both by raising the voltage and by increasing the capacitance of the antenna. Probably Marconi himself favoured the simple monopole antenna. However Fleming, who was given charge of the design of the Poldhu experiment, conjectured that the instantaneous current flow could be maximised by providing many parallel current paths to ground. This led to the use of cone and fan antennas."

So, the transmitter was for sending HF pulses (Morse code) without modulating a carrier. The pulses were detected at the receiving antenna by a 'coherer', based on conduction in a metal powder.

Since the original design is not known for certain, the author, A.G. Boswell from GEC-Marconi, simulated various possible aerials with NEC-4 (s) in the frequency range 62.5kHz - 32MHz and concludes:

"The results show that modern simulation confirms the superiority of the fan and cone antennas that replaced the earlier monopoles in radiating the fast pulses that were required to operate the original coherers. The fan and cone are good radiators over a wide frequency range, and indeed are still the preferred antenna in use at present for radiating picosecond pulses for time-domain microwave applications."

Marconi's pioneering work on wireless was semi-empirical, but to make the first transatlantic broadcast was a monumental achievement. Indeed, when I was a second-year physics student (many years ago), my tutor told me about his PhD viva with Lord Rutherford, who had asked him "Was Marconi an ass?" By 1900, physicists knew that electromagnetic waves travel in a straight line (ignoring diffraction), so they could have told Marconi that the curvature of the earth would make it impossible to transmit a signal from Cornwall to Newfoundland. If anyone told him, he (fortunately) didn't listen. The ionosphere, which reflects the HF radio waves, was not discovered until much later - and, of course, Rutherford did not think Marconi was an ass, but it was a good question for a viva.

[CopperCone]

ah, yes, I imagined the 'disk' to be earth in this case. I wonder if he wanted to do a comparison study of across the united states and across the water, perhaps also wanted to see how it got to japan or china? I'm sure they understood that it radiates omnidirectionally.

I wonder if he wanted to put up a corner reflector facing the ocean...

[bjcuizon]

I was looking at the early trans-atlantic radio experiments and wondered about the antenna Marconi used in Glace Bay and Poldhu. It looks like the freq. is unknown.

Here's some articles that one might be interested in:
https://www.ieee.ca/millennium/radio/radio_differences.html
https://en.wikipedia.org/wiki/Guglielmo_Marconi#Transatlantic_transmissions
http://ethw.org/Milestones:Transmission_of_Transatlantic_Radio_Signals,_1901
The top article says that in 1901 he radiated waves at ~850kHz.

[floobydust]

Marconi's peers said it could not work due to curvature of the earth.
I think he was very empirical, compared to Hertz and Tesla. So somehow he knew "skip".

He had no witnesses so who knows if it actually worked. Later, he officially proved it on a ship 1,550 miles out, at night.

Coherer detectors seem to need 2-3V on them, a lot of signal over 2,100miles.

Scary that Marconi was offered a ride on the Titanic but instead took the Lusitania 3 days earlier. His radio system was invaluable during the sinking, I believe communicating with rescue ship Carpathia.

[cdev]

I've done many "squared off" versions of these frequency independent antennas and they are much easier to construct and tend to work just as well. In the case of the archimedean spiral the square shape is actually electrically a meander line which has the effect of slightly reducing its size.



Quote from: T3sl4co1l on 2017-09-14, 06:04:11

Earth serves as the ground plane in the discone.

It's not symmetrical (because pyramid corners), so the exact pattern will be a little odd.  It will be mostly omnidirectional, which seems less desirable for a transatlantic link, but oh well.  It won't be radiating into the atmosphere, at least (except at the horizon, where you then get ionospheric skip, which is actually the point).  It will also be wideband, which was probably more helpful than not, in those heady days of spark transmitters.

Tim

[cdev]

That would likely work well, it would be huge though, of course.

What ended up being the best antenna for transoceanic broadband communications was the rhombic antenna. They still are used in shortwave broadcasting. The beverage antenna has some similar aspects. The far end towards the signal is terminated with a resistor.  You can turn a rhombic on its side, basically a half-rhombic. Similarly with LPDA, you can do a half LPDA.

A discone with the cone forming the vertical radiator I think potentially is better than with the cone on top. Remember the discussion we had a while back whout the biconical impedance? The same thing applies for this antenna, if you can make the top wider you get the benefit of a lower angle of radiation over a wide range of frequencies.

I know a guy who actually built a HF discone, he used four telephone poles to support it. It survived major storms that brought down his other antennas but not the discone.


He also pulled off a 40 kilometer wifi link, over salt water, using a commodity off the shelf Linksys access point.

Quote from: CopperCone on 2017-09-14, 13:37:41

ah, yes, I imagined the 'disk' to be earth in this case. I wonder if he wanted to do a comparison study of across the united states and across the water, perhaps also wanted to see how it got to japan or china? I'm sure they understood that it radiates omnidirectionally.

I wonder if he wanted to put up a corner reflector facing the ocean...