In EE World videos from the 2023 International Microwave Symposium, engineers received a history lesson on early transatlantic cables through an analysis using today’s methods. Plus, a demonstration of a 1944 radio using a razor blade as a rectifier. Modern equipment provided an automated Morse code signal.

With the Titanic recently in the news, you may have heard James Cameron talk about running simulations on the design of his submersibles. Cameron claimed that his designs used finite element analysis and other mathematical models. Had the Titanic designers had such analysis tools, they might have realized that the doomed ship would sink if it hit an iceberg.

Similarly, what if the engineers and scientists who built early undersea cables had the engineering knowledge and tools we have today? That’s what Ed Godshalk wondered, so he set out to discover why early transatlantic cables failed. He reported his findings at the 2023 International Microwave Symposium (IMS) plenary session.

According to Godshalk, a lack of knowledge about transmission lines and how they affect electrical signals was the major contributor. The other problem, of course, was that electrical wire wasn’t so readily available. Fortunately, some wire was accessible because it was used in women’s hats.

The breakthrough first came in the early 1800s when batteries and wires enabled current-on-demand. That brought on dots and dashes — Morse code — what today we’d call “on/off modulation.” Speeds of 20 words per minute (4 bits/sec) were possible. That brought on the telegraph, with wire strung on poles but having a resistance of about 500 Ω per 100 miles. Telegraphs used the Earth for the return path, so just one wire was needed. Of course, only one transmission at a time to one receiver was possible through each wire. “They only used 100 V at the time,” said Godshalk when commenting on the signal loss. “I don’t understand why they didn’t stack more batteries to achieve longer distances.” Scientists of the day did understand the need for impedance matching to maximize distance. The voltage at the receiver end was just enough to excite the sounder. The photo shows the transmission line schematic.

Godshalk explained why the first successful undersea cable worked after others had failed. Unfortunately, delay and dispersion reduced the data rate to 5 wpm. Michael Faraday’s 1854 paper explained the problem but not how to solve it. Sir William Thompson (“Lord Kelvin”) used Fourier’s Heat Equation, substituting electrical resistance for thermal resistance and charge for heat. He also accounted for capacitance per distance. That let him model current in the cable over time and distance. Thompson’s 1854 transmission-line model looks just as it would today.

From these equations, Thompson could predict pulse spreading — showing how a current pulse spreads as it traveled in a cable. He predicted that the time of peak current from the launch would occur based on the square of the distance from the transmitter. This spreading over a transatlantic cable and a loss in amplitude — think 100 V across 25 kΩ (2000 miles) — would reduce the data rate to an unacceptable 0.5 wpm. Investors were not happy. “You can’t drive an electromagnetic receiver under those conditions,” said Godshalk. You need a different kind of receiver.” Thompson’s Mirror Galvanometer solved the weak signal problem, but not the spreading signal problem. Sine waves, which are not subject to dispersion, solved that problem. The receiver is simply a peak detector and doesn’t care about the waveshape.

In the video, Godshalk goes on to explain how Thompson and others solved problems and invented the equipment needed to make the successful 1858 transatlantic cable. His talk covers the technical, business, and historical aspects of undersea cables. It’s well worth 38 minutes of your time.

Godshalk’s talk wasn’t the only “fun” part of IMS 2023. An exhibit on the show floor provided some history of radio and some hands-on fun. One such exhibit was a design based on a World War II radio that used a razor blade as a rectifier. In this case, a waveform generator provided the Morse-Code modulation signal for a loop antenna. As the video shows, when you move the transmitting antenna inside the receiver coil, you can hear the code. It’s much faster than 0.5 wpm.