Reprinting a item from two years ago. It’s not relevant to present shit, but the line of thinking was fairly original, and it wants to be reprinted for reasons I don’t understand yet! (Possibly to help with my courseware project animating brain networks?)
Needless to say, like ALL of my fairly original and well-researched items, nobody ever looked at it. Now I can be annoyed again because nobody will touch it this time either.
= = = = = START REPRINT:
Multiple choice quiz. Who invented the earth?
A: God, 4004 BC.
B: Random quantum fluctuations evolving atoms and molecules and planets, at a quantum indeterminate date and time.
C: Carl August Steinheil, 1837.
Correct answer is C.
= = = = =
Steinheil was a physics prof who got interested in the new sport of telegraphy, and built his own device. Other telegraphers had tried to use the earth as one side of the circuit but didn’t try hard enough, and stayed with a more reliable closed loop of wire on both sides.
The earliest telegraphs were ‘massively parallel’, with a pair of wires for each letter. When you’re running 52 wires, trimming it down to 26 wires wouldn’t really matter much. In either case they needed a huge conduit or trough, and a whole lot of insulation which wasn’t yet manufactured or understood properly. Even Wheatstone’s elegant and practical telegraph needed 5 separate wires, so adding or deleting a sixth wire wouldn’t make much difference in cost.
After Morse and Breguet switched from parallel to serial, the distinction between one wire and two wires started to make a real difference in percentage terms.
Steinheil was the first to formalize the proper use of ground, and the first to develop effective ways of grounding. His telegraph was distinctive but not very successful.** He’s remembered for inventing the earth.
= = = = =
One aspect of his device was unique and not repeated. All other telegraphs used batteries, either chargeable or dry. The sender closed and opened a circuit to carry the battery’s juice. Steinheil’s device was self-powered and self-contained. The sender was a powerful DC generator. Each ‘keying’ was a half-crank of the generator in one direction or the other.
Here we have Polistra ready to send, HappyStar ready to receive, and our Martian supervising. Both telegraphs are identical, and each contains both sender and receiver.
In this simple setup there’s really no need for ground; it would be easy to run two wires for a few feet. Still, we’re showing the use of the ground path.
The telegraph has three parts.
The left section, shown partly transparent, is the receiver. Steinheil used the recently invented galvanometer principle, with a magnetic coil surrounding two permanent magnets, each pivoted, each moving a pen onto a strip of paper. The paper, as in Morse’s original prototype, is driven by clockwork to move regularly. The takeup and source reels are under the table.
The middle dial looks like it should be the sender, but in fact it was a sort of rotary switchboard to select from several stations, and to select send or receive. Steinheil’s experiment involved several stations in Munich, connected by single wires strung between belltowers and other high structures.
Here’s the innards from Steinheil’s original drawing. I didn’t try to build or animate this.
Multi-layer rotary switches were REinvented in the 1930s, as seen in this Triplett meter schematic.
= = = = =
The big lever on the right is the sender. It turns a dynamo with a commutator, again relatively new in 1837. This is the unique self-powered feature, and also the part that fits best into the intuitive concept of ground.
Here Polistra is sending a short message. I’m not sure why Steinheil used this barbell-style lever. A rotating crank handle would have been easier to use in real life, and also would have been easier on Polistra’s animated hand (for technical reasons in Poser) but she agreed to tolerate the strain for the sake of science.
Now a closeup of the receiver mechanism. Paper was pulled across the viewer area by clockwork under the table. Steinheil’s code resembled Morse, with dots and dashes replaced by opposite polarities. Two permanent magnets are aligned so that each pivots inward for one direction of the current through the coil. Each magnet carries a little capillary pen-tube. When the magnet rotates inward, the pen makes a dot on the paper.
= = = = =
As usual, building this device gave me a chance to think deeply about grounding. In 70 years of dealing with electronics, I’d never really stopped to apply intuition to the question.
We normally start with the partly mistaken notion that the ground simply completes the circuit. The wire carries electrons from sender to receiver, and the ground carries the same electrons back from receiver to sender.
This notion is valid for chassis grounds, which are not actually grounds. A chassis ground uses the metal case, or a large flat area on a printed circuit board, as a wide wire for all return paths. It conducts in the same way as the visible wires or traces, and doesn’t necessarily connect to the earth.
= = = = =
Water and air analogies give a better understanding of real earth grounds. The earth DOESN’T complete the circuit and DOESN’T carry electrons back from B to A after the wire carries them from A to B. Instead, the earth is an effectively infinite source and sink.
We use this notion often for air-moving devices. Here the double-hung sash is pulling in cooler air from below, transporting it around the room, and letting it out as hotter air above. A few of the molecules may accidentally complete the circuit, but most will simply rejoin the infinite mass of the atmosphere.
The air model also lets us compare the behavior of earth for DC and AC. Consider a tube with a closed end. A direct wind will NOT go into this tube because there’s no way out. Ventilation is DC, so it requires a way out. A sound wave WILL go into the closed tube and bounce back and forth in the FINITE elasticity of the tube. Steinheil was thinking bouncily when he developed his two-way code.
= = = = =
A refrigeration system absolutely requires a closed circuit. It uses a toxic material like ammonia or Freon, and it has to switch between high and low pressures under strict control.
Aside from chassis ground, most real uses of grounding are like the double-hung sash with infinitely elastic sink. Noise-reducing grounds rely on the ability of the earth to absorb high-frequency ‘trash’ without returning any electrons or movements back into the protected device. If the earth was a simple return wire, these grounds wouldn’t help.
= = = = =
Steinheil’s experiments showed a water-model intuition. He paid close attention to the water table, and determined that the best ground was a large copper plate sunk below the ‘low tide’ level of the local water table. He seems to have been thinking in closed-loop form, but the best open-loop analogy also comes from water tables and wells. His closed loop was later used literally in the WW1 French ground telegraph.
A well irrigating a crop is the water analogy of the double-hung sash or fan moving air. The well is pulling from an ‘infinitely elastic’ water table and pumping water into the ‘infinitely elastic’ sink of the plants. Some of the water will return directly to the local water table, but most will end up as part of the proteins and sugars created by the plants, and then ultimately return to the earth via decomposition.
= = = = =
Credit footnote: This is the best online source, around p100 in the PDF.
Irrelevant footnote: Steinheil is a perfect aptronym for the man who cured the grounding problem.
** Later in life Steinheil got interested in the new sport of photography, and invented a telephoto lens that did gain tremendous commercial success. His son started a company that lasted into the 1970s.
Polistra's Mill 