So far we have talked about collecting radio intercepts to be decoded, and also using those intercepts for traffic analysis. The next piece of the puzzle is finding out where the signal is coming from. Signals intelligence made extensive use of high frequency direction finding during WW2. Referred to as HF/DF, or colloquially as Huff-Duff, it made a critical contribution.
Direction finding provides a bearing but not a location. Multiple bearings from different stations can be used to find a location. Pretty soon, the world was dotted with hundreds of HF/DF outposts, organized into networks. For example, the Allies Atlantic network consisted of nearly 50 listening posts from Canada, United States and Britain, with plotting rooms in London, Washington and Ottawa. When a signal was spotted, the operator alerted the network on a separate channel and multiple bearings were taken within seconds. Before Ultra, HF/DF was the best method to spot u-boats and re-route convoys.
Some of the radios were continuous tuned across a band using motors. These would stop automatically when a signal was detected and sound an alarm.
Back in 1888, Heinrich Hertz discovered that antennas were directive. Early direction finding was done by rotating a loop. Still works, but rotating a big loop is cumbersome. In 1907, Italians Bellini and Tosi discovered that you could use a pair of orthogonal fixed loops and just rotate a small pickup coil at the center. These systems worked pretty well at lower frequencies and were used for DF in World War I.
High Frequency Direction Finding Technology
As the shortwave bands were discovered, so was a weakness in loop-based HF/DF. Horizontally propagated waves bouncing off the ionosphere gave false directions as they bounced around. Was there a workaround? There was. In 1917 a British pharmacist named Frank Adcock, serving in the Army as an engineer, discovered something neat. If you replaced the loops with two pairs of vertical antennas, they would receive only the vertically polarized signals and provide true directivity. This system was adopted as a navigational aid.
By WW2, both Allies and Axis standardized on the Adcock system for their fixed HF/DF. Technicians varied the height and spacing of the pairs of vertical antennas according to what frequencies they wanted to cover. Typical dimensions were between 10 to 100 feet. Adcock systems had to be isolated so nearby metal structures did not interfere. Signals from Adcock verticals were also combined with a sense antenna to resolve true bearing.
The second big innovation was the use of an oscilloscope to show direction visually. You simply connected each pair of antennas to either the vertical or horizontal traces of a scope. Depending on direction, the pattern and length on the scope showed you direction. This was invented in 1925 as a means to finding the direction of thunderstorms. Because it was electronic rather than mechanical, the Watson Watt system produced results fast. Operators avoided mechanical steering to find a null. But great care had to be taken not to distort the signals when they were amplified for display.
So, by World War 2 wireless intercept operators could tune in a signal and read a bearing in a couple of seconds. Here is a detailed technical description of how these various systems worked, and a simulation of what Huff-Duff looked like on the screen.