As is so often the case, a picture is worth lots of words. Here's a very simplified diagram of a tape deck and related items set up for tape echo production.
tape echo schematic
Figure 1. Tape deck setup for echo production. A signal from a musical instrument enters the record amplifier as shown and then goes to the record head which places the signal on the tape. As the tape moves from left to right, after a short time interval, the tape passes under the playback head which picks up the signal from the tape. The signal next goes through the variable gain playback amplifier and then through a high pass filter. Because tape echo gives a distinct, repeating echo, the effect of this on bass notes can often mess up the rhythm or beat of the music.
     The high pass filter prevents low frequency (bass) tones from going either to the speaker or to the record amplifier, whereas the higher notes of the music such as the accompaniment and the melody do get through which is the desired effect. The diagonal arrow through the high pass filter means that it is adjustable, the exact setting depending on what is being played and the musician's or recording engineer's choice. The playback amplifier likewise has variable gain. If the speed of the tape is 7.5 inches per second, which is one of several standard speeds for many reel-to-reel recorders, and the distance between the record and the playback head is one inch, there will be a delay of 133 milliseconds from the time that any signal is recorded on the tape until the delayed signal is heard at the speaker. This is a little over a tenth of a second.
     Therefore, any musical instrument sound that gets recorded on the tape will be played back after 133 milliseconds. This gives one distinct repeat of the original sound delayed a little over a tenth of a second. Now, by feeding this first delayed signal back into the record head, this will put a replica of the original signal back onto the tape, which will itself be played back again after a delay of another 133 milliseconds. Thus, a person listening to this would hear first the original sound of the instrument and then two separate repetitions, the first after 133 milliseconds, and the second after 266 milliseconds. It is super important here to make sure that each subsequent repetition of the signal on the tape has a lower volume than either the original or the preceding repetition in order to produce a realistic trail of multiple repeat echoes that fade away gradually.
     Adjustment of the level of the delayed signal which is re-recorded on the tape controls how many actual repetitions the resulting echo effect will have. If the delayed signals are reduced very slightly, there will be many repeats, whereas if the delayed signal is reduced more, there may be only one repeat. If, by some chance, the gain is set too high, the delayed repeat will be louder than the original sound. Under this condition, the echo effect will quickly degrade to a series of distorted noises that will bear no resemblance to the original instrument sound. Obviously this is a condition which must be avoided.
     If you are using a standard tape deck as an echo producer, adjusting the playback level controls the number of possible repeats, and changing the tape speed varies the delay time. Most commercial reel-to-reel machines have tape speeds of 3.75, 7.5, 15, and 30 inches per second, although not all machines have all four of these speeds available. In our above example, these would give delays of, respectively, 266, 133, 66.5, and 33.25 milliseconds. From personal experience I can state that a delay of 33.25 milliseconds is too brief to be effective, but the other three delay times are all useful for various effects.
     Some manufacturers, however, developed tape machines specifically for echo production, and some of these were quite elaborate. On many of them, it was possible to move the playback head closer or farther from the record head and thus get a wide range of delay times within the capabilities of the machine. Still another advancement was to include multiple playback heads which sent their outputs to different speakers. In addition, these extra playback heads could also send their signals back to the record head. The results from these machines were excellent, producing a nice multi-channel stereo effect that approached certain types of large room reverb patterns.
     Not all tape decks can produce tape echo. Many cassette decks and low-end reel-to-reel machines use a common record/playback head and thus are useless for the echo effect. In order to produce tape echo, there has to be a separate playback head and also a playback amplifier which can operate independently of the record head and its amplifier. Commercial reel-to-reel machines include these functionalities because it is important for the recording engineer to be able to monitor the signal recorded on the tape as the machine is recording so he can make any minor tweaks as necessary while the recording is going on.
     It is interesting to note that if you are making a recording on a machine that is simultaneously functioning as an echo producer, the recorded version on the tape will have the tape echo included but with one difference. There will always be one fewer repeat on the recorded version that you actually hear when making the recording. The reason for this is that the first echo repeat that you hear is actually the original recorded signal on the tape. If you had the controls adjusted so that while recording you heard only one delayed repeat, there would be no echo recorded on the tape.
     There is also a related tape echo effect that Ken Griffin used on a few of his recordings, and that is a type of "plucked echo" effect. This is done by allowing just one repeat and by not using either high pass filtering or any feedback to the record head. The single repeat is played back at a considerably higher volume than the original sound. The end result of this is a very soft initial sound followed by a louder repetition after a short delay. This louder delayed sound actually becomes the main signal for the music! On certain types of music this gives a kind of plucked sound which is useful to achieve a specific effect. If one is using a mixer and has the tape deck properly connected to the system, it's also possible to have a few delayed repeats follow the initial plucked effect. The drawback of doing this is that for some, it is difficult to play accurately, because the musician hears a considerably louder repetition of everything he plays which is delayed by a little over a tenth of a second.
Page 2.