The Art & Skill of Radio-Telegraphy

-Second Revised Edition-
William G. Pierpont N0HFF

Back to Table of Contents

Chapter 10 - Other Keying Devices And Their Use

In Chapter 9 the regular "straight" keys were discussed at length.  Here we look at all  keying devices.

These may be classified as:

All Kinds of Keys

Innumerable variations of simple mechanical switches may be devised.  Almost any conceivable kind of motion may be used to operate the switch: up and down, sidewise, sliding, squeezing, twisting, etc.   They may be actuated by: human action (finger, hand, arm, foot, lips, neck, breath pressure, etc.),  mechanical or electromagnetic action (e.g., in a relay, to duplicate the keying patterns in a second circuit), etc.

For the handicapped several kinds of keys have been devised to be operated by breath pressure on a diaphragm or piston, etc.  Some of the interesting recent designs take advantage of solid-state circuitry using such things as a) the interruption of a light beam by a finger tapping in front of a photo-sensitive cell, b) the change of capacitance or resistance produced by moving a finger  to approach or lightly touch a fixed metal pad, c) the tone of a human voice humming in  Morse code within the range of a tiny microphone, and other possible means to control the keying.  How does one classify such devices?


After the "original" Vibroplex speed key was introduced in 1904 it soon began to be widely used by commercial telegraphers (see "Bug" below).  Those who used it found its sidewise movements much more comfortable and natural,  and also that it prevented or reduced the risk of glass arm.  However, it was rather expensive.  This idea of easier and possibly faster sidewise movement later led to the simpler, so-called "double-speed" key (also called "sideswiper", or "cootie key"). The paddle of its key-lever closed the circuit whether moved to the right or to the left. It sidewise movement offered relief to operators who had suffered from "glass arm".   (A few operators, following this idea, simply turned their straight key around 90 degrees so it could be used with a one-way sidewise motion.)

In use the operator formed the code characters by using his thumb and forefinger to press the paddle alternately from one side to the other as he formed the successive dits and dahs for each character.  For example, if he typically made the first element of a character to the left (L), whether it was a dit or a dah, its next element was made to the right (R), and so on, so that his pattern of movement was L-R-L-R-L-R . . . or R-L-R-L-R-L . . .  This back and forth motion often  tends to lead to a sort of peculiar rhythm of its own, betraying that a  "sideswiper" is being used.

Commercially made double-speed keys were relatively cheap compared to the Vibroplex, and it was quite easy to make a good home-made one.   No wonder that it became popular in wireless operations, especially among hams, for some years.  (It does not seem to have been used by  landline telegraphers.)  Perhaps its novelty was as much an attraction as its claim for higher speeds and lowered fatigue.

Of passing interest in 1926 was a similarly connected key having two pushbuttons, like typewriter keys or pushbuttons, to be used with two fingers, called the "Cricket" by its  manufacturer. The keys were to be used alternately to form the characters, as with the "sideswiper".  It never became popular.

The "BUG"

Historically the Martin semiautomatic key, introduced in 1906 as the "Auto" and later as the "Vibroplex" - commonly called a "bug" - is listed here second because of its greater mechanical complexity and difference in use..  The patented Vibroplex, by making dots automatically (by the sidewise vibration of its elastically mounted arm),  relieved much of the operator's effort (although he still had to form the dashes manually), and increased his speed potential, while reducing the risk of "glass arm" (by sidewise movement and division of labor between thumb and fingers).

In its various models it became very popular and has been widely used up to the present time.   There have been  many imitations, a few of which also produced automatic dashes.  Normal (right handed) models formed the dits automatically with a right-wise movement of the thumb and the dahs manually with a left-wise motion by one or two fingers against a paddle assembly.   A few designs produced by a few manufacturers provided automatic dahs with by a second vibrating arm.

On the Australian land-lines bug-keys were known as "jiggers". Those  issued by the Sydney GPO Telegraph Office in 1946 had 3 knobs, two of them controlled separate swinging arms, one for automatically  forming dits and one for automatic dahs and the third for manually controlled dahs.  The knobs could be positioned at either end of the base-plate for easy use by right or left handed people.  I have no information as to how these were used.


A "bug" should not slip on the table, and its paddles should be about 2-1/2 inches above the tabletop.  Most teachers recommend a light touch, pivoting the hand on the knuckle of little finger and using as combination of finger action and rolling wrist-motion.   (Long-time speed champion Ted McElroy, however, said the wrist and elbow should be off the table, and a full, free swing of the arm used.)  We may suspect several different styles are equally satisfactory.  (It has been suggested that by holding a pencil in the same hand while sending will help one learn to relax.)

Bug sending should duplicate good hand-key sending.  Handle it easily.  Do not grip its paddles, but only allow the fingers or thumb to touch the side you are pressing on -- not touching the other side.  When a bug is used for radio work there is a tendency to make the dits relatively too light.  As compared to telegraph landline sending, radio requires a heavier style to put the signal through static and interference, and a heavier key will help do this.  So be sure to set heavy enough dits that they are not likely to be swallowed up by moderate static or interference.

Setting the Adjustments of a Bug

Like all keys, bug adjustments are a highly personal matter, varying from one operator to another.  They are also sensitive to the range of speed.  For example, a bug set for 35 wpm operation will do poorly at 18, and vice-versa. --  Remember the rule:  NEVER readjust another operator's bug!

Hugh S. Pettis, K3EC, recommends the following as optimum bug settings:

He cites a common technique for setting correct dit duration (a dit equals unit space) is to clip an ohmmeter across the bug terminals.  First, set it for full scale while holding the paddle against the dah contact.  Then adjust the stationary dit post contact until it gives a mid-scale reading for a series of dits, and finally settles on a full scale reading - a closed circuit.  His personal preference is for the dits to taper off to a closed circuit after about ten dits.    More dits will give a lighter weighting, and if the series leaves the circuit open, it is too light.  Fewer dits will produce a heavier weighting, and if it is fewer than eight, the nominal (8 dit) error signal cannot be made.

Robert. R Hall W9CRO recommends:   (Some adjustments are interactive.)



Key smoothly and easily with a minimum of effort.  Let the bug do the work - you just control it, with the arm resting on the table, touching the paddles loosely (lightly) between thumb and forefinger.  Control it without much motion of the hand or fingers.  A slight twist or roll of the wrist will change from the dit to the dah side.  Relax and enjoy it.   Don't  bat out the dits and dahs out with thumb and forefinger so widely separated and so hard that it tends to push bug around.

There is a marked tendency among some bug users to set the dits too fast relative to the hand- formed dahs and spaces. Hand-formed spaces tend to become too long in proportion.  The result is often a choppy sounding code or to signals which are certainly readable, but tiring to listen to and read.  Katashi Nose KH6IJ points out that "at high speed one cannot put much force on the paddles." He also said that "If you move your whole arm, the law of inertia prevents you from attaining high speeds."


Keyers are electronic devices controlled by paddles similar to those on a "bug" for automatically making dits and dahs, and often incorporate other useful operating features, including buffers and memories.  Many include "iambic" type of operation by a "squeezing" motion which provides for alternate dits and dahs, which further automates sending and in this way reduces total effort.  An iambic keyer will always produce perfect characters, even though they may not be used in our code.

Katashi Nose here says "If you have already mastered a bug, it will take about three weeks to convert to electronic-key sending.  Once you are converted, you are hooked because now your bug fist is ruined [ed. for most people]; an entirely different technique is required."    If your keyer has "forced character spacing" (FCS), use it! This may take several weeks practice, but your sending will be real armchair copy.   It is worth the effort.


Finally, the keyboard (including the use of electronic computers with programs for using their keyboards) automatically makes all characters from a typewriter type of keyboard.  Both keyers and keyboards often include teaching programs for learning the code and/or improving code abilities, as well as having memories for various purposes.  This is about the ultimate in code production.  (Machine sent CW is considered almost a "must" for good copy when signals are very faint, including QRP-- and for very high speed work (hand sending just won't hack that.)

Keyboards also have much to offer the beginner in learning the code initially and for improving one's skills.  What may be possible hand-key speeds?

Psychological testing shows the average rates at which people can tap a fingers:-
On the High side:   9.7 per second,  or 576/minute,  (300 in 31 sec)
Average:  8.6 per second,  or 516/min.  (300 in 35 sec.)
On the Low side:   6.7 per second, or  402/min.  (300 in 45 sec.)

If we assume that a dit is one "tap" and a dah is equal to two "taps" (two nerve pulses: one down and one up), then we may say:
Taps Letters Group Frequency Taps x Frequency
1 E 0.130 0.130
2 T I 0.166 0.332
3 A N S 0.214 0.642
4 D H M R U 0.192 0.768
5 B F G K L V W 0.124 0.620
6 C O P X Z 0.139 0.834
7 J Q Y 0.024 0.168
Average per letter 1.000 3.494
Average for five-letter word = 17.47 taps.

At this rate, assuming the above rates can be maintained for periods of time needed to send messages, news, etc., the slowest keying rate would be 23 wpm, the average 30 and the highest 33 wpm.


The Sydney Australia GPO Telegraph Office in 1946 produced a bug that had two separate swinging arms for dots and dashes. There were 3 knobs:  one for dits, one for automatic dahs and one for manually controlled dashes.  The knobs could be positioned at either end of the base-plate for easy use by right or left handed people.

The Art and Skill of Radio-Telegraphy-Second Revised Edition-
©William G. Pierpont N0HFF