A New DoubleBug

This is another in the series of fully-automatic keys that have been developed at WB9LPU.  It is a work in progress, and this topic will be updated as more work is done.

     - Click here for First Update (2/8/2009) 

        - Click here for May Update - Contest Results  (5/18/2009)                               

Parkwood DoubleBug Number Eight



    This is yet another attempt at making a fully-automatic telegraph key.  It is the eighth one that has had satisfactory performance, but the goals were a bit different this time.  Primarily, the intent was to make a design that would be "bulletproof" in operation and simple enough in construction that it could be made in reasonable numbers in a reasonable amount of time (assuming a sufficiently flexible definition of "reasonable").  It uses a number of concepts that have been developed for other Parkwood bugs over the years and combines them into a small instrument whose characteristics are now being evaluated.

     Here are some of the design features that were incorporated into the prototype shown above -

     - Magnetic control of pendulum motion and paddle tension

     - Single-lever design to eliminate conflict between dot and dash operation

     - Dual pendulums to prevent momentum transfer during dot-to-dash transitions

     - Magnetic sensing of pendulum motion

     - Micrometer adjustment of dot and dash duty cycle (weight)

     - Both magnetic and sliding-weight control of dot and dash speed

Basic Design and Construction

     The prototype is built on a base of Fortal aluminum (a very hard and high-strength aluminum alloy), with other parts made from 2024-T4 aluminum and C360 brass.  [Note: these are high-resolution photos - left-clicking on them will let you zoom in on details.]  Screws for adjustment and construction were of stainless steel.  The view below shows the general layout.  A single lever (with dual finger-

 pieces for a comfortable feel) is pivoted on ball bearings.  The far end of the lever bears two rare-earth magnets that are  mounted in threaded carriers.  A set of magnets in the lever and the rear of the top plate aids in centering the pendulum, which is naturally held near center when at rest because of symmetrical design of the driving mechanism.  There are separate dot and dash pendulums, each pivoted on precision ball bearings.  Each pendulum carries its own (adjustable) magnet that is opposed to the the corresponding magnet on the paddle lever.  When the paddle is moved to either side, the appropriate pendulum "follows" the paddle magnet and oscillates around the new position of the paddle magnet.  The rate of oscillation is determined by the degree of attraction between the magnets and the position of the brass weight along the pendulum.  I have coined the term "magnet follower" for this type of mechanism.  Movement of each pendulum, which bears another magnet at its tip, is sensed by a magnetic reed switch contained in an upright brass tube near the tip.  The relative duration (on-time) of the dot or dash is determined by the position of the reed switch relative to the pendulum swing.  This position is adjusted by a screw-thread mechanism built into the base of the bug.  A closer look at this arrangement is show below.


     Using separate pendulums whose rest position is set by separate stops prevents transfer from one pendulum to the other during dot-to-dash (or v.v.) transitions.  A drawback to the two-pendulum design is that two changes in weight position must be made to change the overall speed.  Experiments are currently being made to determine the sizes of the pendulum weights that will allow them to track together  for speed changes.  Since the system contains numerous non-linearities, this may be difficult.

Final Comments

     Testing of this instrument is in its early stages, and it has not been put on the air as of this writing.  At present the finish is a bit rough, but it will be cleaned up when the testing is done.  When it is "shaken down" sufficiently on the air and on the bench, a final design (which may use a brass base for added weight) will be worked out.  The initial setup of the instrument is a bit involved, but it can be approached rationally by an understanding of the principles of operation, and it should be no harder to get working well than a conventional bug.  For further information on the design features used in this prototype, take a look at some of the articles on my related web site.   

     The last photo is a size comparison with an instrument that we all know and love.  Yes, size does matter, especially if your operating desk is always as cluttered as mine seems to be.


     73 de Rich, WB9LPU

     wb9lpu@gmail.com or wb9lpu@earthlink.net