Updated Can Türk
Every Morse trainer has a speed control. Most have two, and the difference between them is the most consequential setting you will touch.
The problem with slowing down
The obvious way to make Morse easier is to send it slower. At 5 words per minute a dot lasts 240 milliseconds — long enough to hear as a distinct event, long enough to count.
And that is exactly the problem. At that speed you do not hear -.-. as C. You
hear dah, di, dah, di, and you count four separate things and add them up.
The counting works. It gets you through the alphabet. It also becomes the only
way you know how to decode, and counting has a hard ceiling — somewhere around
10 WPM the beeps arrive faster than you can count them, and everything falls
apart at once.
People who learn this way describe hitting a wall. It is not a plateau. It is the counting strategy running out of road.
What Farnsworth does instead
Named after Russ Farnsworth, the method separates two things that a single speed control conflates:
- Character speed — how fast the dots and dashes inside a character are sent
- Effective speed — how fast the text arrives overall
Farnsworth sends the characters at full speed, say 18 WPM, and pads the silence
between them until the overall rate is whatever the learner needs, say 8 WPM.
The character -.-. sounds exactly as it will at 18 WPM — a single rhythmic
shape, too fast to count — and then there is a long pause before the next one
while you work out what you just heard.
You are given time to think. You are not given time to count.
The arithmetic
Morse is defined in units, where a dot is one unit:
| Element | Units |
|---|---|
| Dot | 1 |
| Dash | 3 |
| Gap inside a character | 1 |
| Gap between characters | 3 |
| Gap between words | 7 |
The reference word PARIS is exactly 50 units long including its trailing word
gap, which is where words-per-minute comes from: at W WPM, one unit lasts
1200 / W milliseconds. The Morse code speed chart works
that out for every speed.
Of those 50 units, 31 are the actual keying and 19 are the gaps. The ARRL method for Farnsworth keeps the 31 at the character speed and expands the 19 until the whole thing takes as long as the slower effective speed requires, distributing the extra time across character and word gaps in their normal 3:7 ratio.
So at 18/8 Farnsworth, each character takes the same time it would at 18 WPM, and the silences do all the stretching.
What to actually set
A reasonable starting point, and what this trainer defaults to:
- Character speed: 15–20 WPM. High enough that counting is not an option. Do not lower this, ever. It is the one number that should not move.
- Effective speed: start at 5–8 WPM. This is the dial you adjust.
Then raise the effective speed as you get comfortable, and leave the character speed alone until the two meet. When you are copying at 18/18 you are copying at 18 WPM with standard timing, and you got there without ever learning to count.
If a session feels impossible, lower the effective speed. If it feels easy, raise it. The character speed is not part of that conversation.
The uncomfortable part
Farnsworth is harder at the start. Counting works, and it works immediately, and giving it up means several sessions where you feel worse than you did the day before.
That trade is the whole method: a slower first week in exchange for not hitting a wall in the second month. It is the same reason learning from a chart feels faster and ends badly — the shortcut is real, and so is the bill.
You can set both speeds independently in settings, and hear the difference immediately in the translator by sending the same word at 18/8 and at 8/8.