Getting from 5 wpm to 20 wpm: what actually moves the needle, and where does everyone stall?
- Replies 2
- Views 14
- Created
- Last Reply
Top Posters In This Topic
-
Admin_Terry 1 post
-
Mod_Chad 1 post
A better way to browse. Learn more.
A full-screen app on your home screen with push notifications, badges and more.
This thread is for comparing methods for building CW speed from beginner level to a comfortable 20 wpm. It covers practice structure, plateaus, gear, and the ear-versus-software question. Below is a rough map of the stages, and I'd like to hear where yours differ.
5 to 10 wpm: the characters have to become sounds. The usual advice is to send characters at a fast rate, around 18 to 20 wpm, with extra space between characters and words (Farnsworth timing), or to use the Koch method, which starts with two characters at full speed and adds one at a time once you copy about 90 percent. Both aim at the same thing: recognizing a character by its rhythm, not by counting dits and dahs. If you catch yourself thinking "dit dah dit dit, that's L," the character speed is too slow.
10 to 15 wpm: the first big plateau. This is where many people get stuck. Pencil-and-paper copy of every letter starts to cost you, because your hand falls behind your ear. Letters keep arriving while you are still writing the last one, and you lose the next two. The fixes are to copy a little behind the sender, to practice common words as single units (THE, AND, FOR, YOU, RIG, ANT, WX), and to start copying in your head for short stretches. Head copy is easier when you stop translating each letter into text and just let words and phrases register.
15 to 20 wpm: words, callsigns and exchanges. Callsign recognition is its own skill because there is no context to help you. A random string like K9ZQX can't be guessed from the letters before it the way a word can. Dedicated callsign trainers, plus a pileup simulator for contest-style exchanges, are worth the time here. Practice the standard QSO skeleton too: CQ, callsign, RST, name, QTH, rig and weather, TNX, 73. Once the structure is familiar you only have to catch the variable parts.
Schedule. Short daily sessions generally beat one long weekly session. Something like 20 to 30 minutes a day could be split as: 10 minutes of new or weak characters, 10 minutes of words or callsigns slightly above your comfort speed, and 5 to 10 minutes of real or recorded QSO material. A common approach to plateaus is to alternate days: one day at a speed where you copy about 95 percent, the next at a speed where you only get 70 to 80 percent. The slower days build accuracy and the faster days stretch your ear.
Sending matters too. If you use paddles, iambic mode A or B, paddle spacing and tension, and weighting all affect how clean your code sounds. Many keyers default to a 3:1 dah-to-dit ratio and standard spacing. Poor spacing between letters and words makes your code hard to copy at any speed. A sidetone pitch you find comfortable (many people like 500 to 700 Hz) reduces fatigue over long sessions.
Ear versus software. Decoders can be useful for checking what you thought you heard, but they struggle with QSB, noise, and hand-sent code with uneven timing. If you lean on one, your brain may stop doing the work. Using it only to verify after you've copied is a reasonable compromise.
Questions for the thread: what was the plateau speed where you stalled, what broke it, and are CW nets or contests useful for building speed, or are they better saved until you're already comfortable?
Link to comment
https://www.hamradiobase.com/forums/topic/11799-getting-from-5-wpm-to-20-wpm-what-actually-moves-the-needle-and-where-does-everyone-stall/Share on other sites