SARC Events

SARC Events

SARC Courses
Course Information
Field Day


So You Want To Learn CW II

A Communicator Reprise: Part 2

On 2019-04-12 we published an article on learning to copy CW (Morse Code). At that time we committed to following up with another article on learning to send CW.

Like the first article, this one is based on recent personal experience. The methods and tips that have been working for me may not work for everyone, but they will give a committed learner a place to start, with the knowledge that these methods have worked for at least one other person.

The methods described in this article are ones I used to become comfortable in daily CW QSOs with old friends and a few new friends. These “rag chews” normally last from 30-45 minutes and are at a speed of around 18-20 words per minute (wpm).  
This article will not prepare you to participate in CW traffic nets. Nor will it prepare you to work CW contests without using a computer or CW memory keyer to do the sending. But these are things you can grow into, if you choose, after developing the basic sending skills that are addressed in this article.

What You Need to Know Before You Start

Just as you can’t sing “Auld Lang Syne” if you don’t know the tune, it is essential to know the sound of good CW in order to send clear, readable code. 

Practice at copying good code is thus critical preparation for learning to send. If you have not yet learned to recognize well-formed, well-spaced words in CW, put this article aside until you have had more practice listening.

The minimum useful character speed (not necessarily the same as the average text speed) is 10 words per minute (wpm). (This is my opinion—I have read an expert opinion that learners should start at 25 wpm.) Below this the brain hears too much dead space between the dits and dahs and can’t get a sense of the sound pattern. (Although 5 wpm has been used in the US as a minimum code speed, all authorities I have come across have condemned this speed as being so low as to inhibit progression to higher, practical speeds.)

The average code speed can be effectively reduced (for beginners) by adding extra space between successive words (and, for very beginners, between characters, although this slows recognition of standard words and character groups).
Since CW conveys information using only two lengths of sounds, and all the information is in the sequence of these two lengths of sounds and the spacings between the sounds, timing is everything. The only difference between “5” and “SEE” is the spacing between the dits of the “S” and two “E”s.

There are standards for spacing between dits and dahs in a character, between letters in a word, and between words in a phrase. You can find the details at URL:, but basically the spacing between the dits and dahs in a character is the length of a dit, the spacing between two characters is the length of a dah, and the spacing between two words is the length of the character “H”.

Getting Ready: Decisions, Decisions

Before you can begin you have several decisions to make.

First, you will have to make a choice of sending apparatus, commonly called a “key”. Actually you have several choices, including the classic straight key (“pump”), the bug, dual paddles and keyer, and the single paddle and keyer. 

I’ve seen a variety of odd-ball sending devices, including the up/down microphone buttons on my IC-7000, the buttons on a recycled two-button mouse, a piece of bent metal and a thumbtack, etc. I strongly recommend purchasing a good quality commercial produced paddle to start, however. You can find some advice at URL:,  but be warned that the highly praised Begali paddles are very expensive. I don’t recommend you start there (unless you want to sell it to me very cheaply if you decide to give up on CW, in which case I hope you buy the Begali Adventure with the mounting bracket for the Elecraft KX3 J). I started with a basic black body Bencher BY-1, and still use that for my base station.

If you choose a straight key, you will be unnecessarily limiting your speed, and a future transition to a paddle will be difficult since the hand and finger action is very different. 
A bug, on the other hand, is an invitation to malformed characters (and I’ve heard some doozies on the air). With a bug, the dahs are sent at manual speed, and the dits are sent at automatic speed. There is not necessarily any correlation between the two speeds, or even between the lengths of successive dahs. 

I recommend using a paddle and keyer. The keyer helps ensure characters are correctly formed.  A keyer used with paddles solves the “bad timing” problem—often heard from bug-users—by generating sequences of correctly timed and spaced dits and dahs. With a keyer, holding the dah paddle too long will generate more dahs, not a long dah (which, in CW-speak, represents the “cut” number for zero), as happens with a bug.

A decision you’ll have to make with the paddles and keyer option is whether you want a single or dual-lever unit. 

With dual paddles you can use Iambic keying. Iambic keying generates alternating dit-dah patterns when the two paddles are pressed at the same time. It requires good thumb-finger coordination and fine muscle control. I am convinced it is more mentally demanding that non-Iambic, since the brain needs to consider the particular pattern of muscle movements required for each character, and getting set up for these movements between characters can be challenging (it is for me when I push my speed up). However, there is a small efficiency advantage in using Iambic keying. 

I use Iambic keying, although I have read that high-speed CW operators commonly prefer single paddle keying, which is necessarily non-Iambic. You can find much more on Iambic keying in an opinion piece, from a negative but useful perspective, at URL:   The explanations are very good, and some of the negative points are good ones. (One particularly pointed argument is the following: “In practice, anybody who can send at 5 wpm with a paddle can “squeeze key” effectively. At 20wpm it takes a lot of practice and some people just can’t do it. Above 40wpm the more complicated squeezes are forgotten about even by operators who “squeeze” everything at slower speeds.”)

In the rest of this article, I will assume you are doing Iambic keying with dual paddles and a keyer. (If you choose not to use Iambic keying, however, I will forgive you: just skip over those bits.)

Once you’ve selected and acquired your new paddles (and I’m going to assume you have selected paddles from here on), you will need to adjust the spacings and tension. You should start with the manufacturer’s recommendations, but you can also find very good information at URL:  However you set things up initially, you will likely be fine tuning those adjustments for some time until you become comfortable with them, so be sure to learn the correct procedure for your unit.
Next you will need something to generate the sounds you need for code practice. You can find suggestions in the article by N7RR referenced earlier, or a “build-it-yourself” design at URL: (bottom link).  I purchased a PicoKeyer Plus at first. You can also get (as I did later) the widely supported K1EL Winkeyer WKUSB-SMT, which, afterwards, can be very useful for getting your rig to send CW from your computer (very useful for CW contesting). Both are easy to build kits for those who like to solder. 

If you have a modern HF rig, this will likely be your least (additional) cost option. Modern HF rigs generally can act as a good code practice oscillator if the QSK (sometimes called VOX) setting is turned off, so the transmitter does not generate a carrier automatically when the key or paddle is pressed. (When setting this up, test it by connecting the transmitter to a dummy load, and turning the power down to the lowest setting available.) Keep in mind, however, that you may not have as much flexibility in Iambic keying options as with the specialty keyers.

A modern specialty keyer can be set in one of at least three common modes: Iambic A, Iambic B, and Ultimatic. With both Iambic A and Iambic B keying, pressing both paddles simultaneously causes the keyer to output a sequence of alternating dits and dahs. The technical differences are described in an authoritative article (by John Curtis, President of Curtis Electro Devices) that can be found at URL: A is easier to learn, since, when the paddles are released, the keyer completes the current dit or dah element and then stops.  With Iambic B, if the paddles are released during generation of an element, the keyer will follow whichever element it is currently sending with the alternate (dit or dah) element, so the timing of the paddle action is more critical and must be much more precise (which can be tough for beginners). 

Ultimatic keying is less commonly supported (except by specialty keying chips, such as that from K1EL). With Ultimatic keying the last paddle pressed determines whether the dit or dah is repeated. If the dah paddle is pressed last, for example, the current dit will be completed and then the keyer will generate a recurring series of dahs.

My ICOM IC-7000 supports only Iambic B, so that was how I started and continue today. However, if you have a choice, and choose to use Iambic keying, Iambic A may be easier to learn.

A final decision you will have to make is which hand to use to send. Although I am right handed, I send with my left, which leaves my right hand free to make notes. This is useful when doing fairly fast contacts that are being logged on paper, such as for Summits on the Air (SOTA) activations. It is generally easier, however, to use the dominant hand to send with, since muscle coordination will be better developed and muscle control will be more precise. Whichever choice you make you will likely have to live with, although I know at least one CW op who can send well with either hand.

If you decide to send with your left hand, I suggest you learn, as I did, with the paddle configured for a right-hander, where the left paddle sends dits. The advantage is that you can be a guest op in a right-hander’s shack without having to reconfigure their rig’s paddle settings (a step which is likely to be forgotten and so will confuse one or both of you).

Sending Practice

We’re finally here: you now get to send CW!  You have done your preparations, and obtained, assembled and tested the necessary equipment.

In the beginning you should concentrate on sending one letter at a time, with lots of spacing between letters. Start sending a few standard things, such as your callsign, and those old typing standbys: “The quick brown fox jumps over the lazy dog”, and “Now is the time for all good men to come to the aid of the party”. “My Fair Lady” fans might even try “The rain in Spain falls mainly in the plain.” But these quickly become boring, so you’ll need other material.

I’ve found simple books (hint: this is not the time to pick up that book you were reading on Maxwell’s equations, Laplace transforms, or string theory) and newspaper articles can be useful.

There are four basic steps in sending: 
  1. What to say (the idea to be conveyed);
  2. How to say it (the choice of words and syntax);
  3. How to spell the words; and
  4. How to send the letters in the words. 
Note that the first two steps are common to speech, and so will be familiar. The third step is common to writing and typing, but can be a stumbling block for CW (more on this to come). The fourth step is the one you want to practice first, and using a book as source material will allow you to focus on this step. This step is where you will develop the brain-finger muscle coordination that you’ll need to “talk” in CW, just as, when very young, you learned to coordinate your brain with your tongue, lips, and larynx to speak.
One tip: when sending from books (or other written material), don’t bother with the commas, periods, dashes, and oddities like exclamation marks and quotation marks. You hardly ever need these in real life. 

In real QSOs you can manage very well by abandoning all but a few punctuation marks and pro-signs. Especially useful are the forward stroke used for portable operation (“/“), the question mark (“?”), the pause (<BT>), “73”, the “go ahead” K, the “go ahead only the one station I’m calling” KN, and the “that’s it from me for this QSO” <SK>. 
Finally, you will make mistakes. (If you think you aren’t making mistakes then you aren’t listening closely enough to what you are sending.) 

The formal way of correcting an error is to stop, send a string of eight dits, and then re-start the word in question. In practice the number of dits doesn’t need to be exactly eight, as long as it is clearly more than five (code for the numeral five) and isn’t absurdly large. This is the method I generally use.

Three other methods you might encounter are:
  1. A short series of spaced dits (e.g. dit, space, dit) followed by the corrected word.
  2. A question mark, followed by the corrected word. (Also used to repeat a word where there may have been confusion, such as with an abbreviation such as “SARC”.)
  3. For errors which are obvious (e.g. clear mis-spellings such as “tommorow”), just ignore the error and keep moving, which respects the fact that the operator at the other end can figure out what was intended.
So, with your paddles, keyer, code-practice oscillator set up, and your book in front of you (with adequate lighting), start at Page 1 and work your way through 2-3 pages per session. Keep it up—you are training your brain to make your fingers do the talking.

One final tip: Unless you live alone, you may want to use headphones. (Otherwise in a couple of weeks you may find that you are living alone.)

Keep practicing...

No comments:

Post a Comment


The July-August 2024 SARC Communicator

Hello summer... With another big Summer issue. The July-August 2024 Communicator, digital periodical of Surrey Amateur Radio Communications ...

The Most Viewed...