W5ALT CW Info
Is CW Dead
What is CW?
Is CW Digital?
Why CW Effciency?
Slowing Down a Bug
Bunnell Double Speed
TAC Hole-In-The-Wall Bug
Lionel J-36 Bug
What is CW?
by Walt Fair, Jr., W5ALT
Everyone seems to already know that Morse code or CW is the funny beeping noise that comes from a telegraph key. Well, it's actually a little more complicated, and hopefully more interesting, than that! In this section we'll describe some of the technical aspects of Morse code and CW operating. Hopefully we'll show that there is actually some technically interesting things associated with CW.
It's All in the Timing
The simplest description of CW is turning on and off a radio signal to make long and short pulses. While that is pretty accurate, there are a few more details that need to be taken care of in order to actually send and receive CW.
First, Morse code consists of long and short pulses - everyone knows that. CW operators call them "dits" and "dahs" not "dots" and "dashes", since "dit" and "dah" more closely resemble the timing of the sounds. So if you discuss CW, please don't say "dots and dashes" - OK? But equally important is that the dits and dahs are separated by spaces. It is the relative length of the pulses and spaces that really define Morse code.
It is interesting that the human ear and brain are able to tell the difference between long and short times, but only up to a limit. In some experiments that I did back in the 1980's, I found that my ear/brain combination had a hard time telling dits from dahs when the ratio of the sounds was much less than 2. In other words, if the dahs were less than twice as long as the dits, my brain couldn't easily and quickly figure out whether any given sound was a dit or a dah. On the other hand, making the dahs longer than about 3 times the length of a dit didn't help hardly at all in differentiating the sounds.
Likewise for the spaces between elements (dits and dahs). If the space between characters is less than about 2 times the space between dits and dahs, my brain has a hard time rapidly figuring out whether the space represents the end of a character or simply a space between dits and dahs.
For that reason, standard Morse code is defined to use dahs that are 3 times longer than the dits. In addition, don't forget about the spaces. The space between dits and dahs is equal to the length of a dit and the space between characters is equal to the length of a dah, or 3 dits. There is one more important timing to consider - the space between words. In general that is considered to ideally be twice the space between letters, but if it's a little longer, there's no confusion. If you stick with that timing, you can send perfect CW.
In CW circles, people refer to the sound of an operator sending CW to be his "fist". In practice most people do have some distinctive characteristics to the way they send CW. The differences are due to small differences in timing and to an experienced CW operator sound much like an accent that someone has when speaking a language. Some spoken accents are downright irritating, others sound somewhat funny, others show a lack of education or that the person comes from a certain part of the world. But in general, for communicating, it is known that things work best if people speak with a common accent. That's why for American English the so-called "mid west accent" is accepted as standard. In fact since most all radio and TV announcers are trained to speak with a "mid west accent" when we hear it we might say they have no accent at all and are easy to understand. In contrast a Texas or Bronx accent is easily recognizable.
So if everyone uses exactly the same CW spacing, won't that make everyone sound the same? Yes! And that is the idea - to communicate effectively and be easily understood. There is no pride in having a distinctive fist! The highest compliment you can pay to a true CW operator is to ask them if they are using a machine to send CW when they are using a manual key.
Words and Minutes
When trying to gage code speed, hams and telegraphers usually use the term "words per minute" (WPM). Now we all know what a minute is, but there are words like "I" and there are words like "supercalifragilisticexpialidocious" so what do we mean by words per minute?
Well someone at some time decided that the average word has 5 letters, so if you want to know how many words are sent in CW, you count the number of letters and divide by 5. Well, almost. It turns out that the numbers and other procedural symbols are longer than most letters, so the numbers and procedural symbols count for 2 letters. Therefore my callsign when I operate from Venezuela is YV1/W5ALT, which contains 6 letters, 2 numbers and the / symbol, so that makes a total of 6 + 2*3 = 12 letters or 2.4 words. By the way, I'm sure you're dying to know, "supercalifragilisticexpialidocious" has 34 letters, so it counts as 6.8 words.
But, although that works on the average, it still isn't a very precise way to figure actual code speed, because some letters have lots of dahs and others have only a dit or two. So although the 5 letters per word work on the average, it doesn't work on most any real word you can think of. There is at least 1 word, though, that is exactly as long as the average: the word "PARIS". So, if you want to know how fast you are sending, send the word "PARIS" over and over and count the number of times you send it in a minute. That is precisely your CW speed in words per minute.
So, I'll bet by now you've thought to yourself "Holy moly, at 5 WPM it would take 20 minutes to send a 100 word message describing what a nice time I had over the weekend! That's not very useful!"
Well guess what? You'd be right. 5 WPM is agonizingly slow. Most hams consider 15 - 20 WPM to be a reasonable conversational speed and some operators cruise along in excess of 50 WPM. But also, due to the slowness, CW operators use a lot of abbreviations to get the message across in fewer letters. For example, some common abbreviations are ES for "and", FB (i.e. fine business) for "great" or "fantastic", QTH for "location", GE for "good evening", etc. To be an effective CW operator you not only need to know Morse code, but also learn the common abbreviations used in CW communications.
It is apparent that CW is a digital mode, since the only thing needed is an on-off switch to send code. Since the length of the dits and dahs are not the same, it is not a typical binary code (more on that later!), but since it uses only discreet states to encode information, it definitely is a digital code. (OK, if you want to get picky, it is a form amplitude shift keying using pulse width modulation. Satisfied?) So, if CW is a digital code, we should know what the data rate is in terms of baud, right?
Well, it turns out that's not too hard to figure. Using the word PARIS, it can be seen that there are a total of 49 timing units needed to send the word and the space following it, although some people like to round that to 50. To figure that out, we define 1 unit as the length of a dit. Then the length of a dah is 3 units, the space between dits and dahs is 1 unit, the space between characters is 3 units and the space at the end of the word is 6 units. If you count them out and add the units all up, there are 49, but you can round that to 50 if you want.
So that means that if you send CW at a speed of W words per minute, you are sending (50 W) symbols per minute or (50 W/60) = (0.8333 W) symbols per second. It also means that the time required for a dit - 1 unit - is (60/50 W) = 1.2/W seconds. And since the definition of baud is the frequency of the smallest meaningful unit of data, the CW baud rate is 0.8333 W baud. So when you send CW at 5 WPM, you are manually generating a 4.17 baud signal, while at 20 WPM you're up to 16.67 baud. For comparison, PSK31 is about 31 baud, RTTY is normally 50 baud, and other digital modes get up to the hundreds or thousand of baud.
Confused? Don't worry too much about it. Unless you are just curious, want to use that as a trivia question for a cocktail party, or are writing digital software to handle CW, you really don't need to know. Our ears and brain can learn to copy CW even if we don't know what the equivalent baud rate is!
And What About Farnsworth?
It turns out that when people hear CW at various speeds, the letters seem to change rhythm as the speed increases. Many people find that there is a learning plateau at around 10 WPM or so. The reason has to do with the ear-brain perception, but we won't go into that. Essentially below about 10 WPM people hear individual dits and dahs, but above that speed they seem to hear complete character sounds instead. That's pretty much what I experienced in learning CW - above 10 WPM I essentially had to relearn the entire alphabet.
So, to ease the burden of having to start over and relearn all the sounds when you get to around 10 WPM, the Farnsworth method was developed for sending code. Basically if the speed is above some rate, say 15 WPM, CW is sent with its normal spacing depending on the speed. But below that rate, 15 WPM in the example, the individual letters are sent at 15 WPM but the spaces between letters and words is lengthened so that the average number of words sent per minute is still a lower speed.
Most people find it easier to learn once at 15 WPM and if they want to send slower, just add more space between characters. So now you know what Farnsworth spacing is all about.