CW...well where do I begin. First of all, I must admit that while I enjoy CW, it is not my favorite mode, at least at the present time.

There are some VERY nice aspects about the mode in comparison to other modes. Some of the advantages are:

I also have the opinion that most, if not all, DXpeditions use CW. They use it for some of the advantages listed above. I have also made some 160M CW contacts with relative ease that I feel would not have been possible with SSB.

It is my opinion that it is easier to achieve more DX using CW, with all other factors considered equal.

If 2 hams were to start off with no DX contacts, and 1 ham were to use strictly SSB, and the other only CW, and were to maintain equal status for the following criteria:

time spent on the radio, RF power, antenna, etc.....

I would bet that the ham who had used CW would achieve his DXCC earlier. Why... more DX stations use CW, less power is needed, and easier to filter out noise.

I had gathered a club's logs from members operating in various bands, and both in CW and SSB, for a contest called "Freeze Your Butt Off". (Don't get me started on the Arizona SQoRPions...). We operated QRP. On each band, stations were contacted in the same general geographical area. On SSB, a few stations we contacted operated with 100W, most were above 500W, 1 was QRP himself; now on CW, about 1 or 2 stations were operating with 100W, most were below 50W. So.... same geographical area (per each respective band), most SSB stations running power, and most CW ops were running below 50W. Hmmm..... I do not know what each station's signal strengths were, but I do know that a good contact was made. For the CW operators this was probably also with keeping in mind that the minimum amount of power needed to establish contact was used.

CW operation has many abbreviations in actual usage. For example instead of saying "My name is Dave", many CW operators would send "OP DAVE"

A fist is an operator's on-air style of sending. It is similar to his "radio fingerprint", or better yet, his CW "voice". With experience, many operators can tell who is sending even before he identifies himself. They can do this by recognizing the way he sends his characters.

Cut numbers are used primarily in contesting or in DX QSO's. It is a method to send shorter numerical characters as a substitution. Commonly in CW contesting a RST report of 599 is sent. Instead of sending the number 9, which is "----.", you can "abbreviate" it as the letter N, "-.". It is then generally understood to indicate the number 9. Remember all numbers in CW have 5 elements in each character, letters have 1 to 4 elements. The cut numbers are derived by their beginning character elements of their respective #'s (for #'s 0 through 5); (and the ending character elements for #'s 6 through 9); and "transposing" them to their "equivalent" letter. The # "0" in Morse code is dah-dah-dah-dah-dah, and to abbreviate it, we take the 1st character element, which is "dah", and the "equivalent" letter is "T". The # "1" is dit-dah-dah-dah-dah; so to abbreviate it, it becomes "dit-dah", for the "equivalent" letter of "A".

Another of the cut numbers that is commonly used is "T" instead of "0"

Check for variation on a difference in the above cut numbers table ( I had seen another variation online)

Zero beating:

Definition

Zero beating a CW signal is basically matching the audio frequency of your signal with the audio frequency of the signal of your intended contact station. This is the same as matching your RF signals. It helps to ensure that you are within the passband of the filtering of the other station so that he can hear your signal. Most hams prefer to copy CW within the vicinity of the audio frequencies of 600 to 800 Hz. I myself prefer 600 Hz. This is a matter of personal preference since your ears will be the ones hearing during a contact. Also be aware that this is a matter of personal preference, 2 operators may like to hear 2 different tones. I at times have zero-beated a CW signal before I called a station who was calling CW, only to have him in turn zero-beat me, which threw me off since he moved his frequency.

How to zero beat a signal

Your radio most likely has a way to adjust your CW sidetone; in many cases a knob or a menu setting. Place your radio in CW mode with any filters set wide. In a non break-in (semi or full) mode, send a long dash. (The idea here is not to transmit a signal yet, but to simply hear a sidetone.) Adjust your CW sidetone setting to what you would prefer to copy from another station. Now tune around until you hear a reasonably strong CW station. (Don't become confused by another station sending which will most likely have a slightly different tone; perhaps you can do this with a W1AW CW bulletin). Now again, with your radio in a non break-in (QSK, either semi or full; you don't want to transmit an actual signal yet) mode, send a long dash with a straight key, or a series of dahs with a bug or paddles. Listen to your tone and the station you are intending to zero beat. Adjust your VFO in small increments until your sidetone and the tone of the other station sound the same. It will start to oscillate. Now you are zero-beated, switch your rig over to a break-in mode, or into a transmit mode and you can send to them on a frequency near to them.

I would fully imagine that there are some experienced CW ops who can dial in on their VFO, listen to the tone of the station, and know that they are zero-beated with them, or relatively close.

I have a little different method of zero-beating. I have a Timewave DSP-599 audio DSP filter. When I wish to zero-beat another station, I depress the Tone button and a steady 600 Hz is heard in my headphones. I turn my VFO tuning knob until I hear the tones are the same and then I can send. Usually after I have done this I turn off the tone on the Timewave DSP-599. The Timewave DSP-599 can easily have the sidetone adjusted from 350 Hz up to 1800 Hz. 

A friend of mine Walt WB8E has another way he uses at times. He has a Radio Shack DSP speaker that he has had modified so that the LED lights constantly as he approaches the proper zero-beated signal with his radio. Another friend added a simple circuit to the RS DSP speaker. (BTW, I've picked up a few of them for $5, but seen them as high as $35. They work for crap on SSB, but work very well for CW)

I know some other hams use computer software to decode CW and zero-beat by setting the intended signal at the proper space in their waterfall, much as you would do for PSK31.

In my personal opinion, I think that there are some cases where you may not necessarily want to be exactly zero-beated with another station. One time I feel that this may be helpful, would be in the case of a DX station who is not working split. You may want to be a little off frequency of other stations calling him. This is so that your signal may stand out a little more uniquely than others. Don't go too far off of frequency or you may be outside of the DX stations receive passband and he will never hear you. I guess that amount all depends on how narrow or wide his filter is.

Types of keys/paddles: straight key, bug, iambic paddles, "sideswiper", "vertical key"

History of Morse Code

History of Samuel F.B. Morse (pdf entitled morse.pdf)

Types of codes: American Morse (used by railroads), International Morse (used by hams and once adopted by the maritime trade [ships at sea], and Navy Morse) During the time of the sinking of the Titanic, all 3 were in use on the high seas

Farnsworth vs. Koch methods

CW & QRP clubs

The word PARIS is the standard to determine CW code speed. Each dit is 1 element and each dah is 3 elements. Between each character the space is 1 element. The spacing between each character in a word is 3 elements; and the spacing between each word is 7 elements.

The word PARIS is exactly 50 elements. Note that after each dit/dah of the letter P -- one element spacing is used except the last one. (Intra-Character).
After the last dit of P is sent, 3 elements are added (Inter-Character). After the word PARIS - 7 elements are used.
Thus:
P
di da da di
1 1 3 1 3 1 1 (3) = 14 elements 
A
di da
1 1 3 (3) = 8 elements 
R
di da di
1 1 3 1 1 (3) = 10 elements 
I
di di
1 1 1 (3) = 6 elements 
S
di di di
1 1 1 1 1 [7] = 12 elements 
Total = 50 elements 
() = intercharacter
[] = interword

If you send PARIS 5 times in a minute (5 WPM) you have sent 250 elements (using correct spacing). 250 elements into 60 seconds per minute = 240 milliseconds per element.

13 words-per-minute is one element every 92.31 milliseconds.

The Farnsworth method sends the dits and dahs and intra-character spacing at a higher speed, then increasing the inter-character and inter-word spacing to slow the sending speed down to the overall speed. For example, to send at 5 wpm with 13 wpm characters in Farnsworth method, the dits and intra-character spacing would be 92.3 milliseconds, the dah would be 276.9 milliseconds, the inter-character spacing would be 1.443 seconds and inter-word spacing would be 3.367 seconds.

A dah (or dash) is 3 times as long as a dit. There is also a proper amount of spacing of time between characters and words. There is 1 "dit" space between elements in each character; there are 3 "dit" spaces between characters; there are 7 "dit" spaces between words. .......................(Provide illustration of dit es dah lengths, as well as explanation of PARIS.)

The word PARIS is used as the standard to define a word and code speed given as Words Per Minute (WPM). PARIS takes 50 "dit" spaces of time.

The letter "P" is:

The letter "A" is:

The letter "R" is:

The letter "I" is:

The letter "S" is:

If each dit space equaled 1 second of time, the word "PARIS" would take 50 seconds to send. This is just over 1 WPM of code speed (1.2 WPM to be more precise)...pretty slow.

Now if you were to shorten the time of dit spaces, to say for example, 5 dit spaces per second, the word "PARIS" would now take 10 seconds to send. You would be able to send the word "PARIS" 6 times in 1 minute for...you guessed it, a 6 WPM code speed. An overall 5 WPM code speed would have the dit spacing equal to 1.2 seconds.

Decrease the dit space more so that, if 10 dit spaces were equal to 1 second, the word "PARIS" would be able to be sent in 5 seconds. Now you could send the word "PARIS" 20 times in 1 minute. 60 seconds in 1 minute divided by 5 seconds per word equals 20 WPM.

Finally, for an overall 30 WPM code speed, every 25 dit spaces would equal 1 second; the word "PARIS" would be able to be sent in 2 seconds. If the word "PARIS" is able to be sent in 2 seconds, you can then send it 30 times in 1 minute.

30 WPM overall code speed is quite commonly heard. Some hams who have acquired the skills and lots of practice can send and receive at much higher rates. Generally speaking, any higher than 30 WPM is an exception to the "rule". Those are few and far in between.

If a dit takes 2 dit spaces (one dit, and an empty dit space), and you were sending a string of 125 dits in 1 minute you would be sending @ 5 WPM. (50 dit spaces per word multiplied by 5 WPM equals 250 dit spaces; half of these are the actual dits being sent, the other half are the empty dit spaces; therefore 125 dits being sent)

125 dits/minute = 5 WPM

325 dits/minute = 13 WPM

500 dits/minute = 20 WPM

750 dits/minute = 30 WPM

CW Bandwidth = wpm X 4 (e.g., 40 WPM = 160 Hz)
From the ARRL License Manual 1976:
"With proper shaping, the necessary keying bandwidth is equal to 4 times the speed in words per minute for International Morse Code; e.g. at 25 words per minute, the bandwidth is approximately 100 cycles [Hz]."

Speed & bandwidth

I would also like to say in "defense" of SSB, I feel that sometimes it is nicer to hear a human voice during a contact. A little more personal or "human".