There are two main uses of the concept: one that uses the distinct typeface, whose characters are all minuscule, and one that attempts to use both minuscule and capital letters of the existing typewriter keyboard or the basic ASCII character set to take the place of the distinct characters.
The special typeface does have some letters used as capitals. These are simply of the same form as the minuscules, only expanded in size so as to extend beyond x-height both above the mean-line and below the base-line.
The special typeface has ligatures in the form of digraphs for each of the long vowels, and for "wh", "ng," "sh", and "ch". There are two distinct digraph ligatures for the voiced and unvoiced "th". There is a variant of the "r" to end syllables, which is essentially silent in received pronunciation but with a special sound in General American, Scots English and some other British regional accents. Each of the letters has a name the pronunciation of which includes the sound that the character stands for. For example, there is a backwards "z" to replace the "s" where it is voiced. The name of this letter is "zess".
There was later a little-known modification of the character set to accommodate local ways of pronouncing English, a form of "disambiguation". In the original set, a "hook a" or "two-storey a" was used for the sound in "cat", and a "round a" or "one-storey a" for the sound in "father". But Americans and Canadians generally do not pronounce the words "rather", "dance", and "half" with this latter sound, so a 45th character, the "half-hook a", was devised that could be taken either way by the appropriate linguistic community.