Split character string multiple times every two characters

You were actually quite close. You need to specify the separator-positions as sep = c(2,4) instead of sep = c(2,2):

df <- separate(df, a, c(paste0("V",LETTERS[1:3])), sep = c(2,4))

you get:

> df
  VA VB VC
1 Aa Bb CC
2 AA BB CC
3 AA bb CC

---

In base R you could do (borrowing from @rawr's comment):

l <- ave(as.character(df$a), FUN = function(x) strsplit(x, '(?<=..)', perl = TRUE))
df <- data.frame(do.call('rbind', l))

which gives:

> df
  X1 X2 X3
1 Aa Bb CC
2 AA BB CC
3 AA bb CC

We could do this with base R

read.csv(text=gsub('(..)(?!$)', '\\1,', df$a, 
    perl=TRUE),col.names=paste0("V", LETTERS[1:3]), header=FALSE)
#  VA VB VC
#1 Aa Bb CC
#2 AA BB CC
#3 AA bb CC

If we are reading directly from the file, another option is read.fwf

read.fwf(file="yourfile.txt", widths=c(2,2,2), skip=1)