The classical language allowed a second suffix, a da written after a final consonant that had already closed the syllable. It is called the da drag, the strong da. It was never pronounced in the way an initial da is, and in the later orthography it was dropped from ordinary spelling altogether.
It did not vanish. It survives in canonical texts, in archaic and imperial-period documents, and in manuscripts copied from old exemplars by scribes who reproduced what was in front of them. So the presence of a da drag in a text is a signal, and a reader who notices it learns something about where the text has come from before reading a word of its content.
It is one of several such signals. Old Tibetan orthography has a small set of features that later spelling gave up: the reversed form of the vowel sign i, certain subscript combinations, particular spellings of common particles, and the da drag among them. Together they let a manuscript be placed in a rough relation to the great standardisations of the language, the revisions of vocabulary and spelling undertaken to regularise Buddhist translation.
None of this is decorative. Tibetan spelling was standardised deliberately, by decision and at a known period, and the archaisms are what the standardisation left behind. A scribe reproducing them is transmitting a text older than the rules being taught in his own day.