I will try to answer your question but (unfortunately) I have no conclusive evidence to support what I have written below other than what I've observed. It's just pure conjecture from a little bit of snooping around the net.
First off, you cannot actually visually distinguish (eyeball) an engraved note from a lithographed example. In Canada, we had the change occur on our Ones & Twos around the same time as when this Indonesian 100 Rupiah was issued (1984). They actually tested the lithographed backs (a cheaper printing method) on test notes (but I digress). They changed printing techniques in mid prefix (so we know that up to ___ # notes had engraved backs & after ___# they were lithographed for prefixes AFF, EAK). The only way one can actually know is if this type of data which was passed on to collectors. The Bank of Canada provided the change-over serial number data to Charlton (our catalogue who published it) & that's why we know & I'm unsure how helpful my best guess can be. I cannot find any change-over point on either the SCWPM, the banknote museum nor any other site at the moment.
From checking with the banknote museum
P-122a is likely prefix AAA to HAA (assuming there's no IAA) but the change-over could be before this. If you check out their example, P-122b prefix is KBQ & I found this
P-122b listed (deduced) by PMG on eBay. It happens to have prefix JAD & I believe that PMG would exercised due diligence to ascertain the exact change-over point. Unfortunately, there's more eBay sellers who have listed their P-122's (100 Rupiahs) as the first (a) version incorrectly which certainly doesn't help clarify where the change-over point is. If there's somebody from Indonesia, who has a catalogue or better info, it would help define that change-over point.