neil,

we've just been talking about a portion of the required code and assumed you'd include addition stuff to control when it began and ended the table lookup.

it sounds like your routine starts, and runs forever not just through your intended table, but past the end and across all memory addresses to FFFFh and then back around to 0000h again, forever.

could you post what you have now and we'll make some suggestions...

;-----

movc a,@a+dptr doesn't do much to the microprocessor resources. it does change dptr; it changes only the accumulator contents, overwriting the contents prior to its execution.

since the accumulator changes by overwriting the contents with the value pointed in memory, two subsequent movc a,@a+dptr instructions would (in most cases - exception not noteworthy) access a different value and ovewrite the accumulator:

;-----
movc a,@a+dptr
movc a,@a+dptr
;-----

in this example, the first movc would load a character from the table, and then the next movc would use THAT CHARACTER as the new index or offset into the table to pick up yet another character or possibly an opcode byte beyond the table.

usually this is a mistaken construct. there are applications where this might be used. quickly sequencing through access tables like binary trees are one situation though a little decision logic is needed to determine which branch to take.

for example, a data structure named a priority queue often uses a structure where a binary branch is either twice the last offest or twice plus one. i'll drop that topic as j.guy has been playing topic police recently. :)

priority queues are great things in micros, but its a topic for gurus and not novices.

duh