PICASM- That's just the bloody stupid thing that I'll have forgotten. I wouldn't be at all surprised-- I'll go do it now!
SSYLEE- The program in question was written to test a new board with an 887. It checked out with all the other ports used as digital outputs, but failed with a6 floating and a7 low.When it failed, the fusebits were checked first. A new board was fitted with just the processor , in order to eliminate any loading effects, and tried again. It was at this point that it was noticed that the processor supply current was pulsing to 90 milliamps, with nowhere external to go.
I supposed that the chip might be faulty, and tried 2 more (from the same batch) with the same results. I then programmed an 886, with the same program, and it worked.
SO WHAT NOW
Just tried the decoupling- no help. It was a good idea,though.