--_002_3ef6s9xc27ln2skaterswarbrick_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Spehro writes: >>> You might want to look at the states of all the relevant registers on y= our >>> debugger. >>=20 >> Yep, they come up as the datasheet claims (of course). But the debugger >> also thinks everything should work... > > Rather than MPSIM, can you use a hardware debugger? Otherwise you're > working blind. Of course there is always the possibility it works only > when debugging (usually some kind of initialization issue), I've been a bit busy at work, but finally found some time to spend on this tonight. I realised that the Pickit3 does actually do debugging, so played around with it a bit, only swearing slightly at the horrific Java monstrosity that is MPLAB-X. Basically, it seems that the problem was enabling low-voltage programming. I still don't understand why this interacted so strangely, but disabling it seems to have fixed things. Grr. Thank you everyone for the really helpful responses! Rupert --_002_3ef6s9xc27ln2skaterswarbrick_ Content-Type: text/plain; name="ATT00001.txt" Content-Description: ATT00001.txt Content-Disposition: attachment; filename="ATT00001.txt"; size=208; creation-date="Fri, 11 Jan 2013 22:22:09 GMT"; modification-date="Fri, 11 Jan 2013 22:22:09 GMT" Content-Transfer-Encoding: base64 LS0gDQpodHRwOi8vd3d3LnBpY2xpc3QuY29tIFBJQy9TWCBGQVEgJiBsaXN0IGFyY2hpdmUNClZp ZXcvY2hhbmdlIHlvdXIgbWVtYmVyc2hpcCBvcHRpb25zIGF0DQpodHRwOi8vbWFpbG1hbi5taXQu ZWR1L21haWxtYW4vbGlzdGluZm8vcGljbGlzdA0K --_002_3ef6s9xc27ln2skaterswarbrick_-- .