> two 27pf caps to ground on each pin of cristal

> Also changed cap values, used other cristals, and other
> 16F6284s !!!!!!!!!!!

I always use 47pF. It's a little slow starting but it keeps going

>, but it usually stops oscillating after a few minutes. If I put the
> scope probe on T1OSO or T1OSI, it starts oscillating again

Sounds like a physical fault. I'd check for a short and remove
any flux that may be bridging tracks. How's the humidity where
you are ? But you say you've changed the caps and crystal. Often
that soldering/de-soldering would change or remove the fault

Are the caps really grounded ?

I would have suggested a weirdness in your s/w that turns TMR1
off or does something strange to RB6 / RB7. That doesn't explain
why the turn-off time changes wrt a parallel resistor

However, as you don't know the cause of the problem, you can't
rule anything out, neither obvious nor obscure

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics