Robert Rolf wrote:
> Harold Hallikainen wrote:
>>> alan smith wrote:
>>>>  From my experiance, you need to do at least ONE bulk erase with 5V,
>>>> and then forever more, it will work fine with 3.3V.  Its hit and miss
>>>> if you don't do the one time bulk erase, but
>>>>  try that and see if it works for you.
>>> I can confirm the same bizarre requirement. 16F876As. Would not verify
>>> using ICSP @ 3.3V.
>>> After one bulk erase at 5V (all parts were tolerant of 5V) loaded fine,
>>> and bootloader
>>> then worked correctly. Weird, but now I just put in the bootloader
>>> (ROMZap) @ 5V
>>> and then all is well.
>>>
>>> Robert
>> Is there a chance the chips were code protected? It seems the only way to
>> get rid of that is a bulk erase, which, of course, requires 5V. You can,
>> however bootload at 3.3V with code protect enabled.
>>
>> Harold
> 
> I didn't specifically look at the state of the config.
> These were factory fresh parts, already mounted, and I just needed to
> 'make them go' yesterday. The default for an erased CONFIG is NOT code protected,
> but I suppose it was possible that factory testing left them with CP set.
> I'll see if I still have one lying around from that batch to see what state it
> was in. Seems to be a logical explanation for the observed problem.

Derned if I didn't see the same thing today!  The same chip that I had a 
problem with before - PICkit2 complained, I overrode, and it failed to 
verify - works now, after having programmed it at 5V.

So I tried the other two chips from the same batch of sample 16F886s... 
and both worked the first time at 3V target power.  Wrote and verified 
with no errors, and functionally it looks fine.

Well, I guess the ICD2 will be a useful tool for other reasons.  :)
-- 
Timothy J. Weber
http://timothyweber.org
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist