Re: [avr-libc-dev] atmega324 variants (P, A, PA) signature

[Joerg Wunsch]

As rr wrote:

> Sorry, but I'm using linux, so I can't help you about AVR Studio.

I'm using FreeBSD, so I'm in similar troubles here. ;-)

To any Windows + AVR Studio users out there: please check whether the
current XML part description file still has the bug, and if so, file a
bug report to Atmel.

> > There's no difference from a library point of view anyway.

> Then, why only mega324a and mega324pa have different signature bytes?

The library doesn't use the signature bytes, they are only copied over
from the XML files into the header files.

> > I thought there's an appnote epxlaining the transition to the "A"
> > devices, isn't there?

> Yes, of course! Sorry, I meant mega324a and mega324pa.

There was no plain (i.e. without "P" or "A") ATmega324 at all.  Only
the ATmega644 has been available without a suffix letter, and unlike
its siblings, it lacks the second USART.  Later on, the picopower
devices came up on the market, so the smaller siblings were only
produced as ATmega324P and ATmega164P.  To add to the confusion, an
ATmega644P variant was then introduced that not only added the
picopower feature set (compared to the plain ATmega644), but also the
second USART.

Now, with the transition to the "A" die-shrink devices, one would have
expected to only see ATmega164PA/324PA/644PA anymore, but it appears
that there's enough market request for also having the non-picopower
variants available, so we can also see an ATmega324A (rather than only
the ATmega324PA).  Obviously, as both have different feature sets,
they require different device IDs.  (In most other cases, the "A"
variants are identical in their feature set to the non-A devices, so
they can use the same device ID.)

-- 
cheers, J"org               .-.-.   --... ...--   -.. .  DL8DTL

http://www.sax.de/~joerg/                        NIC: JW11-RIPE
Never trust an operating system you don't have sources for. ;-)