Peter Mueller wrote:
Am Sonntag, 12.09.04 um 12:47 Uhr schrieb Arnd-Hendrik Mathias:
in fact the circuit plan is only half of the truth because it is a
generic schematic. It only figures out the logical level regardless of
the concrete example. The documentation states the plan to communicate
witha PC-RS232 interface. This doesn't only mean inverting the logical
level of the signal but furthermore converting the low TTL signal -
which might be about 3V (?) high while not transmitting anything, 0V
for the start bit and again 3V (? usually depends on HW) for any 1 set
Bit and Stop Bit - to the logical inverted +-12V signal of a RS232
interface (self explanatory you cannot expect this level from a max.
RS232 expects at least +/- 3V on the transmitter side (+/- 2V on the
receiver). it also allows for more than +/- 12V.
The maxim chips usualy double and invert the supply voltage, so you'll
get +/- 6V at 3V VCC.
using only 0V for the low signal isnt according to the RS232 specs, but
works on many PC hardware. i wouldn't use it in a product but for
developement cictuits its ok.
3.3V powered MSP ;°). Possible Hardware you could use is stated in the
section "Demonstration Circuit" in slaa078a.pdf or any other standard
RS232 converter (e.g. MAXIM, etc....).
Hope this helps a little bit. Have much fun
Arnd-Hendrik
Hi Arnd-Henrik,
I already have an adapter that converts RS232 levels to levels which can
be used from the MSP. What I want to get rid is the inverter in the
application note. At the moment the signal coming from the MSP is low
active, but what I need is a high active signal from the capture compare
unit ...
as you implicit point out, that is only possible with a software uart.
the serial hardware of the msp430 does not support inversion.
As I wrote playing with the ccr settings was not successful so far.
you cant invert the signal in the hardware, but you can invert the bit
before shifting in: invert the carry bit:
xor #BIT0, r2 ; inverted input
or you can invert the whole byte at the end, which is probably better,
performance wise.
inverting the interrupt edge for the startbit detection should be
trivial too.
chris