MPC5200 baud rate frequency

[Didi]

I have a Freescale MPC5200 board I designed a while ago which
comes out of reset (I am writing the flash with a debug monitor
which has served me over the years - I wrote it first for the 68020,
then
I made a CPU32 version, then an MPC8240 version etc...), and the
monitor says something through the UART (I use PSC_6).

But the baud rate is totally off the mark.

I made the other system say the same through its UART
and, measuring the length (43 characters, enough to get an idea
if speeds match) I see the new system is apr. 2.7 times slower
than it should be.

Great, I though I'd just write a 2.7 times lower value in the counter
timer and see what happens. Well, the thing is, nothing happened.
Same time (2.7 times what it should be) again. I just love that kind
of thing when you make dramatic changes only to achieve no change.
I am writing to the so called Counter Timer Upper and then Lower
registers, treat them as 8 bit wide (i.e. I write using a .b access,
stb) and I write the value of 215, which should yield a 19200 baud
given the 132 MHz IPB clock. I already described the effect of
changing the 215 to 79 (or sort of, I lost the precise number, it was
not exactly 2.7, so it may have been 78 or 80, not that it matters).

I am tired now and will reluctantly go to sleep (I hate doing that
with
unsolved basic issues like that).
Hopefully someone will have an idea what am I doing wrong which
will help me tomorrow...

Thanks,

Dimiter

 

[jasen]

But the baud rate is totally off the mark.

I made the other system say the same through its UART
and, measuring the length (43 characters, enough to get an idea
if speeds match) I see the new system is apr. 2.7 times slower
than it should be.

Great, I though I'd just write a 2.7 times lower value in the counter
timer and see what happens. Well, the thing is, nothing happened.
Same time (2.7 times what it should be) again. I just love that kind
of thing when you make dramatic changes only to achieve no change.
I am writing to the so called Counter Timer Upper and then Lower
registers, treat them as 8 bit wide (i.e. I write using a .b access,
stb) and I write the value of 215, which should yield a 19200 baud
given the 132 MHz IPB clock. I already described the effect of
changing the 215 to 79 (or sort of, I lost the precise number, it was
not exactly 2.7, so it may have been 78 or 80, not that it matters).

I am tired now and will reluctantly go to sleep (I hate doing that
with unsolved basic issues like that).

Not being familar with that hardware I see three possible mistakes.

I hope the new day has brought you success.

1> you were writing the correct value but later overwriting it
2> you were writing the correct value but to the wrong location
3> you were writing the correct value but with an incorrect setup

It may be worth checking the errata in the hardware documentation.

Bye.
Jasen

 

[Didi]

Not being familar with that hardware I see three possible mistakes.

I hope the new day has brought you success.

Ah, it did, thanks. There had been more to it than just the baud rate,
I rewrote the UART initialization code in a less ambitious manner
(thus more
reasonable for a first-time out of reset CPU type here) and it worked.

Dimiter