18F67K22 UART NO-GO if OSC=64MHz

[avatarengineer]

Using latest rev 5.017.
UART doesn't play on 18F67K22 when PLL used and OSC=64MHz.
Looking for typical 9600Baud with 64MHz Osc to work.
All works Ok with 16MHz Osc (no PLL).

Tried many variations of #USE232, etc...
Anyone with proven code example?

Thanks,
AE

[Ttelmah]

Have a look at the link from this thread:

<http://www.ccsinfo.com/forum/viewtopic.php?t=51799>

May be the same problem....

[avatarengineer]

didn't work

[PCM programmer]

Explain what you mean by "didn't work". Do you mean you get garbled
output ? Partial or completely garbled on all characters ?

Post your test program. Post something like this for your main():

[avatarengineer]

Not a newbie here.

within Main loop is;

IF(kbhit()) {
Charin=fgetc(Comport1);
...do stuff...
}

works at 16MHz, not at 64MHz.
in other words, at 64MHz, there is no kbhit.

[Ttelmah]

Answer the rest of PCM_programmers questions:

"Also post the #include line for your PIC, #fuses, #use delay(), and
#use rs232(). Post any other setup code. Keep the program short.
Also tell us if you are using the internal or external oscillator. If external,
are you using a crystal or a oscillator package ? Also tell us the Vdd
voltage of your PIC.

Finally, tell us how you are testing it. Are you typing characters at low
speed from a terminal program and looking for them to be echoed back ?
(As shown in the code above). Or something else ?".

[Ttelmah]

As a comment, check the supply voltage actually _at_ the PIC pin.

There are some threads on the MPLAB forums with various 'having problems at 64MHz" connections, and in two cases the people were powered off an ICD, and when they switched to 64MHz, the load went up enough to cause the supply at the PIC pin to droop, and the chips started to behave 'peculiarly'....

Best Wishes

[sezgin ozan]

Hi, if you use pll then your rate is also multiplied by 4 and becomes 38400,

sezgin

[Ttelmah]

[ilker07]

for example you set uart

#pragma use rs232(baud=14400, xmit=PIN_G1, rcv=PIN_G2, stream=TTL, errors)

Then baudrate on serial terminal 14400 * 8 =115200

[Ttelmah]

NO.

Blatantly wrong.

With CCS, if you set the clock statement correctly, the baud rate is
what you set in the #USE RS232. Nothing else.

What you post is wrong.

It would only behave as you describe, if you incorrectly set the #USE
DELAY, with the crystal rate, not the clock rate, and then manually turn
on the PLL yourself, rather than letting the compiler do this for you.

Incorrect #FUSE/#USE settings.

[temtronic]

Since it works at 16MHz and not 64MHz, build both versions and dump the listings to see IF the compiler is setting the PLL and UART bits properly.
If you use the PLL (x4) , I'd expect the UART BRG to be /4 to get the same speed (9600 in your case )

It may be a compiler bug ?

Have you tried previous compiler versions ?

[Ttelmah]

If you look Jay the original problem was in 2014. The latest poster
just attached a completely 'wrong' statement to the end of the thread.
Dozens (hundreds?) of us here have since used this chip at 64Mhz OK.
Just using correct clock settings, fuses and baud rate settings.