Default clock for a MSP430G2553 (LaunchPad)

[xv4y 46]

Hi,

 

What are the default clock settings (BCSCTLx registers) for an MSP430G2553 in Energia ?

 

I have my beacon running but the clock is slowly drifting (30 seconds in 12 hours) and the beacon sequence is not synchronized anymore so the receivers are not decoding it.

I think this is because I have interrupts call too often (each 1/256th of second) and the MCU too slow, but I am not sure...

 

Thanks,

Yan.

[CorB 64]

Hi,

 

You might have a look into this:

http://electronics.stackexchange.com/qu ... ld-weather

 

The DCO is rather sensitive for temperature changes.

 

cheers

 

CorB

[Rickta59 589]

What are the default clock settings (BCSCTLx registers) for an MSP430G2553 in Energia ?

 

It defaults to 16MHz for the msp430g2553. The value of the define 'F_CPU' is determined by which

board is selected in the Energia IDE. This value is stored in the boards.txt file and each board can have

its own .f_cpu. value. Here is the default entry for the msp430g2553:

 

lpmsp430g2553.build.f_cpu=16000000L

The core code uses the F_CPU value to decide at compile time which precalibrated setting to use

for the BSCCTL1 and DCOCTL registers.

 

void initClocks(void)
{
#if defined(CALBC1_16MHZ_) && F_CPU >= 16000000L
BCSCTL1 = CALBC1_16MHZ;
DCOCTL = CALDCO_16MHZ;
#elif defined(CALBC1_12MHZ_) && (F_CPU >= 12000000L)
BCSCTL1 = CALBC1_12MHZ;
DCOCTL = CALDCO_12MHZ;
#elif defined(CALBC1_8MHZ_) && (F_CPU >= 8000000L)
BCSCTL1 = CALBC1_8MHZ;
DCOCTL = CALDCO_8MHZ;
#elif defined(CALBC1_1MHZ_) && (F_CPU >= 1000000L)
BCSCTL1 = CALBC1_1MHZ;
DCOCTL = CALDCO_1MHZ;
#else
       #warning No Suitable Frequency found!
#endif
/* SMCLK = DCO / DIVS = nMHz */
BCSCTL2 &= ~(DIVS_0);
/* ACLK = VLO = ~ 12 KHz */
       BCSCTL3 |= LFXT1S_2; 
}

 

-rick

[xv4y 46]

Hi,

 

Thanks for your replies.

CorB, I forgot to tell I hooked up a 32KHz Xtal to the LaunchPad external clock pads.

I am rather used to solder SMD but this one was not the easiest to solder since the board is already populated...

 

Rick, yes I saw the second code you give but missed the board.txt...

I think my problem is somewhere else... the drift is worst than I thought!

I also use the delay() function which make usage of the WatchDogTimer.

I will try to use only my own "delay" using the RTC lib.

 

Regards,

Yan.

[energia 485]

Note that Energia uses the internal VLO for ACLK and not the 32KHz XTAL. See hardware/msp430/cores/msp430/wiring.c

 

/* ACLK = VLO = ~ 12 KHz */
BCSCTL3 |= LFXT1S_2; 

 

This makes me realize that there is there is no way for the user to set the ACLK to the external XTAL using a higher level API.

 

-robertinant

[xv4y 46]

Note that Energia uses the internal VLO for ACLK and not the 32KHz XTAL. See hardware/msp430/cores/msp430/wiring.c

-robertinant

Hi Robert,

 

Does this mean that I am unable to set ACLK so it uses the XTAL ?

The results I have are in line with my expectations...

 

I think I found my error but I have to run a long test to check the RTC drift.

I set CCR0 to 128 (as I want to mesure 128 periods of the xtal oscillator) but I read that CCR0 counts N+1 not N.

No it is to 128-1 and hover 10 minutes the clock is ok.

TI documentation is not really precise about this, and the examples not all good.

Sometime you find CRR0 = N-1 but nobody explains why...

 

Yan.

[oPossum 1,083]

For 128 cycles, you have to set CCR to 127. This is explained in SLAU144 section 12.2.3.1: "The number of timer counts in the period is TACCR0+1"

 

I think you can set the ACLK source to whatever you want after Energia has done it's initialization.

[energia 485]

Hi Yan,

 

You can set XTAL as ACLK source and can do so by setting BCSCTL3 in wiring.c. But I am sure you already figured that one out :-) It would be nice if this was possible using a higher level API.

 

Anyways, happy you got it to work!

 

-robertinant

[xv4y 46]

Hi,

 

Now my beacon is working and signals have been received by another station.

The only problem is that the output power is a bit low (100mW) and propagation conditions not great at this season for the band I choose (30 meters band is more a winter band), but this RF related, not MSP430 related.

 

The clock drift is around 1 second less for 4 hours but I have to measure it.

It seems in line with the 20ppm rating of the xtal.

I will have to correct the drift either by software (adding a leap second), by changing the XCAP_x value in BSCTL3 or by hardware pulling a little bit the xtal (adding an adjustable cap)...

Perhaps I can read one pin and depending off the presence of a jumper, set XCAP_x value...

 

Thanks for your help.

Yan.