Lpc2000

I need LPC2132 to run @32MHz for 2mSec (my time measurement periode)
and then drop speed to 4MHz for 100mSec. This cycle continues forever. 
Is it possible to change MSEL while running? I am afraid of Capture 
range, Locking range and stabilizing period etc.

Anybody experianced with this?

Vimal Amin

--- In lpc2000@yahoogroups.com, "Vimal Amin" <vimal19542000@y...> 
wrote:
>
> I need LPC2132 to run @32MHz for 2mSec (my time measurement periode)
> and then drop speed to 4MHz for 100mSec. This cycle continues
> forever. 
> Is it possible to change MSEL while running? I am afraid of Capture 
> range, Locking range and stabilizing period etc.
> 
> Anybody experianced with this?

Not reprogramming the PLL, but you can connect and disconnect it to 
switch between the crystal frequency and the PLL frequency.  This 
switching can be done instantaneously, when you just leave the PLL 
running all the time.  I have used this as a workaround for the old 
buggy bootloader on LPC2114 that didn't work at high cclk speeds.

Btw, the PLL input requires at least 10 MHz, so you can't go down to 
4 MHz.

SYS->PLLCON  = 0x01;  /* PLL = enabled, not connected */
SYS->PLLFEED = 0xAA;
SYS->PLLFEED = 0x55;

... running at slow Fosc speed

SYS->PLLCON  = 0x03;  /* PLL = enabled and connected */
SYS->PLLFEED = 0xAA;
SYS->PLLFEED = 0x55;

... running at full PLL speed


Karl Olsen

Thanks Karl. You are right.

But there is a bit, PLOCK. I am not very clear about application.

Can I do as follows?
Disable PLL (Not just switch). I neet to conserve power. And while enabling
again, I check this bit again befor switching to PLL again.

Regards

Vimal


[Non-text portions of this message have been removed]

---- Original Message ----

From: "Digital Systems, Partner" <dsystem@...>
To: <lpc2000@yahoogroups.com>
Sent: Friday, January 20, 2006 6:42 PM
Subject: Re: [lpc2000] Re: changing PLL multiplier on fly - PLLCFG
(MSEL)

> But there is a bit, PLOCK. I am not very clear about application.
>
> Can I do as follows?
> Disable PLL (Not just switch). I neet to conserve power. And while
> enabling again, I check this bit again befor switching to PLL again.

I guess you can.  Other than the temporary PLL disconnecting, I have only 
done the normal PLL enabling at startup:

SYS->PLLCFG  = 0x23;  /* P=2, M=4 */
SYS->PLLCON  = 0x01;  /* Enable PLL */
SYS->PLLFEED = 0xAA;
SYS->PLLFEED = 0x55;
while (!(SYS->PLLSTAT & 0x400));      /* Wait until PLOCK */
SYS->PLLCON  = 0x03;  /* Enable and connect PLL */
SYS->PLLFEED = 0xAA;
SYS->PLLFEED = 0x55;


To conserve power, disconnect and disable PLL, and enter Idle mode as much
as possible.  Powerdown mode would be better, but only an external interrupt
or reset will bring you out of it.

When waking up, enable but not yet connect the PLL.  According to the
datasheet, the PLL takes 100 usec to lock.  Don't trust this number, wait
until the PLOCK bit is set, which tells that it has locked.

While you wait for it to lock, you can continue working at the low clock.
If you have nothing useful to do at the low speed, you'll conserve most
power by enabling the PLL interrupt so that you get an interrupt when PLOCK
is set, and while waiting for the interrupt, you enter Idle mode.  When the
PLL has locked, your PLL interrupt handler is called, and in that you
connect the PLL, and start the real processing.  This means that the CPU is
idle, not active, during the ~100 usec lock time.

Also remember to disable all unused peripherals in PCONP, and don't pull any
P1.x pins low (neither by external circuitry, nor by having them as GPIO
outputs that are low), as there are permanent internal pull-ups on these
pins.

Karl Olsen

Karl Olsen wrote:

>---- Original Message ----
>From: "Digital Systems, Partner" <dsystem@...>
>To: <lpc2000@yahoogroups.com>
>Sent: Friday, January 20, 2006 6:42 PM
>Subject: Re: [lpc2000] Re: changing PLL multiplier on fly - PLLCFG
>(MSEL)
>
>  
>
>>But there is a bit, PLOCK. I am not very clear about application.
>>
>>Can I do as follows?
>>Disable PLL (Not just switch). I neet to conserve power. And while
>>enabling again, I check this bit again befor switching to PLL again.
>>    
>>
>
>I guess you can.  Other than the temporary PLL disconnecting, I have only 
>done the normal PLL enabling at startup:
>
>SYS->PLLCFG  = 0x23;  /* P=2, M=4 */
>SYS->PLLCON  = 0x01;  /* Enable PLL */
>SYS->PLLFEED = 0xAA;
>SYS->PLLFEED = 0x55;
>while (!(SYS->PLLSTAT & 0x400));      /* Wait until PLOCK */
>SYS->PLLCON  = 0x03;  /* Enable and connect PLL */
>SYS->PLLFEED = 0xAA;
>SYS->PLLFEED = 0x55;
>
>
>To conserve power, disconnect and disable PLL, and enter Idle mode as much
>as possible.  Powerdown mode would be better, but only an external interrupt
>or reset will bring you out of it.
>
>When waking up, enable but not yet connect the PLL.  According to the
>datasheet, the PLL takes 100 usec to lock.  Don't trust this number, wait
>until the PLOCK bit is set, which tells that it has locked.
>
>  
>
I am also considering varying the clock to save power, won't this raise 
havoc with the peripherials?  AFAICT, the uart baudrate, timer interrupt 
values, etc., all will have to be reprogrammed to match the new PCLK value?

Regards,

TomW


-- 
Tom Walsh - WN3L - Embedded Systems Consultant
http://openhardware.net, http://cyberiansoftware.com
"Windows? No thanks, I have work to do..."
----------------------------------------------------

---- Original Message ----

From: "Tom Walsh" <tom@...>
To: <lpc2000@yahoogroups.com>
Sent: Friday, January 20, 2006 10:47 PM
Subject: Re: [lpc2000] Re: changing PLL multiplier on fly - PLLCFG
(MSEL)

> I am also considering varying the clock to save power, won't this
> raise havoc with the peripherials?  AFAICT, the uart baudrate, timer
> interrupt values, etc., all will have to be reprogrammed to match the
> new PCLK value?

Yes.  You can reduce havoc by switching between PLL multiplier=4 and
APBDIV=4, and PLL disabled and APBDIV=1.  Except for a little disturbance
during the switching, the peripherals continue unchanged.

Karl Olsen

PLL settings for power conservation...

Thanks a lot to Karl to confirm my assumptions.

Warm regards

Vimal Amin