Re: PXA270 - Sleep & Wake
- From: "Jim Merkle" <jimmerkle@xxxxxxxxxxx>
- Date: Tue, 3 May 2005 09:28:08 -0700
We found the hardware problem.
1) PWREN drops from 3.3V to Ground.
2) VCC_CORE drops from 1.6V to Ground
3) GPIO outputs assume their PGSRx defined states
I've tried multiple core frequencies. An errata mentioned 91MHz, so I was
trying that.
A=0, 2N=3, L=7, B=0, HT=0, F=1, T=0 (91Mhz)
One clue that we tracked down is we have an external 32K RTC oscillator,
with which we are driving the TXTAL_IN input. The loader enables the 32K
clock. When we entered sleep, this 32K clocked stopped. The problem was
the 32K clock source is an open collector output, with a pull-up resistor
going to a PWREN switched supply. When the PWREN signal dropped, so did the
supply, so did the pull-up, so did the 32K clock. With a cut & jump, we now
have an external 32K clock as well as sleeping and waking.
Thank you everyone!
Jim
"voidcoder" <voidcoder@xxxxxxxxx> wrote in message
news:%23R9Ba76TFHA.2976@xxxxxxxxxxxxxxxxxxxxxxx
> Are You sure that CPU is switched to SLEEP? Check the core voltage
> to see if You really enter SLEEP. What is the core frequency when
> You are trying to enter SLEEP mode?
>
>
> "Piet" <pietdesomere@xxxxxxxxx> wrote in message
> news:b498cce3.0505020418.1dcc1148@xxxxxxxxxxxxxxxxxxxxx
>> Has GPIO 0 been configured as an input in the GPDR register?
>>
>> P
>>
>>
>> "Jim Merkle" <jimmerkle@xxxxxxxxxxx> wrote in message
> news:<eeUlY$QTFHA.3308@xxxxxxxxxxxxxxxxxxxx>...
>> > I'm having difficulty putting my PXA270 to sleep and then waking it via
>> > GPIO_0.
>> > In order to more rapidly test this, I put the following module into my
>> > loader, which can be invoked as a command:
>> > No attempt is being made in this example to preserve any specific
>> > state,
>> > conserve power, or preserve SDRAM. Should the processor wake from
> sleep,
>> > the loader will display the reset cause and the values of the PSSR and
> RCSR
>> > registers.
>> >
>> > // Prototype assembly function that pokes R0 into CP14, register 7
>> > extern "C" void WritePWRMODE(DWORD dwSleepMode);
>> > #define PWRMODE_SLEEP 0x00000003 // Value for cp14: Reg7 to induce
> sleep.
>> >
>> > // Perform minimal steps necessary to configure the PXA270 for sleep
>> > and
>> > // wake up from the Philips, then initiate a sleep.
>> > STATUS CmdSleep(int argc, char * argv[])
>> > {
>> > volatile PMRC_REGS * pPmrcRegs = (PMRC_REGS *)PWR_BASE_U_VIRTUAL;
>> >
>> > // Configure the state of the GPIO outputs during sleep
>> > /*
>> > ; PGSR0: 0 .. 31
>> > PGSR_nEnCF33V GPIO_13 ; 0: enable 3.3V to User CF card,
>> > 1:
>> > turn off 3.3V
>> > PGSR_nEnCF50V GPIO_14 ; 0: enable 5.0V to User CF card,
>> > 1:
>> > turn off 5.0V
>> > PGSR_nEnUtilPwr GPIO_22 ; 0: enable 5V supply, 1: dissable
> 5V
>> > supply
>> > PGSR_FF_nRTS GPIO_27 ; For "normal" RS-232, we want this
> line
>> > high, for RS-485, we want this low
>> >
>> > ; PGSR1 32 .. 63
>> > nFFUART_RS422 GPIO_32 ; If we want the RS-232 transceiver
>> > enabled, raise this line
>> > PGSR_nPOE GPIO_48
>> > PGSR_nPWE GPIO_49
>> > PGSR_nPIOR GPIO_50
>> > PGSR_nPIOW GPIO_51
>> > PGSR_nPCE2 GPIO_54
>> >
>> > ; PGSR2 64 .. 95
>> > PGSR_FF_nDTR GPIO_82
>> > PGSR_nPCE1 GPIO_85 ; another PCMCIA bit
>> >
>> > ; PGSR3 96 .. 120
>> > PGSR_EnCoreLow GPIO_100 ; 1: 1.0V core voltage
>> > */
>> > pPmrcRegs->pgsr0 = GPIO_13 | GPIO_14 | GPIO_22 | GPIO_27;
>> > pPmrcRegs->pgsr1 = GPIO_48 | GPIO_49 | GPIO_50 | GPIO_51 | GPIO_54;
>> > //pPmrcRegs->pgsr1 = GPIO_32 | GPIO_48 | GPIO_49 | GPIO_50 |
>> > GPIO_51
> |
>> > GPIO_54;
>> > pPmrcRegs->pgsr2 = GPIO_82 | GPIO_85;
>> > //pPmrcRegs->pgsr3 = GPIO_100;
>> > pPmrcRegs->pgsr3 = 0;
>> >
>> > // Configure the wake-up event
>> > pPmrcRegs->prer = GPIO_0; // (GPIO_0 rising edge) 0x00000001
>> > pPmrcRegs->pfer = GPIO_0; // (GPIO_0 falling edge) 0x00000001
>> > pPmrcRegs->pwer = GPIO_0; // (GPIO_0 event) 0x00000001
>> >
>> > // Configure the PCMCIA output drive, 13Mhz oscillator during sleep
>> > pPmrcRegs->pcfr=0;
>> >
>> > // Sleep Configuration register
>> > // Set Ramp up for power domains, ignore VDD_FAULT, Keep P1 domain
>> > powered with clocks running
>> > //pPmrcRegs->pslr = BIT_31|BIT_30|BIT_27|BIT_26|BIT_22|BIT_3;
>> > pPmrcRegs->pslr = 0xCC400008;
>> >
>> > printf("CmdSleep: Going to sleep...\r\n");
>> >
>> > // Wait for the last character in the above line to get out
>> > msWait(1); // wait one millisecond
>> >
>> > WritePWRMODE(PWRMODE_SLEEP); // Initate SLEEP (0x03 - regular
> sleep)
>> >
>> > // Do not return!!!
>> > while(1);
>> >
>> > return STATUS_SUCCESS;
>> > } // CmdSleep
>> >
>> > The processor appears to be sleeping peacefully, but it refuses to wake
> when
>> > pulsed on GPIO_0. I can tell the processor has entered sleep since the
>> > settings in PGSRx have taken effect, driving each of the GPIOs to its
>> > requested state. (There's still the possibility we have a hardware
>> > problem...)
>> >
>> > Any ideas?
>> >
>> > Thank you,
>> >
>> > Jim Merkle
>
>
.
- References:
- Re: PXA270 - Sleep & Wake
- From: Piet
- Re: PXA270 - Sleep & Wake
- From: voidcoder
- Re: PXA270 - Sleep & Wake
- Prev by Date: Re: Why am I getting an assert when I try to Load a bitmap that I have added to the image?
- Next by Date: Re: audio stream type identification
- Previous by thread: Re: PXA270 - Sleep & Wake
- Next by thread: about usb-to-serial(RS232) device driver problem...
- Index(es):
Relevant Pages
|
