topic Re: Power up problem :LTC3589 PMIC - i.mx536 in i.MX Processors

Power up problem :LTC3589 PMIC - i.mx536

yekta — Thu, 29 Oct 2020 08:56:41 GMT

Hi,
I'm using LTC3589 for powering freescale imx536 processor.
My assumption was  that without accessing using I2C_SDA and I2C_SCLK ,
using the default values in the  LxDTVx and BxDTVx ,
and using R1 and R2 resistor values I can power up the system .

But during the measurements I have observed the following problem:

Altough Power_ON is continously high following signals 
have only the the intended  values for 14 msecs within one second period.
--------------------------------------------------------------
LTC_WAKE ,
LDO2_1V3 , 
LDO4_2V8 ,
SW1_1V1  ,
SW2_1V3  ,
SW3_2V5  ,
VDD_ANA_PLL

Following signals are continously high (at expected values)
-----------------------------------------------------------
LDO1_1V3,
LDO3_2V8,
SW4_3V3

My Board configuraion is as follows :
For power on sequencing ,I have connected the enables and regulator outputs as follows.

By forcing the POWER_ON pin to high  the WAKE signal will be high.

WAKE= ENable-2    => SW2_1V3  will be up 

SW2_1V3 =Enable-3 => SW3_2V5  will be up 

Sw3_2V5 =VDD_REG of imx53  imx53 internal regulators will be up VDD_ANA_PLL=1.8V
will be generated

VDD_ANA_PLL=En_LDO2=En_LDO34 =>SW1_1V1 , LDO3_2V8 ,LDO4_2V8,LDO2_1V3  will be up 

VDD_ANA_PLL=Enable of TPS54218(for 1V8 DDR2 voltage generation) 

DDR2_1V8=Enable-4 => SW4_3V3 will be up

Re: Power up problem :LTC3589 PMIC - i.mx536

FelixVidal — Thu, 12 Apr 2012 13:59:22 GMT

I've also used this PMIC for our design and, I also had some problems... I assume you are using LTC3589-1 or LTC3589 -2 (and not LTC3589). I recomend you see the following documents:

a) From Freescale, MX53UG.pdf, pages 105 and following,

b) from Linear the last revision of LTC3589 datasheet, page 46.

Re: Power up problem :LTC3589 PMIC - i.mx536

yekta — Thu, 12 Apr 2012 16:43:52 GMT

I ordered LTC3589-1 but somehow procurement bought LTC3589 .

Do you thing that the following mode of the operation is possible?

VSTB is connected to ground and Power_on is driven by a watchdog timer ,which goes high 400msecs after power up .

No need to access the I2C registers of PMIC , to keep the regulators alive ?

Re: Power up problem :LTC3589 PMIC - i.mx536

FelixVidal — Thu, 12 Apr 2012 17:03:56 GMT

There is something I don´t understand. If you are using the LTC3589, what do you do with the output of LDO3 (for this reference is +1.8 V?. But if you don't use the I2C and you use the LTC3589-1, how do you turn on LDO4 (only switches on by I2C)? .

Re: Power up problem :LTC3589 PMIC - i.mx536

yekta — Wed, 06 Jun 2012 18:29:02 GMT

I replaced the LTC3589 with LTC3589-2 (samples) .The problem is solved.

According the LTC support:

"For the LTC3589 (or -1), no regulator can have more than 300mV on it when the regulator is enabled, or it will refuse to start. For the LT3589-2 this rule has been removed, and the regulator will start even when there is 300mV or more of voltage on the output when the regulator is enabled.

Usually with applications processors this is leakage from a regulator that is on to one that has not yet started, resulting in failing the 300mV prebias rule."

Thanks to the support.