AM335x PMIC_POWER_EN

Mauro Cleris

Other Parts Discussed in Thread: AM3352

Hello,

we are using AM3352 processor with TPS65217B PMIC.

We give the 'power_off' command through LINUX kernel and after 2s an ALRM2 event pull-down the PMIC_POWER_EN signal and the SHUT-DOWN sequence is started. Up to here everything is clear.

We note that after the 3V3 domain is SHUT-DOWN, it seems that the AM3352 pulls HIGH the PMIC_POWER_EN pin again: this starts the POWER-UP sequence and the system restarts.

Is this behavior correct? Looking at the datasheet we cannot understand the reason of this behavior.

Anybody can help us?

Many Thanks,

Mauro

over 11 years ago

0 Biser Gatchev-XID over 11 years ago

TI__Guru**** 393215 points

Hi Mauro,

PMIC_POWER_EN could be set high by an alarm or external wakeup event (AM335X TRM Rev.I, section 20.3.3.8). These should be the things to check.

0 Biser Gatchev-XID over 11 years ago in reply to Biser Gatchev-XID

TI__Guru**** 393215 points

Can you post how is EXT_WAKEUP connected, and also RTC_KALDO_ENn, CAP_VDD_RTC, RTC_PWRONRSTn?

0 Mauro Cleris over 11 years ago in reply to Biser Gatchev-XID

Prodigy 170 points

Hello Biser,

thanks for the fast reply.

EXT_WAKEUP: conncted between TPS65217B (WAKEUPn pin) and AM3352 with 10kOhm pull-up res at 1V8 RTC power domain.

RTC_KALDO_ENn: connected to GND through a 10kOhm res.

CAP_VDD_RTC: connected to GND through a 1uF cap.

RTC_PWRONRSTn: conncted between TPS65217B (LDO_PGOOD pin) and AM3352. In this case the signal level is scaled through a resistive partitor (10kOhm - 12.1kOhm).

Hope this is enough.

0 Biser Gatchev-XID over 11 years ago in reply to Mauro Cleris

TI__Guru**** 393215 points

Everything seems correct on the hardware side. Have you checked if the EXT_WAKEUP signal stays stable when this happens? If there is a transition on EXT_WAKEUP it can trigger PMIC_POWER_EN.

0 Mauro Cleris over 11 years ago in reply to Biser Gatchev-XID

Prodigy 170 points

Hi Biser,

what I cannot understand is if the behavior is correct or if it is anomalous.

Now i am going to check the signals you suggested.

0 Biser Gatchev-XID over 11 years ago in reply to Mauro Cleris

TI__Guru**** 393215 points

I am trying to help you find the reason that causes this behaviour. This signal can not toggle on its own.

0 Mauro Cleris over 11 years ago in reply to Biser Gatchev-XID

Prodigy 170 points

Hi Biser,

I attach a scope image, the PWR_EN signal is driven low by the AM3352,

after a while the LDO_PGOOD signal starts to go down and the PWR_EN is driven high (this is not clear).

The nWAKEUP signal goes to a strange level around 1V, even this is not clear because there is an external 10kOhm pull-up resitor

0 Mauro Cleris over 11 years ago in reply to Mauro Cleris

Prodigy 170 points

Here the image...sorry!

0 Biser Gatchev-XID over 11 years ago in reply to Mauro Cleris

TI__Guru**** 393215 points

Hi Mauro,

From what I see it looks like the VRTC power rail, to which nWAKEUP is pulled up, gets overloaded and VRTC drops when the processor enters standby. This generates a front on nWAKEUP which takes the processor out of standby. Can you check VRTC behaviour to confirm if this is the case?