PF3001+IMX6ULL temperature shutoff mechanism for longer duration

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

PF3001+IMX6ULL temperature shutoff mechanism for longer duration

1,031 Views
vivekrajpara
Contributor III

Hi,

We have a requirement of temperature shutoff mechanism with IMX6ull + PF3001, we need help in root causing one issue as described below. Looking forward to a helping response to the issue.

Set up:

We are keeping the device ON at 85 deg C constant temperature for 15 -16 hours. In this condition except for PMIC, MPU, and DDR, all other components are in powered off. The device should be recovered when the temperature comes below 70 deg C.

Issue:

If we increase the temperature of PCB up to 85 deg C and keep for a shorter duration like 1-2 hours. In this condition, the processor is successfully able to come out of this U-Boot condition when the temperature cools down to 70 degC.

It doesn't work properly if we keep PCB 85 deg C for longer duration like 8-15 hours (Randomly observed in 9 hours, 13 hours in 2 devices). In this condition, it is not coming to the normal condition even after cooling down the system to 70 deg C.

Analysis:

-> When we observe SOC_IN and HIGH_IN power, they are off when the issue is reproduced although PMIC input is proper and uninterrupted.

-> PMIC_POWERON_REQ line is becoming 0 in this condition.

-> PMIC case temperature is 86deg C constant throughout the test.

-> IMX6ULL case temperature is 87deg C constant throughout the test.

Questions:

1) Is there any condition where MPU turn off PMIC_POWERON_REQ line?

2) Is there any way PMIC goes in shutoff mode for a longer duration if the temperature cycle lasts ( i.e. 85 C for more than 16 hours)

3) Is there any temperature condition in which the processor stops working and shut itself.

4) There is PANIC temperature alarm in the IMX6ULL TEMPMON section, which resets the system.

      -> Will PMIC_POWER_ON_REQ PIN turned off when PANIC alarm condition generates Chip reset signal?

      -> How can we calculate time after which PANIC alarm resets the chip once the temperature reached to PANIC threshold?

Thanks,

Vivek

Labels (2)
0 Kudos
3 Replies

823 Views
gusarambula
NXP TechSupport
NXP TechSupport

Hello Vivek,

You can find more information on how the thermal driver works on the i.MX Linux Reference Manual included as part of the BSP documentation.

There shutoff state is dependent only on the temperature monitor and it won’t track how long the offending temperature was. If you are testing as soon as the temperature reaches 70°C it would be that on the longer heating scenario more heat is accumulated, and it may take a bit longer for the processor junction temperature to reach a workable level temperature. Have you checked how extra time it takes for the processor to turn on?

The panic interrupt temperature can be set on the TEMPMON_TEMPSENSE2n[PANIC_ALARM_VALUE] bits. You can find more details on the i.MX6ULL Reference Manual.

I hope this helps!

Regards,

0 Kudos

823 Views
vivekrajpara
Contributor III

Hi gusarambula,

Thanks for the reply, 

As mentioned earlier, we are keeping our i.MX6ULL based board at  85 °C constant for 16 hours. We observed that after some hours every supply gets down to 0 V i.e. SOC_IN and HIGH_IN go to 0V, VSNVS is still 3V and PMIC_POWERON_REQ goes low. Hence the PMIC goes to power off mode.

Did you have any insights on the below questions?

1) Is there any condition where MPU turns off PMIC_POWERON_REQ line?

2) Is there any way PMIC goes in shutoff mode for a longer duration if the temperature cycle lasts ( i.e. 85 C for more than 16 hours)

3) Is there any temperature condition in which the processor stops working and shut itself.

Regards,

Vivek

0 Kudos

823 Views
gusarambula
NXP TechSupport
NXP TechSupport

Hello Vivek,

There shutoff state is dependent only on the temperature monitor and it won’t track how long the offending temperature was. So regardless of how many hours the offending temperature was present the processor should be able to startup again as soon as the internal junction temperature is low enough.

There is indeed a threshold in which the processor will shut down to protect itself from damage due to the high temperature. This is the PANIC_ALARM_VALUE that is setup on the TEMPMON_TEMPSENSE2n register. The PMIC_POWERON_REQ signal would be changed accordingly.

I hope this helps!

Regards,

0 Kudos