LS1088A shuts down during -40C operation

[Arxy]

We have a problem similar to https://community.nxp.com/t5/Layerscape/LS1088A-Thermal-problem-40-Degree/td-p/1265909

In our case, our design using LS1088AXE7Q1A has been working successfully at room temperature with ~12 boards for several months. However, when starting up and running at -40C, the LS1088A shuts down less than 30 minutes after startup with message:

  thermal thermal_zone0: core-cluster: critical temperature reached, shutting down

  reboot: HARDWARE PROTECTION shutdown (Temperature too high)

DDR and other temperature sensors on the PCB report ~-30C.

So far this has occurred on 2 of 3 units. Is there any way to fix this to ensure operation at the cold temperature?

[hanifakhairunisa]

If the LS1088A shuts down at -40°C, make sure it matches that temperature by checking its data sheet. Make sure the power supply is stable and sufficient, and all components have low temperatures. Manage thermal stress and avoid rapid temperature changes. Protect from condensation and moisture. Check that the firmware handles low temperatures correctly and check the reliability of all physical connections. Enable detailed diagnostics to determine the cause of the failure. you can contact me via whatsapp mod

[emma32]

We are encountering the same problem. The provided link for updating the device tree is not useful. This link sets the maximum temperature threshold for the unit. For our board, which utilizes the extended temperature range version of the LS1088ARDB we have adjusted this threshold to 105°C instead of the default 95°C. While the manual for the LS1088A specifies that the thermal management unit reads temperatures above 0°C the LS1088As rating suggests it operates down to -40°C.

Regards: servidor avanzado ff 

[yalani]

You need to use a third-party service to use Spotify Receiptify and get a receipt of your most played songs. Go to receiptify.herokuapp.com.

https://receiptifyus.com/

[nisafitri]

In our case, our design using the LS1088AXE7Q1A has been operating successfully at room temperature with around ~12 boards for several months. However, when powered on and running at -40C, the LS1088A shuts down less than 30 minutes after powering on with the message:

thermal thermal_zone0: core-cluster: reaches critical temperature, shuts down

reboot: HARDWARE PROTECTION shutdown (Temperature too high)

The DDR and other temperature sensors on the PCB report around ~-30C.
I've also read several articles on ngobrol games, they also found the same thing, so I haven't found a solution.

To date this has happened on 2 of the 3 units. Is there a way to fix this to ensure operation at cold temperatures?

[czynest]

In this scenario, it's advisable to adjust the alert and critical parameters within the device tree.

Important: Ensure that you do not surpass the predefined temperature parameters for the microprocessor.  Providing guidance on modifying device tree parameters with a specific caution about not exceeding temperature limits, emphasizing the importance of careful adjustments.

[rboileau]

@Hector_Villarruel See above about my request for more information and/or access to the link.

[Hector_Villarruel]

Hello @Arxy 

In this case you should modify the alert and critical parameters on the device tree.

Please note that you must not exceed the set temperature parameters of the microprocessor.

Best Regards,

Hector V

[rboileau]

Hi Hector,

I'm not able to access the link you provided above.
https://community.nxp.com/t5/iMX-and-Vybrid-Support/Regarding-asap-thermal-monitor-by-i-MX-8M-Mini/m...

Is there another way to get access to this information ?

Our design using LS1088AXE7Q1A has been working successfully at room temperature with ~12 boards for several months. However, when starting up and running at -40C, the LS1088A shuts down less than 30 minutes after startup with message:

thermal thermal_zone0: core-cluster: critical temperature reached, shutting down
reboot: HARDWARE PROTECTION shutdown (Temperature too high)

We can’t read the processor temperature directly since it doesn’t report <0°C (per the reference manual, but also verified). The DDR and other temperature sensors on the PCB report ~-30C, so the processor is obviously not too hot.
So far this has occurred on 2 of 4 units. I also see another question on the NXP forum from several years ago with a similar problem.
Is this a known issue with this processor? Is there a solution to have it work reliably at -40C°?



Roderic

[Hector_Villarruel]

Hello @Arxy 

As I mentioned on my previous reply you have to  modify the alert and critical parameters on the device tree.

Please note that you must not exceed the set temperature parameters of the microprocessor.

Link:

linux-imx/arch/arm64/boot/dts/freescale/fsl-ls1088a.dtsi at lf-6.1.y · nxp-imx/linux-imx · GitHub

Just in reference here is the PATH:

linux-imx/arch/arm64/boot/dts/freescale/fsl-ls1088a.dtsi

On Line 139.

BR,

Hector 

[shawnkarow]

We have this issue as well. The link you point at to update the device tree is not helpful. This sets the MAX temperature the unit can run at. We have updated that to 105 (instead of 95) for our board as we are using the extended temperature range version of the LS1088ARDB. The manual for the LS1088A says that the thermal management unit will only read temperatures above 0C but the rating on LS1088A use says to -40.

We see the same error running at -30C. The sensor for the temperature manger says invalid, as expected due to above note, over and over again but then suddenly says too high, shutting down. We have at least 10 other thermal sernsors in our unit all reading negative temperatures once this occurs. The polling period in the device tree for the temperature is set to 5 seconds in the area you point to (line 139) as well. It is hard to believe that the CPU in a period of 5 seconds suddenly goes from below 0 to over 105 in the space of 5 seconds when it is in a cooling chamber with all the other parts around it still reading -30C. This also does not happen on every unit we test.

[shawnkarow]

Follow up on continued testing. We have added some logging to the kernel driver for monitor of thermal and we have caught the error in action. We get the expected failed to read message as we are running at -40C and the chip is only supposed to read temperature between 0-110. However suddenly it goes from being unable to read to reading a temperature that is too hot. Any advice on how to mitigate this situation would be appreciated.

[ 518.895108] thermal thermal_zone0: failed to read out thermal zone (-61)
[ 519.887110] thermal thermal_zone1: failed to read out thermal zone (-61)
[ 524.172576] thermal thermal_zone0: failed to read out thermal zone (-61)
[ 525.007254] thermal thermal_zone1: failed to read out thermal zone (-61)
[ 528.395107] thermal thermal_zone0: failed to read out thermal zone (-61)
[ 530.124835] thermal thermal_zone1: failed to read out thermal zone (-61)
[ 530.465083] thermal thermal_zone0: trip_type == THERMAL_TRIP_CRITICAL, we read temperature (117000)
[ 530.474373] thermal thermal_zone0: core-cluster: critical temperature reached, shutting down
[ 530.483114] reboot: HARDWARE PROTECTION shutdown (Temperature too high)
[ 531.521064] thermal thermal_zone0: trip_type == THERMAL_TRIP_CRITICAL, we read temperature (111000)
[ 531.530168] thermal thermal_zone0: core-cluster: critical temperature reached, shutting down
[ 531.538640] reboot: HARDWARE PROTECTION shutdown (Temperature too high)

[Hector_Villarruel]

We consider that we need you to provide us with a picture of your processor, the one that is working properly and the one that is not, which is presenting this situation.

In order not to compromise your information,
Could you please send them to us by creating a technical support case?

Also, when creating a case, please let us know that this situation is linked to the following number: 00588182.

You can create a new ticket using the technical support web located in the following link:

https://support.nxp.com/s/

Best Regards,

Hector Villarruel