This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AM3352: Temperature difference with AM3357

Part Number: AM3352
Other Parts Discussed in Thread: AM3357, , TPS65261, TPS65251

Hi Team.

 

There is the temperature difference of 8.4degC between AM3352 and AM3357 on the customer board at the following condition.

Ambient temperature:72degC

AM3352 : 78.7degC

AM3357 : 87.1degC

 

Do you have any idea for this at same peripheral usage and frequency?

Do you have any idea we have to check?

Is this reasonable difference?

 

I use PET for AM335x. But I couldn’t get any big Power (W) difference to increase Tjunction.

 

Would you please give me your advice?

 

Thanks and Best regards,

Kuerbis

  • Hi Team.

     

    Here is the additional information. They are using a different PMIC as below.

    AM3352 with TPS65261

    AM3357 with TPS65251

     

    Thanks and Best regards,

    Kuerbis

  • Kuerbis,

    A few questions:

    1. Are these the same board designs or different board designs?  Mutual heating from other components may cause the processor to have a different temperature.  Similarly, board size, shape, construction, orientation and density all impact thermal behavior.
    2. Are they running the same software or different software?  Software execution patterns can have a large impact.  Also, clock settings can habve a significant impact.
    3. Have you conducted this experiment on single boards and providing readings or have you done this on a group, like 10 boards, and provided averages?  Process strength variations between chips with the same part number can result in noticeable power dissipation levels.
    4. Do these parts have the same package?
    5. Do these parts have the same temperature range?

    Notably, the AM3357 has additional, active PRU/ICSS blocks.  These may also add to power consumption, even if the software is not using them.

    Tom

  • Hi Tom

     

    Thank you for your reply.

    The customer is trying to replace AM3357 & TPS65251 with AM3352 & TPS65261 in the same product. So I want to confirm the reason because the temperature and the current (Current consumption difference is only 6mA) was different.

     

    Here is the reply for your questions.

    1. different board designs due to PMIC footprint is different.

    2. the same software.

    3. Only one board.

    4. The same package.

    5. The same temperature range. -40 to 105

     

    Thanks and Best regards,

    Kuerbis

  • Kuerbis,

    Device operating temperature is increased due to either increased power consumption or reduced thermal dissipation.  If you claim to see higher operating temperature and the power consumption is the same, then you either have a new board that dissipates heat less or you have other devices on the board that are generating more heat.

    Drawing conclusions from a single board and processor can be misleading.  To have confidence in your results, you need to make measurements on several boards of each type.

    You mention current consumption was nearly same - only 6ma difference.  Please provide the steady state current and voltage for each of the primary power supplies.  You need to calculate the total power consumption - which is the same as your thermal dissipation.

    Tom

  • Kuerbis,

    To add to Tom's inputs. Please look at table Table 3-1 in http://www.ti.com/lit/ds/symlink/am3359.pdf.

    AM3352 has a top frequency of 1GHz as compared to am3357 which has highest frequency of 800MHz.

    Higher frequency will imply higher voltage/higher OPP.


    So if your cpu is running at the top frequency on AM3352(1 GHz) then definitely power consumption
    will be higher hence the can expect higher temperatures.

    Regards,
    Keerthy

  • Kuerbis,

    Is there and further action needed on this discussion?  If there is no activity in the coming week, it will be closed.

    Tom