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.

TPS659038/9 RANGE bit change on the fly

Hello,

My customer would like to know what would happen if the RANGE bit of SMPSx_VOLTAGE register changes on the fly.  If I look at the register map, I see the following sentences.

Note: RANGE cannot be changed on the fly when the SMPS is Active.  To change the operation voltage range, SMPS has to be disabled.

Note: For Dual-phase and triple-phase modes, RANGE=1(1V to 3.3V) is not supported.

The customer did a try to confirm if this is true, but they saw strange behavior when they set RANGE=1 to SMPS12.  The SMPS12 output rose up to 2.27V and went down to 0.0V after that.

Questions

=========

1. Would you please tell me what is the correct behavior when RANGE bit of SMPS12_VOLTAGE changes on the fly?  If the register map is correct, the RANGE=1 setting should be ignored, so the output voltage doesn't change.

2. If the description of the register map is correct, what would be the root cause of the behavior my customer saw?

Best regards,

 Yoshikazu Kawasaki

  • Kawasaki-san,

    the RANGE-bit MUST NOT be changed while the SMPS is active. Please have your customer to disable it (in case of EVM via the GUI, SMPSxx_CTRL register, then change the range (and voltage), and re-enbale it only after that.

    If the RANGE is changed during active mode, the device may get (permanently) damaged!

    BR,

    Frank

  • Sorry, I missed to comment on your 2nd item: Indeed, SMPSs operated in Multi-phase must not be operated in RANGE=1 at all (potential damage to the device may occur, if done anyhow), so above proceedings apply to SMPS-rails operated in single phase mode only, while your example of SMPS12 would be multiphase and therefore must not ever be set into RANGE=1.

    Please note, changing the range is a means for evaluation. In an actual system, the voltages will/should not change significantly (only e.g. to support AVS or DVFS), besides, there is an overlap between the range-regions (Range=0: 0.7V ... 1.65V, Range=1: 1V ...3.3V), so one should be able to select the appropriate range before power-up and therefore no need to change it during operation, even in case of a voltag change.

    BR,

    Frank

  • Hello Frank,

    Thank you very much for your quick reply.

    It means their observation was expected and the register map shows the wrong information because it says "cannot be changed".  Would you please correct the register map description?

    The customer cannot avoid the RANGE=1 when the SW doesn't operate correctly since any settings could happen.  In this case, how the customer can do to avoid unexpected output voltage which may damage the device?  The reason they did a try is to think about the case when the S/W doesn't operate correctly.  It could happen and PMIC should protect that if the RANGE=1 "can be changed".

    Best regards,

     Yoshikazu Kawasaki

  • Kawaski-san,

    the PMIC does not block write access to (voltage-)registers. Consequently, customer must ensure SW-integrity. However, even if the range-bit would be blocked (which it isn't), there is still a 3x voltage change possible just by changing the voltage-bits (e.g. 1V to 3.3V) which would most likely damage downstream hardware. One means to support a sanity-check is the watchdog (per default disabled, but could be enabled by software), which could be configured to create  an interupt and power-down the PMIC.

    BR,

    Frank

  • Hello Frank,

    Thank you very much for your quick reply.  I informed what you told me to the customer and they understood that, but they'd like to correct the datasheet and register map for RANGE bit in order not to confuse customers.  I asked that to the owner of the datasheet and register map, so I have no more question on this ticket.

    Best regards,

     Yoshikazu Kawasaki