LMBCTLEN, LMAUTOEN, LMEN, LMSTAT Relation Hello NXP Support, - Is there any conflict if we enable LMAUTOEN, LMEN at the same time as in the following figure? - What is suggested configuration? Manual mode (via LMEN) or automatic mode (via LMAUTOEN)? - In which scenario, we have to use manual mode since there is already automatic mode? Thanks. BR/Kumamon Re: LMBCTLEN, LMAUTOEN, LMEN, LMSTAT Relation Hello @kumamon,
1. No conflict, actually, you can find a relevant note inside the Power Management chapter from the S32K3XX Reference Manual:
2. & 3. Automatic last mile auto-enable feature allows you to use the last-mile regulator (instead of the boot regulator) during chip startup and standby mode recovery.
In "manual mode", LMEN must be enabled before PLL is enabled. After Standby mode exit, you must re-enable the PMC last-mile regulator before transitioning to faster clock frequencies.
You can refer to Figure 180. Last-mile regulator enable sequence:
Regarding LMBCTLEN, it must be set to 1 if external BJT between VDD_HV_A and V15 is used on the PCB. For example, in S32K3X4EVB, if you want to use the NPN, set LMBCTLEN = 1.
Best regards, Julián Re: LMBCTLEN, LMAUTOEN, LMEN, LMSTAT Relation Hello @Julián_AragónM , Thanks a lot for your reply. You must write 1 to PMC.CONFIG[LM_EN] field before transtioning to faster clock frequencies irrespective of the setting of PMC.CONFIG[LMAUTOEN] field. This is because of the reduced clock speed when using the boot regulator. 1. What is the reason behind the above statement you mentioned, because I think automatic mode can do it by chip itself, why still need software to interfere in this case? And is the following understanding correct? Only automatic mode: not ok. Only manual mode: ok. Manual mode + automatic mode: ok. 2. It looks like "Manual mode + automatic mode" is always better than "Only manual mode"? In which scenario, automatic mode is not expected? Is it the scenario where we just strictly want to use boot regulator in chip startup time (for example)? 3. In legacy project, we even did not turn on last mile regulator when we use higher clock (160 MHz) than FIRC. We now know it violates the statement in RM. And we are estimating the risk, do you know what will be worst result here? Thanks. BR/Kumamon Re: LMBCTLEN, LMAUTOEN, LMEN, LMSTAT Relation Hello @kumamon,
Let me confirm the behavior with the internal team, in order to provide a consolidated answer regarding LMAUTOEN + LMEN functionality.
Regarding question 3: If PLL is enabled but the last-mile regulator is not enabled, the chip will be powered by the boot regulator. Under PLL clock, the core demands a large current and heat will accumulate on the boot regulator. This will overheat the chip until it is damaged.
Following the reference manual is mandatory: "As long as LMEN=0 software must make sure that the system clock is on FIRC clock or slower. To use higher clock speed software must set LMEN=1."
I will provide a response for Q1. & Q2. as soon as I get confirmation from the internal team.
Best regards, Julián Re: LMBCTLEN, LMAUTOEN, LMEN, LMSTAT Relation Hello @Julián_AragónM , No problem, thanks in advance. BR/Kumamon.
View full article