While I play with STM32 a lot on the software side, I still miss a lot of hardware knowledge so I want to confirm my thinking before making a test PCB. I use STM32L4R5 GPIO to control LM3800 power sequencer. GPIO pin is directly connected to LM3880 EN pin (active high). To guarantee that LM3880 is not enabled by a glitch during STM32 power up, I want to put external pull-down resistor to guarantee it is set to low. Of course, once the STM32 is on, I want to enable LM3880 and it will be mostly high all of the time. Now, regarding the value.
- STM32 datasheet says internal pull-up (I guess this is what i have to look for regarding high state later) is 40kOhm (25-55 min/max)
- LM3880 datasheet says EN voltage should be over 1.4V max (1V min)
My understanding is that I have to consider the voltage divider that will happen between internal pull up and my external pull down. Choosing 56K resistor, gives me 1.66V to 2.28V at EN pin when I set GPIO pin high (I use 3.3V and the divider min max). Before high is set, pull down will give me the required low on EN.
Is my thinking correct or I miss something?
EDIT: I received excellent comments and understood my wrong thinking about the pull down resistor concept. Reconsidering the full power design, I decided to update the question with the following:
The STM32 has to control two DC/DC convertors. It is easily done by software, but I need external pull down resistors to guarantee both chips are off during STM32 power up (takes a long time due to huge arrays) before it takes control of the GPIO pins. My thinking is:
Chip 1, TPS54202, has EN level of 1.2V, powered by 12V, output 5V. Has internal pull up current source of 0.7 uA. Following the previous discussion, 0.7uA would mean (12V - 1.2V) / 0.7 uA or 15.4MOhm equivalent resistance. I would need an external resistor, for example 100K and this will give me around 0.08 V at EN when STM32 is still not initialized. After STM32 is ON, I will use the GPIO pin as open drain and drive it low/high when necessary.
Chip 2, LP2985, 12V to 10V LDO, active high, no internal current source. So, I guess it is ok just to put the same pull down resistor 100K just to guarantee the chip is not enabled until STM32 is powered on and takes control of the GPIO.
The low current through resistors should be ok for GPIO to drive them easily.