Delay to wait until strapping pin works as normal-function pins?

[michcfr]

Hello,

The ESP32 Datasheet (§2.4 Strapping Pins) states that after reset, the strapping pins work as normal-function pins.
What is the delay by which these pins work as normal-function pins?

I need to know this delay because I have connected the strapping pins of the ESP32 board to another host MCU GPIOs. So, the host MCU need to wait this delay to not interfere with the ESP32 during its boot time.


Regards,
Michel

[ESP_Sprite]

That's in the datasheet, page 14.

[michcfr]

That's in the datasheet, page 14.

thank you.
So the delay is, at least 1ms, after the CHIP_PU signal goes high?


regards,
Michel

[ESP_Sprite]

Yes, with 'goes high' defined here as >0.6V.

[PeterR]

@ESP_Sprite

So the delay is, at least 1ms, after the CHIP_PU signal goes high?

Are strapping pins only checked on hard power up/reset or are strapping pins also checked on CPU restart?
What is the ESP32 crashes and reboots? Will the strapping pins again be checked again during the first 1mS?

EDIT: Similarly watchdog & ESP brown out etc.

[ESP_Sprite]

From the datasheet: 'During the chip’s system reset release (power-on-reset, RTC watchdog reset and brownout reset), the latches of the strapping pins sample the voltage level as strapping bits of ”0” or ”1”, and hold these bits until the chip is powered down or shut down.' So not on a CPU restart, crash&reboot only if the reboot is caused by the RTC WDT and it's configured to do a system reset, yes on brownout.

[PeterR]

Thanks!

@michcfr
So you need to be a little careful. You cannot (easily) use the RTC Watchdog to reset. You would also need to be careful with your brown out thresholds. Your second MCU should brown out & go down (& stop intefering with the strapping pins) before the ESP. A voltage supervisor would ensure both CPUs are reset and is best practice - especially if you have other logic devices.

[michcfr]

Thanks!

@michcfr
So you need to be a little careful. You cannot (easily) use the RTC Watchdog to reset. You would also need to be careful with your brown out thresholds. Your second MCU should brown out & go down (& stop intefering with the strapping pins) before the ESP. A voltage supervisor would ensure both CPUs are reset and is best practice - especially if you have other logic devices.

Thank you guys!
It is even more tricky than expected...
To me more specific on my request: the strapping pin IO15 of the ESP32 is connected to the second MCU and there is a pulldown resistor on the line. So, I need the IO15 to be HIGH during this period. Any suggestion of the schematics?

thx

[PeterR]

What are you using IO15 for?
Why do you have a pull down on IO15? Why not a pull up?
Why use IO15?

Its not going to be easy to turn a pull down into a pull up EDIT on the fly!
The only thought which springs to mind is to have the second MCU control the ESP boot. The MCU could then set IO15 high before doing whatever it normally does with IO15.

[michcfr]

What are you using IO15 for?
Why do you have a pull down on IO15? Why not a pull up?
Why use IO15?

Its not going to be easy to turn a pull down into a pull up EDIT on the fly!
The only thought which springs to mind is to have the second MCU control the ESP boot. The MCU could then set IO15 high before doing whatever it normally does with IO15.

IO15 is used to do software bit bang in order to have single wire serial communication between the ESP32 and the second MCU
It is pull down because I have been advised to add 8kΩ-5MΩ pull-down resistor to reduce externally induced interference. The distance are very short between the ESP32 and the second MCU so maybe it is useless...and it will solve my issue.
IO15 is the only remaining pin on my PCB. Maybe I should reorganize it.