GPIO20 redundance (not be used)?

User avatar
urbanze
Posts: 176
Joined: Sat Jun 10, 2017 9:55 pm
Location: Brazil

GPIO20 redundance (not be used)?

Postby urbanze » Tue Jun 19, 2018 2:08 pm

Why software can't use GPIO 20? I need this last pin in my product.

WiFive
Posts: 1815
Joined: Tue Dec 01, 2015 7:35 am

Re: GPIO20 redundance (not be used)?

Postby WiFive » Tue Jun 19, 2018 4:56 pm

You plan to expose the die and wire bond to it?

User avatar
urbanze
Posts: 176
Joined: Sat Jun 10, 2017 9:55 pm
Location: Brazil

Re: GPIO20 redundance (not be used)?

Postby urbanze » Tue Jun 19, 2018 7:45 pm

WiFive wrote:You plan to expose the die and wire bond to it?


I want to change one pin from LED RGB to 20 (before in 15) to use analog function in 15. Just blink LED wired in top layer pcb. IDF (gpio.h) doenst support gpio_num_20 :(

User avatar
martinayotte
Posts: 108
Joined: Fri Nov 13, 2015 4:27 pm

Re: GPIO20 redundance (not be used)?

Postby martinayotte » Tue Jun 19, 2018 8:48 pm

As WiFive mentioned, GPIO20 isn't exposed on QFN package ...

User avatar
urbanze
Posts: 176
Joined: Sat Jun 10, 2017 9:55 pm
Location: Brazil

Re: GPIO20 redundance (not be used)?

Postby urbanze » Tue Jun 19, 2018 9:05 pm

martinayotte wrote:As WiFive mentioned, GPIO20 isn't exposed on QFN package ...



Why WROOM has this pin in package? just to fool me? LOL :lol:

:shock: One WROOM pin map in google, shows gpio 20 in wroom, but I looked another and say's "NC"...... Ok... But why gpio20 cannot be used? 6-11 is flash, 0/2/15/12(?) is boot, and 20?

User avatar
kolban
Posts: 1570
Joined: Mon Nov 16, 2015 4:43 pm
Location: Texas, USA

Re: GPIO20 redundance (not be used)?

Postby kolban » Wed Jun 20, 2018 3:50 am

My guess is that internally in the MCU, there is the logical notion of the existence of additional GPIOs such as GPIO 20. Now imagine you are the designed of the physical package that is the ESP32. You are now told that you can only have a finite set of pins surrounding its edge. You map the ones that you know you absolutely must have. By the time you are done, you have found that you still have logical functions that, in theory, could be exposed ... but in reality you have no where to expose them to as you have used up all the pins on the package for other functions. What are you to do? Perhaps you could make a physically bigger package? Or, maybe, as in this case, you simply don't expose them to the outside of the package. They become "un-usable". Do you continue to document their logical existence? I think yes. If internally the logical pins have numbers that don't change, then you would again notice that some are "missing".
Free book on ESP32 available here: https://leanpub.com/kolban-ESP32
Available for ESP32 consulting.

ESP_Sprite
Posts: 1663
Joined: Thu Nov 26, 2015 4:08 am

Re: GPIO20 redundance (not be used)?

Postby ESP_Sprite » Wed Jun 20, 2018 4:27 am

I think the pin was intended for the /CS line of a flash chip internal to the package. While it was never used as such, the PSRAM code does use it as an intermediate internal signal wire to delay a signal by one or two clock cycles.

WiFive
Posts: 1815
Joined: Tue Dec 01, 2015 7:35 am

Re: GPIO20 redundance (not be used)?

Postby WiFive » Wed Jun 20, 2018 5:54 am

ESP_Sprite wrote: the PSRAM code does use it as an intermediate internal signal wire to delay a signal by one or two clock cycles.


False :D https://github.com/espressif/esp-idf/co ... 6L521-L549

Who is online

Users browsing this forum: No registered users and 4 guests