Hi Xabi, what do you mean? Are you talking about controlling outputs while the ESP32 is in Deep Sleep mode?

Sorry Rui, I sent the question before writing it. I edited it 🙂

If you need more information, please let me know

Hi Xabi,
Sorry for taking so long to get back to you, but I’ve been receiving a high volume of emails and I wasn’t able to keep up with everything.

I would also like to know how to do that. According to the documentation it should be very easy… You would call this function to hold a state during deep sleep:

esp_err_t rtc_gpio_hold_en(GPIO_NUM_XX);

And you would call this function to disable/unlock states after a deep sleep (so, you could change the pin value):

esp_err_trtc_gpio_hold_dis(gpio_num_tgpio_num)

According to a source on the ESP32 Forum: http://bbs.esp32.com/viewtopic.php?t=4040

Something like this code should keep GPIO 14 HIGH during deep sleep:

void setup() { 
  pinMode(14, OUTPUT); 
  digitalWrite(14, HIGH); 
  esp_err_t rtc_gpio_hold_en(GPIO_NUM_14); 
  delay(1000); 
  esp_deep_sleep(10000000000); 
}

However, I can’t make it work for me… I’ve also found this repository with some examples, but using the ESP IDF: https://github.com/krzychb/ulp-loop.

I’ve tried to run these commands on Arduino IDE, but it also didn’t work for me:

esp_err_t rtc_gpio_init(gpio_num_t GPIO_NUM_14);
esp_err_t rtc_gpio_set_direction(gpio_num_t GPIO_NUM_14, gpio_mode_t RTC_GPIO_MODE_OUTPUT_ONLY);
esp_err_t rtc_gpio_set_level(gpio_num_t GPIO_NUM_14, uint32_t t=1);
esp_err_t rtc_gpio_hold_en(gpio_num_t GPIO_NUM_14);
esp_deep_sleep(10000000000);

Unfortunately I don’t have a working example for you… But I’m sure something small is missing to make it work.

I hope that helps. Regards,
Rui

Hi Rui, 
Thanks for the information! I tried it but I still couldn’t make it work. Did you make it work? Or do you know anyone that made it work? 
Thank you very much for your support 🙂
Xabi

I didn’t have the time to make it work either. In theory, it should work, but I’m sure it’s something lower level that is missing that doesn’t allow it to make it work.

Regards,
Rui

Hi Rui,
Good news here! I made it work! Here you have the simple code that worked for me:

#include "driver/rtc_io.h"
gpio_num_t pin_MOSFET = GPIO_NUM_15;
void setup() {
rtc_gpio_init(pin_MOSFET);
rtc_gpio_set_direction(pin_MOSFET,RTC_GPIO_MODE_OUTPUT_ONLY);
rtc_gpio_set_level(pin_MOSFET,0); //GPIO LOW
delay(5000);
rtc_gpio_set_level(pin_MOSFET,1); //GPIO HIGH
esp_sleep_pd_config(ESP_PD_DOMAIN_RTC_PERIPH, ESP_PD_OPTION_ON);
esp_sleep_enable_timer_wakeup(60*1000*1000);
esp_deep_sleep_start();
}
void loop() {
// put your main code here, to run repeatedly:
}

On the other hand, I’m having problems with my transistor. The BS170 is not giving enough current for a Vgs of 3.3V. I made a little research over the internet and, for micros like ESP32 which give 3.3V through their outputs, it’s needed MOSFETScalled logic level MOSFET (work for Vgs=3.3V). Could you give me a good reference of a logic level MOSFET that would work well with the ESP32 and is going to be used to control a sensor that needs up to 150mA?
I hope you find that code useful 🙂

Regards,
Xabi

Thanks for sharing your code and I’m glad you made it work. I’ll definitely give it a try.

Have you tried the 2N7000? I often use that N−Channel MOSFET with the ESP8266.

That’s exactly the one I ordered from Aliexpress to try! I’m still waiting for it, but from it’s datasheet curves I’m not sure if it will give me 150mA for 3.3V in Vgs. At least you confirmed I’m not in the wrong direction 😉
I’ll close this question as resolved.
Thanks for the support Rui.

Regards,
Xabi

I’m also not 100% sure… so you’ll need to test, but I think it should work for your needs. Thanks. Regards,
Rui

I have just tried the 2N7000 to switch a red led (in line with a 70 ohm resistor) using a 3.3v power source and it only worked after I had to put a 10K ohm pull down resistor on the gate. A very strange result occurred that perhaps someone can explain was that every time I touched the gate with a wire of any sort it turned on the transistor switch, this wire was not connected to a power source. I did make sure the transistor was the right way around, I found that if you switch drain with source it just powers it on permanently like reversing a diode.  My only interpretation (I am a novice) was that the floating voltage in the wire was activating the transistor, but when I used a 10K resistor as a pull down all worked fine. The 2N7000 does have a min gate activation voltage of 0.8, so perhaps the floating voltage was enough to activate it???. Any one else had this occur to them or do they have an sound explanation?