32k_N XTAL - TXCO

ESP_Sprite
Posts: 9769
Joined: Thu Nov 26, 2015 4:08 am

Re: 32k_N XTAL - TXCO

Postby ESP_Sprite » Thu May 11, 2023 9:12 am

Can you set CONFIG_SLOW_CLK_CAL_CYCLES to 0? This should disable calibration and would make the ESP32 use the external clock input without trying to see if it works first. Then please make sure that the clock indeed keeps running at 1sec/1sec when in deep sleep to verify it actually works.

Lancsrick
Posts: 30
Joined: Mon Apr 10, 2023 5:48 pm

Re: 32k_N XTAL - TXCO

Postby Lancsrick » Thu May 11, 2023 9:20 am

Will do thanks. Is the best way to verify 1 sec/1sec to do a millis() output after 15min or so?

ESP_Sprite
Posts: 9769
Joined: Thu Nov 26, 2015 4:08 am

Re: 32k_N XTAL - TXCO

Postby ESP_Sprite » Fri May 12, 2023 6:53 am

No, because when the ESP32 runs as normal, the clock is derived from the 40MHz crystal. Best to print gettimeofday() output, then deep sleep for e.g. 15 minutes, then wake up and print it again. The deep sleep shuts down the 40MHz crystal making the ESP use the RTC clock for timekeeping.

Lancsrick
Posts: 30
Joined: Mon Apr 10, 2023 5:48 pm

Re: 32k_N XTAL - TXCO

Postby Lancsrick » Fri May 12, 2023 6:56 am

Thanks.

Lancsrick
Posts: 30
Joined: Mon Apr 10, 2023 5:48 pm

Re: 32k_N XTAL - TXCO

Postby Lancsrick » Tue May 16, 2023 9:23 am

Thank you for all the help finally got to the bottom of this. My pull up value was too high - once I changed to just 10k then things seem to be ok. That's a pull up to a 680k/220k voltage divider off 3.3v.

I don't get any calibration message on startup, but the serial monitor error message on not finding the external oscillator vanishes so I think I'm golden now.

Thanks again.

ESP_Sprite
Posts: 9769
Joined: Thu Nov 26, 2015 4:08 am

Re: 32k_N XTAL - TXCO

Postby ESP_Sprite » Wed May 17, 2023 7:50 am

Interesting - guess there's something with the rise time that trips up the ESP or something. Thanks for posting the solution!

Who is online

Users browsing this forum: No registered users and 92 guests