Add support for 32k XTAL as RTC_SLOW_CLK source

- RTC_CNTL_SLOWCLK_FREQ define is removed; rtc_clk_slow_freq_get_hz
  function can be used instead to get an approximate RTC_SLOW_CLK
  frequency

- Clock calibration is performed at startup. The value is saved and used
  for timekeeping and when entering deep sleep.

- When using the 32k XTAL, startup code will wait for the oscillator to
  start up. This can be possibly optimized by starting a separate task
  to wait for oscillator startup, and performing clock switch in that
  task.

- Fix a bug that 32k XTAL would be disabled in rtc_clk_init.

- Fix a rounding error in rtc_clk_cal, which caused systematic frequency
  error.

- Fix an overflow bug which caused rtc_clk_cal to timeout early if the
  slow_clk_cycles argument would exceed certain value

- Improve 32k XTAL oscillator startup time by introducing bootstrapping
  code, which uses internal pullup/pulldown resistors on 32K_N/32K_P
  pins to set better initial conditions for the oscillator.
This commit is contained in:
Ivan Grokhotkov
2017-04-24 18:36:47 +08:00
parent 8131c77860
commit 6353bc40d7
16 changed files with 330 additions and 136 deletions

View File

@@ -27,6 +27,7 @@
#include "esp32/ulp.h"
#include "soc/soc.h"
#include "soc/rtc.h"
#include "soc/rtc_cntl_reg.h"
#include "soc/sens_reg.h"
#include "driver/rtc_io.h"
@@ -321,15 +322,11 @@ TEST_CASE("ulp timer setting", "[ulp]")
uint32_t counter = RTC_SLOW_MEM[offset] & 0xffff;
CLEAR_PERI_REG_MASK(RTC_CNTL_STATE0_REG, RTC_CNTL_ULP_CP_SLP_TIMER_EN);
// compare the actual and expected numbers of iterations of ULP program
float expected_period = (cycles_to_test[i] + 16) / (float) RTC_CNTL_SLOWCLK_FREQ + 5 / 8e6f;
float expected_period = (cycles_to_test[i] + 16) / (float) rtc_clk_slow_freq_get_hz() + 5 / 8e6f;
float error = 1.0f - counter * expected_period;
printf("%u\t%u\t%.01f\t%.04f\n", cycles_to_test[i], counter, 1.0f / expected_period, error);
// Should be within 15%
TEST_ASSERT_INT_WITHIN(15, 0, (int) error * 100);
// Note: currently RTC_CNTL_SLOWCLK_FREQ is ballpark value — we need to determine it
// Precisely by running calibration similar to the one done in deep sleep.
// This may cause the test to fail on some chips which have the slow clock frequency
// way off.
}
}