From 8ee657f0e5969ea874eecaef494bcab71936f864 Mon Sep 17 00:00:00 2001 From: Marian Buschsieweke Date: Tue, 14 Dec 2021 22:46:33 +0100 Subject: [PATCH] cpu/nrf5x_common/periph_temperature: Increase resolution MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Modify the periph_temperature implementation to expose the full resolution of the temperature sensor. The accuracy of the sensor will likely be less than 0.25 °C, but typically temperature sensors have a precision in the order of their resolution. Exposing the full resolution can therefore be useful to monitor relative temperature changes. --- cpu/nrf5x_common/periph/temperature.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/cpu/nrf5x_common/periph/temperature.c b/cpu/nrf5x_common/periph/temperature.c index b7547ef056..37805dc83b 100644 --- a/cpu/nrf5x_common/periph/temperature.c +++ b/cpu/nrf5x_common/periph/temperature.c @@ -24,7 +24,11 @@ #include "saul_reg.h" #include "phydat.h" -void temperature_read(int16_t *temp) +/** + * @brief Read the temperature in E-02 °C + * @return The measured temperature in E-02 °C (e.g. 1825 would mean 18.25 °C) + */ +static int16_t temperature_read(void) { /* Start temperature measurement task */ NRF_TEMP->TASKS_START = 1; @@ -32,22 +36,24 @@ void temperature_read(int16_t *temp) /* Wait for temperature measurement to be ready */ while (!NRF_TEMP->EVENTS_DATARDY); /* takes 36us according to manual */ - /* temperature is in 0.25°C step, so just divide by 4 */ - *temp = (int16_t)NRF_TEMP->TEMP >> 2; + /* temperature is in 0.25°C step, multiply by 25 to convert to E-02 °C */ + int32_t temp = NRF_TEMP->TEMP * 25; /* Clear data ready bit and stop temperature measurement task */ NRF_TEMP->EVENTS_DATARDY = 0; NRF_TEMP->TASKS_STOP = 1; + + return temp; } static int _read_temperature(const void *dev, phydat_t *res) { - (void) dev; - temperature_read(&res->val[0]); + (void)dev; + res->val[0] = temperature_read(); res->val[1] = 0; res->val[2] = 0; res->unit = UNIT_TEMP_C; - res->scale = 0; + res->scale = -2; return 1; }