pablozg / freeds

Derivador de energía solar excedente
GNU General Public License v3.0
100 stars 35 forks source link

Bug: "Failed to read term temperature from DS18B20 sensor" message and automatic control interruption #113

Open Csontikka opened 1 week ago

Csontikka commented 1 week ago

Bug: "Failed to read term temperature from DS18B20 sensor" message and automatic control interruption

Version: 1.1.0020 Beta Build: 05042024 Web: 1.1.5 Hardware: PWM10A 2200W MOSFET Trailing Edge Dimmer + ESP32 with additional cooling (relay4) Connected Temperature Sensor: DS18B20 (for measuring boiler temperature)

Description of the Issue: On 2024-10-28 at 16:33, the error message "Failed to read term temperature from DS18B20 sensor" appeared in the console (see attached image), which led to the interruption of automatic PWM control. PWM was enabled in AUTO mode. Despite this error, I continued to receive real-time boiler temperature data in Home Assistant, suggesting that the issue was transient; however, it still resulted in a control shutdown.

Approximately 2 hours later, at 18:49, I noticed the issue and saved the configuration through the device’s web interface, which I believe restarted the internal processes. Following this, the device resumed operation immediately and has been running without issues since.

Interesting Observation: While the DS18B20 error interrupted the automatic control, temperature data continued to update in Home Assistant. I suspect the error was temporary, yet it still blocked the system's operation. The device is equipped with additional cooling (relay4), so overheating could not have caused this issue. Since then, I verified the sensor connection, which is stable and functioning correctly.

Attached Images: The attached images display the error messages in the console and the data from the two relevant sensors (boiler temperature and PWM percentage values).

freeds_1 freeds_2

Request: I am looking for a solution that prevents a momentary sensor error from stopping the entire system. If a sensor occasionally gives a faulty reading, it should not lead to a complete control shutdown.

If you need any further information, please let me know!

Regards, Csontikka

Csontikka commented 1 day ago

The issue has recurred 9 days later. The device exhibited the exact same behavior as before, so I won't go into detail about the events. However, this suggests that the problem is not isolated but rather a recurring, systematic issue that requires further investigation.

Thank you in advance for taking the time to address this!

pablozg commented 1 day ago

Hola, no especificas la configuración que tienes en el apartado de temperaturas para tratar de replicarlo, aunque en mi opinión no es un fallo como crees, lo más probable (al no disponer de tu configuración) es que la temperatura del termo estuviera dentro del rango configurado (por encima de los 50 grados en tu gráfica) por lo que una vez detenido por el fallo de lectura del termo, este no se pone a funcionar hasta que la temperatura esté por debajo del umbral de activación (vuelvo a repetir que no has puesto una captura de tu configuración), por lo que una vez hubiese llegado a esta temperatura se habría activado de nuevo hasta llegar a la temperatura de corte.

Saludos.

Csontikka commented 2 hours ago

Saludos,

I have taken multiple screenshots of the settings that seem relevant, which I have not modified since. The boiler's heating range was set between 30 and 70 degrees, and the temperature never approached these values. However, the previously recorded boiler temperature graph clearly shows that the control did not stop when reaching a specific temperature. The temperature remained unchanged when the PWM turned off.

Regards, Csontikka

image image image image

Csontikka commented 1 hour ago

Dear Pablozg,

This issue continues to concern me, so I took a detailed look at the error logs from November 6. I noticed one additional detail that might provide a useful clue for you.

The graphs clearly show that at the moment the error occurred, the status of the "FreeDS ojle Error" sensor briefly changed to "Problem."

Additionally, just a few minutes before the error, at the same temperature, the PWM was still switching correctly.

I've also taken a screenshot related to the specific time of the issue, which shows the status of the sensors. I hope this visual evidence, along with the details provided, can assist in identifying the problem.

image

I tried to annotate the screenshot as thoroughly as possible to highlight the key points.

Regards, Csontikka