Troubleshooting

The controller's entities are missing / unavailable

• Is the controller added as an ESPHome device in HA (not merely flashed)? See First controller.
• Is the controller online? The hardware overview shows online/offline plus the firmware state.

Outputs/inputs are not assigned

• Panel → General → Match entities for that controller.
• Names must be unique per controller. Umlauts and ß are folded to ASCII (ä→ae, ö→oe, ü→ue, ß→ss).

"Firmware drift" warning

The settings differ from the flashed firmware. This is only a warning — runs keep working. Reflash the controller to clear the drift.

A run is logged as emergency

The on-device emergency shutdown fired. The most common cause: the run duration ends too close to emergency_shutdown_min. Reduce the run time to ≤ emergency_shutdown_min − 1 min (the panel warns about this).

The cistern level is wrong

Check the calibration — the support points and the current raw value.

The cistern level jitters at rest

That is normal — a pressure sensor is noisy. Smoothing on the input (tab Hardware) and a dead band on the source calm the display down; both are explained under Water sources.

GardenControl: the pump switches the wrong relay terminal

On the GardenControl board the two relay outputs are crossed: the screw terminal printed R1 is driven internally by a different pin than the name in the manufacturer's firmware suggests (confirmed by the manufacturer — the printed label is what counts). GardenESP takes this into account from version 0.13.73.

After updating to 0.13.73 you have to reflash the controller — until then it still switches the other terminal. After the flash, check once which terminal actually pulls in: if it is the wrong one, move the wire to the other relay terminal or change the output in the settings from R1 to R2 (or vice versa).

OTA flash fails: "Error resolving IP address … Is it connected to WiFi?"

This message appears when the ESPHome device name (the mDNS hostname gardenesp-steuergeraet-<short code>) has changed — for example because you changed a controller's short code. ESPHome then wants to send the new firmware to the new name (…-<new>.local), which does not exist on the network yet: the running controller still announces itself under its previous name. A chicken-and-egg problem — the new name only comes into being through the flash.

Solution — a one-off transitional flash to the current address:

• Flash over USB (which bypasses name resolution entirely), or

• temporarily specify the still-valid address under wifi: in the YAML:

  wifi:
    ssid: !secret wifi_ssid
    password: !secret wifi_password
    use_address: <previous-name>.local   # the controller's previously valid name, or its fixed IP
    ap:
      ssid: "${friendly_name} Fallback"
      password: !secret wifi_ap_password
  

  Then Install → Wirelessly: ESPHome connects to the old address and flashes the firmware carrying the new name. After the restart the controller is reachable under the new name → remove use_address again.

Because the mDNS name has changed, the controller may show up in HA under Discovered as a "new" device → add it again once; you can remove the old device entry afterwards. Your settings, lines and history are preserved (they do not hang off the device name).

"Invalid encryption key" when adding the device (even though the key is correct)

A message like Invalid encryption key: received_name=gardenesp-steuergeraet-a, received_mac=…. First the ordinary case: the key you enter must match exactly the one the device was flashed with (api_encryption_key from the ESPHome secrets.yaml at the time of the flash). If you changed the key afterwards → reflash and enter the same value.

A common trap with several instances: check the IP / the received_mac in the message. If the IP does not match your device, HA has connected to a different device with the same name (an mDNS collision — two controllers with the same short code on the same network, see First controller). Then it is not the key that is wrong but the target: give the controllers different short codes (unique across the network) and reflash — or add the device temporarily via its fixed IP (Add integration → ESPHome → host = IP).

Further help

Please open an issue — with your HA version, your GardenESP version and (if possible) the relevant log excerpts.