Homebrew by Weight - La Marzocco Linea Micra

01 - Homebrew By Weight.md

This is my 'homebrew' version of the brew-by-weight functionality that comes with a Linea Mini plus the LaMarzocco + Acaia proprietary scales. Given I have a Micra, this functionality isn't supported out of the box.

It uses:

• ESPHome to connect to my Acaia Lunar scales over Bluetooth using a forked version of https://github.com/open-lcc/esphome-acaia-component (my forked version fixes the flow calculation)
• The LaMarzocco custom HomeAssistant integration: https://github.com/zweckj/lamarzocco

How it works:

• input_number.coffee_target_weight: used to set the desired weight
• input_number.coffee_drip_duration: used to estimate how long (in secs) the coffee continues to drip after the pump is stopped.

1. Turn on the scales and wait for the connected symbol to show on the scales (next to the timer).
2. Start the shot. Homeassistant will tare the scales automatically and start the timer on the scales.
3. Once the shot timer reaches 10 seconds, the brew-by-weight automation should trigger
4. The automation continues to monitor the current weight and flow rate from the scales, continually updating the coffee_estimated_drip_weight template sensor.
5. Once the current weight reaches the threshold of (current_weight + coffee_estimated_drip_weight >= target_weight) it turns off the machine (only way to stop the pump) and turns it back on.

Adjust the input_number.coffee_drip_duration over time to get more accurate.

02-esphome.yaml

external_components:
  - source:
      type: git
      url: https://github.com/adampetrovic/esphome-acaia-component
      ref: log_flow
    refresh: 0s

esphome:
  name: coffee-controller
  friendly_name: coffee-controller

esp32:
  board: esp32-s3-devkitc-1
  framework:
    type: arduino

logger:
  level: DEBUG

api:

ota:

wifi:
  ssid: !secret wifi_ssid
  password: !secret wifi_password

  ap:
    ssid: "Coffee-Controller"
    password: "password"

bluetooth_proxy:
  active: True

esp32_ble_tracker:

ble_client:
  - mac_address: <mac address of your scales>
    id: acaia_client

acaia:
  ble_client_id: acaia_client
  id: acaia_1


binary_sensor:
  - platform: acaia
    connected:
      name: Acaia Connected
      id: acaia_connected

sensor:
  - platform: acaia
    weight:
      id: scale_sens
      name: Scale weight
      accuracy_decimals: 1
      unit_of_measurement: g
      filters:
        - lambda: return x * 1000;
    flow:
      name: Scale flow
      id: flow_sens

button:
  - platform: template
    name: Tare Scales
    id: scale_tare
    icon: "mdi:scale"
    on_press:
      - acaia.tare: acaia_1
  - platform: template
    name: Start Timer
    id: start_timer
    icon: "mdi:timer-outline"
    on_press:
      - acaia.start_timer: acaia_1
  - platform: template
    name: Stop Timer
    id: stop_timer
    icon: "mdi:timer-outline"
    on_press:
      - acaia.stop_timer: acaia_1
  - platform: template
    name: Reset Timer
    id: reset_timer
    icon: "mdi:timer-outline"
    on_press:
      - acaia.reset_timer: acaia_1

03-automations.yaml

input_number:
  coffee_target_weight:
    name: Coffee Target Weight
    unit_of_measurement: g
    initial: 42
    min: 10
    max: 60
    step: 1

  coffee_drip_duration:
    name: Coffee Drip Duration
    unit_of_measurement: s
    initial: 2.5
    min: 0
    max: 10
    step: 0.1

sensor:
  - platform: template
    sensors:
      coffee_estimated_drip_weight:
        unique_id: coffee_estimated_drip_weight
        friendly_name: Coffee Post-Shot Drip Weight
        unit_of_measurement: "g"
        availability_template: "{{ states('sensor.coffee_controller_scale_flow') | float > 0 }}"
        value_template: >
          {{ (states('input_number.coffee_drip_duration') | float * states('sensor.coffee_controller_scale_flow') | float) | round(2) }}

automation:
  - id: start_shot_setup_scales
    alias: "[Coffee] Setup scales on shot start"
    mode: single
    trigger:
      - platform: state
        entity_id: binary_sensor.micra_brewing_active
        to: "on"
    condition:
      - "{{ is_state('binary_sensor.coffee_controller_acaia_connected', 'on') }}"
    action:
      - service: button.press
        target:
          entity_id: button.coffee_controller_reset_timer
      - service: button.press
        target:
          entity_id: button.coffee_controller_start_timer
      - service: button.press
        target:
          entity_id: button.coffee_controller_tare_scales

  - id: stop_shot_setup_scales
    alias: "[Coffee] Stop scales on target weight"
    mode: single
    trigger:
      - platform: template
        value_template: "{{ states('sensor.coffee_controller_scale_weight') | float >= states('input_number.coffee_target_weight') | float }}"
    condition:
      - "{{ states('sensor.micra_shot_timer') | float >= 10 }}"
    action:
      - service: button.press
        target:
          entity_id: button.coffee_controller_stop_timer

  - id: brew_by_weight
    alias: "[Coffee] Brew by Weight"
    mode: single
    trigger:
      # make sure we're actually brewing a valid shot
      - platform: template
        value_template: "{{ states('sensor.micra_shot_timer') | float >= 10 }}"
    condition:
      - "{{ is_state('binary_sensor.coffee_controller_acaia_connected', 'on') }}"
    action:
      - wait_template: "{{ states('sensor.coffee_controller_scale_flow') | float > 0 }}"
        timeout: 
          seconds: 5
        continue_on_timeout: false
      - repeat:
          # stop when actual_weight >= target_weight - estimated_drip_weight
          # estimated_drip_weight is calculated above
          until: "{{ (states('sensor.coffee_controller_scale_weight') | float >= (states('input_number.coffee_target_weight') | float - states('sensor.coffee_estimated_drip_weight') | float))}}"
          sequence:
            - delay:
                milliseconds: 100
      - service: switch.turn_off
        target:
          entity_id: switch.coffee_machine
      - delay:
          seconds: 2
      - service: switch.turn_on
        target:
          entity_id: switch.coffee_machine

@joeshmoe9898 i'll update the readme in the coming days with a proper set of instructions for the ESPHome :) Hang in there.

@adampetrovic any chance you could provide a rough outline of the instructions? I've got an esp32 (a CYD) up and running and have connected it, the scale, and the machine together in HA. However, I'm having trouble with loading the automation to do the actual work though... When I try to paste in the automation.yaml as an automation on the esp32, HA gives the following error:

"Message malformed: extra keys not allowed @ data['input_number']"

Thanks in advance!

@taupeshuitun the automations.yaml is meant for homeassistant, not ESPHome.

@adampetrovic thank you so much for putting this together.
Have you experienced very unstable Bluetooth connection? IN the ESP32 logs it seems to connect, but immediately after it disconnects and none of the HA buttons result in any action on the scale.
I have removed the ACAIA integration from HA and remove the scale from the ACAIA app in order not to have duplicate connection requests to the scale.

[10:08:23][I][esp32_ble_client:227]: [0] [60:8A:10:4D:C1:6E] Connected
[10:08:23][D][ACAIA:062]: Registering for notifications
[10:08:23][W][ACAIA:069]: No control service found at device, not an Anova..?
[10:08:23][D][esp32_ble_client:188]: [0] [60:8A:10:4D:C1:6E] cfg_mtu status 0, mtu 247
[10:08:23][D][ACAIA:232]: GATTC Event incoming, type 18
[10:08:23][W][component:237]: Component esp32_ble took a long time for an operation (53 ms).
[10:08:23][W][component:238]: Components should block for at most 30 ms.
[10:08:25][D][button:010]: 'Start Timer' Pressed.
[10:08:25][D][ACAIA:273]: Starting timer

Hey @adampetrovic, cheers for this!

With the new cloud only communication with the machine are your timings still working? I guess I'll just adjust the drip duration to include the latency with shutting off the machine.

@Subtletree The only thing impacted by the new firmware and its delay is the trigger to the brew_by_weight automation. That being delayed by 4 seconds won't really have a measurable impact on the result. Really we only care about brewing being active at some point, so we can start calculating the coffee_estimated_drip_weight.

The logic of brew by weight in this case is only dependent on sensor.coffee_controller_scale_flow being accurate. This is a value that is calculated internally in the esp32 component (esphome-acaia-component).

It's also worth noting that the reliance on the ESP32 controller firmware is optional now. If instead, you use ESP32 bluetooth proxies in your HomeAssistant installation, you can use the Acaia component bundled in the HomeAssistant core now. Which supports everything we're doing here: https://www.home-assistant.io/integrations/acaia/

Volume flow rate: Calculates the current flow rate (in mL/s) while brewing.
Weight: The weight currently shown on the scale.

I will look to provide a blueprint to perform brew-by-weight when I get a chance, which should significantly lower the barrier to entry for people to get started.

It's also worth noting that the reliance on the ESP32 controller firmware is optional now. If instead, you use ESP32 bluetooth proxies in your HomeAssistant installation, you can use the Acaia component bundled in the HomeAssistant core now. Which supports everything we're doing here: https://www.home-assistant.io/integrations/acaia/

Volume flow rate: Calculates the current flow rate (in mL/s) while brewing.
Weight: The weight currently shown on the scale.

I will look to provide a blueprint to perform brew-by-weight when I get a chance, which should significantly lower the barrier to entry for people to get started.

This would be great. Thank you very much Adam