Siemens SX66T096EU/44 dishwasher Questions

[LiamHD]

I own a SX66T096EU/44 dishwasher from Siemens. It was manufactured in Nov. 2011 (according to the FD number).

I would love to integrate into my Homeassistant installation. Does anybody know details about this dishwasher / the "T"-series?
Does is also use the d-bus protocol?

Out of curiosity I opened up the front panel and found the display unit:

The backside looks like this:

We have the following connectors on it:

and:

I assume that this cable is coming from the "main" unit:

Could this be the d-bus?

Could I "plug into" the d-bus at this location of the machine or do I have to connect on the "main" unit? I think it's enough room up here near the display unit and I would not need to get the complete machine out of the kitchen cabinet.

[hn]

The “thick” connector with the pink internal wires, where the red plug goes, is most likely a D-Bus. But don’t just take my word for it — test it yourself, because I’m not taking any responsibility if it explodes 😏

[tubist2004]

If you have any possibility to (safely!) measure the voltages, I would go for the purple one. Its signals are nicely routed to the next (empty) connector, that could be the point you might get the easiest access to the bus.

Could be the connection to an optional next device or a port for a service technician.

[LiamHD]

The “thick” connector with the pink internal wires, where the red plug goes, is most likely a D-Bus. But don’t just take my word for it — test it yourself, because I’m not taking any responsibility if it explodes 😏

Yes. You are correct. This is the D-Bus. Also @tubist2004 comment is correct. The empty connector is a very easy access to the D-Bus.

Today I tested @Bouni's board. I connected it to this empty connector. Unfortunately I have some issue with it:

When I connect @Bouni's board while the dish washer is already powered up the board boots (red led is on) and it's also available via wifi and I see it's normal boot up logs in the ESPHome console.

Pressing any button on the dishwasher's panel doesn't do anything. The panel does it's normal "button pressed"-beep but the dish washer does not react as it does normally. For example when I change the selected program the remaining time on the display does not change. Also I don't get any log-output on the ESPHome console about this interaction.

I assume that the beep is generated directly on the panel's board but the information that the button was pressed does not reach the dish washer's "main" board because the D-Bus is somehow jammed.

Powering up the dishwasher while the board is connected is even worse. This directly leads to the situation that the dish washer's panel goes into a fail-state (flashing 3 leds and not showing any remaining time).

I think I measured the D-Bus Pins correctly because the ESP boots up correctly and when I disconnect the additional board, power down the dish washer, wait some time and power it up again, then everything is back to normal again. Lucky me I didn't brick the dish washer ;-)

Might the problem be that I connected @Bouni's board via the control-panel (on the top of the door) instead connecting it directly to the "main"-board (down at the bottom of the machine)?

@hn or @Bouni do have an idea how to fix this issue?

This is my ESPHome config:

esphome:
  name: geschirrspuler
  friendly_name: Geschirrspüler

external_components:
  - source: github://hn/bsh-home-appliances@master

esp32:
  board: esp32-c6-devkitc-1
  variant: esp32c6
  framework:
    type: esp-idf

# Enable logging
logger:

# Enable Home Assistant API
api:
  encryption:
    key: "..."

ota:
  - platform: esphome
    password: "..."

wifi:
  ssid: !secret wifi_ssid
  password: !secret wifi_password
  min_auth_mode: WPA2

  # Enable fallback hotspot (captive portal) in case wifi connection fails
  ap:
    ssid: "Geschirrspuler Fallback Hotspot"
    password: "..."

captive_portal:

output:
  - platform: gpio
    pin: 
      number: GPIO07
      inverted: true
    id: activity_led

status_led:
  pin:
    number: GPIO06
    inverted: true

uart:
  id: dbus_uart
  rx_pin: GPIO4
  tx_pin: GPIO5
  baud_rate: 9600
  
bshdbus:
  uart_id: dbus_uart
  on_frame:
    - logger.log:
        format: "Received frame dest 0x%02x cmd 0x%04x: 0x%s"
        args: [dest, command, format_hex(message).c_str()]

binary_sensor:
  - platform: bshdbus
    dest: 0x24
    command: 0x2006
    binary_sensors:
      - id: bsh_dw_rinse_aid
        name: Klarspüler
        icon: mdi:glass-cocktail
        lambda: return (x[0] >> 1) & 0x01;
      - id: bsh_dw_salt
        name: Salt
        icon: mdi:shaker-outline
        lambda: return (x[0] >> 0) & 0x01;
      - id: bsh_dw_door
        name: Tür
        device_class: door
        lambda: return !((x[0] >> 3) & 0x01);

sensor:
  - platform: bshdbus
    dest: 0x25
    command: 0x2008
    sensors:
      - id: bsh_dw_remain
        name: Restzeit
        device_class: duration
        state_class: measurement
        unit_of_measurement: min
        accuracy_decimals: 0
        lambda: return x[0];

[hn]

• Double, triple check everything: cabling, config. Check if DATA is 5V with a standalone module (@Bouni's board without anything connected). If you have access to an oscilloscope or logic analyzer check if the pin is high during and after boot.
• Best option: Connect the ESP32-Module directly to the power module. I think (not know!) the EPG60603 has at least one D-Bus on the left side.
• The ugly way: To work around this possible problem, disable the pullup resistor R4 by cutting the trace marked in red on the PCB with a very sharp knife (obviously, this change cannot be undone easily):

As, always, I can not guarantee for anything.

[LiamHD]

• Best option: Connect the ESP32-Module directly to the power module. I think (not know!) the EPG60603 has at least one D-Bus on the left side.

I now did this and wit works perfect. I connected it to the power module the free dbus connector marked here:

Space is so limited down there that I had to remove the "Flottenspeicher" which was quite a hassle (you need this tool for that) before I could remove the power module.

See my PR #82

[hn]

... does not reach the dish washer's "main" board because the D-Bus is somehow jammed.

Yes.

This is my ESPHome config:

Looks ok.

Might the problem be that I connected @Bouni's board via the control-panel (on the top of the door) instead connecting it directly to the "main"-board (down at the bottom of the machine)

Maybe. I’ve never tried that. The control panel likely has a 10k pull-up, and Bouni’s board also has a 10k pull-up (≈5k total). The components might not be able to pull the bus low, meaning the bus could be stuck. Still, that would surprise me — it seems odd. But I'm not an EE.

Triple-check you cabling, even the plain D-Bus cable. Is there a short-circuit betweend DATA and GND somewhere?

[LiamHD]

Maybe. I’ve never tried that. The control panel likely has a 10k pull-up, and Bouni’s board also has a 10k pull-up (≈5k total). The components might not be able to pull the bus low, meaning the bus could be stuck. Still, that would surprise me — it seems odd. But I'm not an EE.

You might be correct. As I now got it working by connecting it to the power module without any changes to my ESPhome config.

[hn]

@Bouni 's board V1.0 has a switchable (on/off) pullup resistor, so installation at the front panel might be possible with that adapter revision.

[LiamHD]

Triple-check you cabling, even the plain D-Bus cable. Is there a short-circuit between DATA and GND somewhere?

I already did that. No short-circuit on the cable. I checked that while the cable was not connected. I can check it again while everything is connected (and the machine is powered down) - but I would expect that this doesn't make a difference.

[MichaelHeimann]

Hi @LiamHD - I've had the same issue today. BSH Board had the red led on but the front panel had all LEDs lit up and didn't come to a normal start and started showing an error after a while (and beeping). This was even with the BSH board on the system-PCB - and not on the front-PCB of my Bosch WTWH7540

My issue was that I copied the WT47W5W0 config too much - I didn't modify the uart: section for the BSH Board. Your config looks fine in that way, but maybe you can doublecheck there, maybe activate debug again.

Hope it helps.

[LiamHD]

Thanks. So what did you change to make it work?

[MichaelHeimann]

Well, I modified the uart section from this:

uart:
  id: dbus_uart
  rx_pin: GPIO6
  baud_rate: 9600

to this:

uart:
  id: dbus_uart
  rx_pin: GPIO4
  tx_pin: GPIO5
  baud_rate: 9600
  debug:
    direction: BOTH
    dummy_receiver: false
    after:
      # Trigger after 1 byte to ensure the LED blinks immediately
      bytes: 1
    sequence:
      - output.turn_on: activity_led
      - delay: 20ms
      - output.turn_off: activity_led

Your config seems to have the right GPIO pins for the BSH Board - did you maybe change that also in the config but didn't yet update the firmware?

[LiamHD]

I now got it running by connecting to the power module directly - not by connecting it to the control panel. See PR #82

[3DJupp]

Wanted to add a small ESPhome-related quick and dirty-trick for saving the total duration, storing info about the cycle and current state. Non-persistent, as the ESP will shut down after the cycle is finished. Not perfect, of course, but might be some inspiration.

Relevant changes (total duration+state sensor)

globals:
  - id: last_program_time
    type: int
    restore_value: false
    initial_value: '0'
binary_sensor:
  - platform: bshdbus
    dest: 0x24
    command: 0x2006
    binary_sensors:
      - id: bsh_dw_door
        name: Tür
        device_class: door
        lambda: return !((x[0] >> 3) & 0x01);
sensor:
  - platform: bshdbus
    dest: 0x25
    command: 0x2008
    sensors:
      - id: bsh_dw_remain
        name: Restzeit
        device_class: duration
        state_class: measurement
        unit_of_measurement: min
        accuracy_decimals: 0
        lambda: return x[0];
        on_value:
          then:
            - lambda: |-
                // nur speichern, wenn sinnvolle Restzeit und Tür offen ODER Startknopf nicht gedrückt
                if (x > 0 && (id(bsh_dw_door).state || !id(bsh_dw_btn_start).state)) {
                  id(last_program_time) = (int)x;
                  ESP_LOGI("timer", "Gespeichert (über Restzeit): %d", id(last_program_time));
                }
  - platform: template
    id: bsh_dw_duration_total
    name: "Letzte eingestellte Laufzeit"
    lambda: return id(last_program_time);
    device_class: duration
    state_class: measurement
    unit_of_measurement: min
    accuracy_decimals: 0
    update_interval: 10s
  - platform: template
    name: "Status"
    id: bsh_dw_status
    icon: mdi:play-pause
    update_interval: 5s
    lambda: |-
      if (id(bsh_dw_door).has_state() && id(bsh_dw_door).state) {
        return {"paused"};
      }
      if ((id(bsh_dw_door).has_state() && !id(bsh_dw_door).state) &&
         (id(bsh_dw_remain).has_state() && id(bsh_dw_remain).state > 0)) {
        return {"running"};
      }
      return {"idle"};

[jowi24]

Hi @LiamHD, I have the exact same machine (SX66T096EU) and ran into almost the identical situation.

Short answer: don't use the connector up at the display/control panel — go for the main power module at the bottom instead. There are two dedicated pins there (meant for a service technician) that gave me clean, stable D-Bus access on the first try.

Longer version, including the expensive way I learned this: I also started at the display connector first, and while probing it with a multimeter to check for voltage, a probe slipped and briefly shorted two pins. That took out the whole power module, and I ended up paying €315 for a replacement power + front module. Not saying that's necessarily what's happening in your case — could easily be a wiring/connector issue as you describe — but given the panel is going into a fail-state, I'd personally stop probing around up there and move down to the power module instead.

I wrote up the whole process (plus how I reverse-engineered the specific commands from logs with some AI-assisted analysis), including a video:
https://smarthome-aber-sicher.de/en/bsh-maschinen-hacken (video: https://youtu.be/NSkQ78OTgHY)

My ESPHome config for this exact model is here, might save you some time since it's the same SX66T096EU: https://github.com/jowi24/bsh-dbus-sx66t096eu

Happy to help if you have questions about the pinout at the power module.