Airtouch 4 integration (AUS)

Same thing this morning, NodeRed showed all nodes had fired but the HA log showed ‘Connection error in receiver.’ When I manually fired that flow half an hour later everything worked fine.

What I had done was turn on NodeRed’s ping function that tries to keep the connection to the HA server more alive and disable my hourly tweaking of an AC zone to try to keep the AirTouch responding.

Tonight I’ll try something different and include a couple of extra nodes at the start of the flow to tell the AirTouch to start heating mode and put a bit of a delay between them. Hopefully that gives the AirTouch time to wake up.

Well that worked fine. No error messages at all and the ac turned on as expected!

Marian, thanks again for your work in developing and maintaining this integration, I’ll send you a donation today.

If it’s useful, the ac was running all day but the same error occurred when NodeRed tried to talk to AirTouch after a period of maybe a couple of hours of not talking to it. Does that make sense?

I’ve gone back to an hourly tweak of a zone to see if that keeps the connection alive.

Thanks guys.

I will do a small update soon where the zones should update even when AC is off.
Should be the last important thing until a firmware update.

2 Likes

Fantastic, thank you. I still get the ‘Connection error in receiver’ error messages despite trying to keep the AirTouch connection alive every 10 minutes so it must be something else. The only way I can see it is to look in the log.

Is it possible please to return an error code/message from your integration so that NodeRed can see that an error occurred?

You can set up email or app notifications relatively easily. Here’s a few things I have in my configuration.xml

Change the log levels

logger:
  default: error
  logs:
    # log level for HA core
    homeassistant.core: info
    custom_components.polyaire: info

Set up an email server so you can email notifications

# Email notifications
notify:
  - name: EmailServerName
    platform: smtp
    server: smtp.yoursmtpserver.com
    port: 587
    timeout: 15
    sender: [email protected]
    encryption: starttls
    username: smtp-username
    password: smtp-password
    recipient:
      - [email protected]
    sender_name: Home Assistant

The email definition goes along with this automation

alias: Notifications System Log All
description: ''
trigger:
  - platform: event
    event_type: system_log_event
condition: []
action:
  - service: notify.EmailServerName
    data:
      title: Home Assistant log entry made
      message: '{{ trigger.event.data.message[0] }}'
mode: single

I also have a service “notify.mobile_app_tim” which goes to my phone, but I can’t remember how I set it up. Some info here.

1 Like

Although I use smtp and mobile notifications for other things I didn’t know about alerting on log entries. That’s super helpful, thank you!

I’ve just upgraded to release 1.4.4. I can see that each zone now has a binary_sensor for zone turbo, but I don’t seem to have any spill sensors. Should they have been created automatically?

I don’t think spill is done by this integration - at least I’ve never seen it.

OK, might be a more general bug then as the change log lists it against release v.1.4.3
Changes:

  • added spill sensor and turbo sensor
  • added workaround for zones not updating temperatures

The code was indeed already available to add the spill sensors to the integration but there was a small bug. I’ve submitted a pull request a few days ago to fix this bug. You could already manually change this in the following file: custom_components/polyaire/binary_sensor.py.

        OLD LINE   spill_sensor = AirTouchGroupBattery(airtouch, group)
        NEW LINE   spill_sensor = AirTouchGroupSpill(airtouch, group)

Thanks for the heads up and fix tamorix.

Seems to me this is the benefit of the integration being part of HACS rather than HA, things get fixed much quicker. It’d be a bit like 3rd party apps needing to wait for Apple to release a new version of iOS otherwise.

I think yes, however it would be great if HA updated to a more recent version…

@mihailescu2m Hi Marian, would it be possible to build in a sense-check of the temperature readings from the ITC sensors please? Every one of the sensors in my system spiked to 154.7 C in the early hours of this morning. I don’t remember it being qquite that warm in reality! But it sure messes with the graphs.


Thanks.

@spry-salt I had a similar thing. Using the Apex Charts card here’s how I solved it - the transform below. It’s not a perfect solution but it prevents spikes in the graphs.

type: custom:apexcharts-card
graph_span: 8h
header:
  show: true
  title: Room Temperatures 8h
  show_states: true
  colorize_states: true
chart_type: line
update_interval: 15m
yaxis:
  - id: a
    min: ~20
    max: ~25
all_series_config:
  stroke_width: 2
  curve: smooth
series:
  - entity: climate.itc_living
    attribute: current_temperature
    name: Lounge
    color: '#0091ea'
    transform: if (x > 35) {return 21} else {return x};
1 Like

Guys, Life got in the way - but got back to this yesterday

Installed 1.4.4
Running HassOS 8.4
HA 2022.7.5

Installed 1.4.4 yesterday and it was running fine all night - i have an automation that sends (through MQTT) the ITC temps every minute to my Node Red instance (on another machine entirely) so i can act on them

I have a Daikin 3 phase unit
I have the Daikin standard controller installed
And i have Airtouch 4 with 5 zones with dampers and 5 x ITCs

I have a Programme setup on the Airtouch console to turn the unit on at 6am and run for an hour until 7AM every day - that ran fine.

at about 9am this morning i tried to use the Thermostat card to turn the unit on - it indicated it was running - but went to the main part of the house that this AC services about 20 minutes later and the unit was not running - Green light on the Airtouch console, so turned it off and then back on at the console - these events show up in the logbook so it appears to be communicating OK

Checked 10 minutes later and the unit still was not running - so went to Daikin console and turned it on there - left it for 10 minutes and still the unit had not started up

Had to once again go to the circuit breaker and turn it off for 30 seconds and then back on to reset everything and get the main console responding again.

Unlike last time the Airtouch app on my phone (android) and Tablet (IPAD) were still responsive - but did nothing.

Its almost like something in the integration is blocking at the console.

I note that i never had this with the original integration from the start of this thread

Should i just assume i am some weird edge use case and revert back to the old integration (non push) and live with it or is there some more troubleshooting i can do with this ?

Craig

I suspect you’re a weird edge case. Works fine for me on the latest version. The worst I get is sometimes HA doesn’t get updates, but I’ve never once had to touch the breaker for my head pump. In your place I’d disconnect HA from your air con entirely for a month or two and see if that happens without HA.

Thanks Tim - yes i had it disconnected for 26 days from the last time around so feel that is pretty conclusive.

I will leave it without the integration for the next week or so and then revert back to the non push one and see where that leaves me

Craig

I’m running the PUSH version on HA 2022.8.2 and am no longer (touch wood) having problems with it responding.

I don’t use the AirTouch app and I don’t have any programs set up on the AirTouch tablet itself. All my AC automations run from NodeRed running on the same VM as HA.

Earlier I had problems with AirTouch ‘waking up’ when the first automation kicked off each morning. I added a couple of NodeRed calls that set one zone to Damper mode, wait a few secs, then set it back to ITC mode about 10 minutes before the first automation starts. That seems to be fine now.

Maybe a point of difference is that my AirTouch stays on all night, set to a low temperature so the AC doesn’t kick in unless the temp gets super cold.

thanks for that - i have noted this post and will then factor that in once i go back to this integration.

In the meantime so i can get a baseline i am going to go back to the non-push integration (which i previously used successfully) and run it for a couple of weeks in the same manner as i did with the Push version

Craig