missing heating data from GW1000 / WS80 an incorrect (?) battery information?

[Max Malycha]

Hi,
after a breakdown of my old device (WS1) I've bought the german version of the GW1000 (firmware V2.2.0 after update) as a bundle with ultrasonic wind-sensor (WS80), classic dip-bucket (WH40) and outdoor temp-sensor  (WH32) from Froggit (HP1000 SE Pro Ultrasonic).
The connection with my old weewx-installation on RHEL7 (3.9.1 rpm) runs without any problems. I've installed the driver via wee_extension and it runs out of the box. All necessary data were received.

A great hug for the work in the development of the driver!

Today I've connected the power-cord from the box with the wind-sensor because it will get cold and misty for the next few days and I wanted to test the correct working of the heating. So I ran the driver in console mode to get the mapped name for heating-voltage and heater-temperature. 

Gosh - in the output in debug level 3 there is no information about the heating:

[root@server ~]# PYTHONPATH=/usr/share/weewx python -m user.gw1000 --test-driver --debug=3

Using configuration file /etc/weewx/weewx.conf

debug level is '3'

IP address obtained from station config

Port number obtained from station config

Battery state filtering ('False') using the default

Interrogating GW1100 at 192.168.3.243:45000

2022-12-11 20:10:09 CET (1670785809): barometer: 1010.7, dateTime: 1670785809, inHumidity: 43, inTemp: 19.8, luminosity: 0.0, maxdailygust: 6.7, outHumidity: 94, outTemp: -0.3, outTempBatteryStatus: 0, pressure: 1010.7, rain: None, rainBatteryStatus: 1.44, rainRate: 0.0, stormRain: 0.0, usUnits: 17, UV: 0, uvradiation: 0.0, wh32_sig: 4, wh40_sig: 4, windBatteryStatus: 3.16, windDir: 115, windGust: 1.9, windSpeed: 0.8, ws80_sig: 4

Really strange. Also there is a difference between the driver data for the devices battery voltage and the information on ecowitt.net weather services dashboard:

Outdoor T&RH Sensor

Normal

Sonic Array

3,16 V

Rainfall Sensor

1,6 V

The difference of the voltage from the rain-bucket in ecowitt.net and the driver is not so big - but it's about 10% ...

I've made a test with an actual weewx installation (4.x) on a raspi and got the same information - no heating-information and the same difference in the voltage.

Is it a bug or an undocumented feature of the driver? ;)

Best regards

Max

[gjr80]

Hi Max,

Today I've connected the power-cord from the box with the wind-sensor because it will get cold and misty for the next few days and I wanted to test the correct working of the heating. So I ran the driver in console mode to get the mapped name for heating-voltage and heater-temperature. 

Gosh - in the output in debug level 3 there is no information about the heating:

[root@server ~]# PYTHONPATH=/usr/share/weewx python -m user.gw1000 --test-driver --debug=3

Using configuration file /etc/weewx/weewx.conf

debug level is '3'

IP address obtained from station config

Port number obtained from station config

Battery state filtering ('False') using the default

Interrogating GW1100 at 192.168.3.243:45000

2022-12-11 20:10:09 CET (1670785809): barometer: 1010.7, dateTime: 1670785809, inHumidity: 43, inTemp: 19.8, luminosity: 0.0, maxdailygust: 6.7, outHumidity: 94, outTemp: -0.3, outTempBatteryStatus: 0, pressure: 1010.7, rain: None, rainBatteryStatus: 1.44, rainRate: 0.0, stormRain: 0.0, usUnits: 17, UV: 0, uvradiation: 0.0, wh32_sig: 4, wh40_sig: 4, windBatteryStatus: 3.16, windDir: 115, windGust: 1.9, windSpeed: 0.8, ws80_sig: 4

This is expected behaviour. The Ecowitt gateway driver (formerly the GW1000 driver) uses the Ecowitt gateway device API to interrogate the device and obtain data. Unfortunately Ecowitt has chosen not to include heater voltage/temperature data in the API so hence the Ecowitt gateway driver cannot provide such data. I do have some ideas on how the Ecowitt gateway driver will be able to obtain this data, but this is someway off being developed.

Really strange. Also there is a difference between the driver data for the devices battery voltage and the information on ecowitt.net weather services dashboard:

Outdoor T&RH Sensor

Normal

Sonic Array

3,16 V

Rainfall Sensor

1,6 V

The difference of the voltage from the rain-bucket in ecowitt.net and the driver is not so big - but it's about 10% ...

There has been some issues with WH40 battery state over the years. The early hardware/firmware version did not report battery state yet the Ecowitt gateway API reported 1.6V (presumably as did the gateway device when it uploads to ecowitt.net). The later version of the WH40 reports the actual battery voltage in 100mV increments. The Ecowitt gateway driver should detect the different WH40 hardware/firmware versions and either omit the WH40 battery state data (early hardware/firmware WH40) or return the correctly decoded battery voltage (later hardware/firmware WH40). This capability was added in v0.5.0b5 of the Ecowitt gateway driver so it is possible there is an error in the driver. Seeing 1.6V makes me suspicious of Ecowitt reporting non-existent battery data, but 1.44V not being a multiple of 100mV makes me suspicious of the driver. Let me look into this further to see how best to troubleshoot.

Gary   

[Max Malycha]

Hi Gary,

thanks for the informations. Sad that Ecowitt does not give the information about heating via the API. But by the way - the little USB-dongle is a great solution for getting data from multiple sensors in an accetable way.

... The early hardware/firmware version did not report battery state yet the Ecowitt gateway API reported 1.6V (presumably as did the gateway device when it uploads to ecowitt.net). The later version of the WH40 reports the actual battery voltage in 100mV increments. The Ecowitt gateway driver should detect the different WH40 hardware/firmware versions and either omit the WH40 battery state data (early hardware/firmware WH40) or return the correctly decoded battery voltage (later hardware/firmware WH40). This capability was added in v0.5.0b5 of the Ecowitt gateway driver so it is possible there is an error in the driver. Seeing 1.6V makes me suspicious of Ecowitt reporting non-existent battery data, but 1.44V 

 OK, tomorrow I will check the battery-voltage in the rain-sensor with a dig. multimeter and report the actual values ...

Best regards

Max