Unstable W100 + E1

[nzjrs]

Hi,

I’ve got a W100 associated with an E1 TRV and an automation which fires the boiler when the E1 requests heat.

Based on the heating traces the W100 looks to have had a bug and not communicated with the E1 for more than a day, which caused the boiler to turn on for the whole day and waste a heap of energy.

This isn’t good.

Here are the sampling settings for the W100. It seems to have not taken a measurement for a whole day and this causes the E1 to be stuck on.

(It’s 12 degrees because I’m away from home. But this means it needs to be reliable!)

[gtxteo]

Hi try with my tutorial

Tutorial - Troubleshooting Aqara W100 whit Aqara Valve E1

Hi everyone. I followed @gafich10 tutorial to change the temperature setting in custom mode. Since then, the W100 has been regularly connected to E1 without any association errors. So, I switched from standard mode to custom mode. I no longer see any log issues on E1. Before that modification, I often had a “connection error” with the E1 valve. Now, however, I don’t. In practice, the settings to modify on the W100 device are as follows: Set: Temperature and humidity sampling settings from sta…

[gafich10]

Hello, glad to hear. I don’t know what the reason is, but this problem is solved when you set W100 to manual mode with a polling period of 10 minutes. The problem completely disappears.
If you analyze it, then I need to raise this problem with the developers more persistently. It has become something massive. Tutorial - Troubleshooting Aqara W100 whit Aqara Valve E1

[nzjrs]

Ok I’ll change the sampling period to custom 10 minutes.

In standard it only worked for a couple of hours. In High it worked for about a month before this bug.

Is custom actually fixing the problem or it will now only make a mistake every few months?

[gtxteo]

Currently, after a month-long test, I have not encountered any disconnect problems. Please note that the data transfer mode, temperature, and humidity are set to standard or high. This always tends to preserve battery life. This can be bypassed with the custom mode.

[gafich10]

@nzjrs

Is custom actually fixing the problem or it will now only make a mistake every

I I set this parameter from the very beginning, the W100 sensor has never fallen off. By the way, I turn off the humidity altogether, it is updated ~once an hour. I save battery.

[nzjrs]

I assume sampling period the interval over which the temperature measured (ie 1Xmin average)?

I’ll set both to manual. Aqara probably doesn’t wake up the device reliably.

This should mean they wake up the device at T0,T0+1,T0+6

[gafich10]

Why don’t you want to set 0.2°C, this gives an additional trigger on E1 about the temperature change?

[gtxteo]

One minute is to much. Please set Threshold and period 10 minut and 0.5°.
So you preserve your battery status

[gafich10]

@nzjrs

This should mean they wake up the device at T0,T0+1,T0+6

One minute in the upper field is the update rate of the W100 dial. If you set the period, it becomes 10 minutes in total, by the way, this reduces unnecessary signals for E1 adjustment.

[gafich10]

I forgot to say. I have W600 and E1 and W100 connected Zigbee directly to the gateway. To eliminate possible disconnections due to repeater problems.
In Ukraine, the industrial network often disappears now (you know why), and such a Zigbee connection is very reliable for operational work without a repeater.

[nzjrs]

Understood. In this case both devices are within 2m of the hub, in the same room (central hallway). The hub also has battery backup and never loses power. This can only be a software bug and not a communication issue unfortunately.