Shelly 2.5 240v Inputs not working as expected at lower voltage

Hello there everyone!

I’ve used a Shelly 2.5 running esphome to automate my central heating and hot water, to replace an old Honeywell 7 day programmer. All working great!

Initially, I used the old programmer as a backwards compatible switch for both the heating and water via the programmer’s button overrides, which worked as expected. All the states in HA were correctly reflected.

I would now like to get the status of my two 240V thermostats instead on the Shelly’s two binary sensors so I can track the time it takes for the hot water/heating to reach temperature.

I have been trying to get this working, but the binary sensor keeps saying On when the thermostats aren’t sending 240v.

Taking a closer look at the inputs with my meter, I can see that they are floating around 16-17VAC when open and 240V when they are closed.

The strange thing is, the old programmer’s live 240V switch lines float much higher than this when off (78V and 44.5V respectively and even just over 100V) and yet the binary sensor always showed the correct value of “Off” until the button was turned on. It’s always reported the correct state when using the programmer as the switch both in the Shelly firmware and esphome.

I don’t understand why at an even lower voltage the binary sensor shows as “On” when really it should be “Off”.

I wouldn’t have thought that such a low voltage would set the HIGH state on the sensor, given that the programmer never does? What am I missing and is it possible to get the correct state from the binary sensors in esphome?

Many thanks
Rich

Be careful when using shelly 2.5 or similar with the thermostat output. Some thermostats seem to connect the output to neutral when “off” instead of really disconnecting it.

And there is around 130V between N and SW1/2 which you don’t see if you only test the thermostat output and don’t test it against the shelly input. This in turn causes the input of the shelly to recognize it as “on” AND will also heat it up => BAD
You need a dry contact as input for the shelly. You could use a relay between the thermostat and the shelly to ensure this.

Hi k8n

Thanks for your reply and insight on this - very much appreciated. I hadn’t expected that as they both looked like they were live switched.

When I measured between SW and L, it was reading ~220V, which I believe does confirm the presence of a neutral on the switch line in the “off” state? You were also right, the Shelly was getting hot - running a lot cooler now.

Am I right in thinking that was essentially creating a low resistance short circuit, from looking at the schematic?

I’m going to take a different approach and use some temperature sensors with another Shelly running at 12V instead!

Thanks again
Rich