I have been thinking about building myself a wireless IOT thermostat as a project for a while. One consideration I have thought about is power consumption. I looked at my current thermostat and as far as I can tell it does not have a C cord to draw power from. Instead it runs on 2 double A batteries. As far as I can tell its been running for over a year + on these batteries with no signs of needing new ones. I cannot understand how it can be this efficient as it uses relays to switch the furnace and AC on. From all the data sheets Ive looked at relays tend to draw hundreds of mW. I believe the ones I am looking at are ~ 100mW, so at 3.3V, that is 30mA. Two heavy duty AA batteries have between 1000-3000 so assuming a 3000 A hour capacitance and a 10% duty cycle that should be 1000 hours of battery life or about a month. Of course the batteries are running other things, like a electric thermometer, micro controller, RTC, and an LED display, all which I would assume require at least 1 mA of power to run continuously.

So does anyone know how my thermostat does it? Are they using latching relays? Are they sleeping the MC?

My plan was to have a raspberry pi do the heavy lifting for processing sensor data through out my house and have a esp8266 run off of 1 or 2
NCR18650B batteries, and have it wake every minute to check on/off commands, but this wouldn't work with standard relays and the relays and eps8266 would kill the batteries in 1-2 months. I'd appreciate any insight into how my current thermostat is so efficient and how I can make mine more efficient.

 

Do you have a diagram on your current thermostat ?
And could it be tht the battery is a backup for the controller ?

 

The relay in a smart thermostat is a latching type. It needs only a short pulse (~10mS) to flip it one way or the other. Once flipped it is held in place magnetically. Average current is therefore very low. The MCU spends much of its time sleeping to conserve energy.

 

I believe its power may be normally extracted from the AC voltage from the furnace.
The batteries are just for backup (or perhaps used when the furnace is on).

 

again I'm pretty sure it's not drawing from the furnace because the only cords available are those that switch the heat, ac and fan. I'm guessing my thermostat is using switching relays and just sleeping a lot now that I am doing more research. I'll see if I can get a picture of my thermostats insides.

 

I believe its power may be normally extracted from the AC voltage from the furnace.

The stat I have is a Horstmann model. It draws no power from the mains. Its only connections are a wire pair to the SPST switch contacts of the latching relay. Runs for just over a year on a pair of AAA alkaline cells.

 

If they are low power relays and only on when the heat, ac, or fan are on, then the total energy drain won't be they high if you also use an LCD display and a low power micro (running at a low clock rate) to control things.
The highest power draw would likely be the receiver circuit, if it's wireless.
You can minimize power consumption by pulsing the power to the higher power devices, such as the receiver and temp sensor at a low duty-cycle, once per second or so.

You could also steal a small amount of power from the furnace lines when they are inactive (small enough to not activate the furnace relays).
That way the battery would only need to power the circuit when the furnace is active.

 

Looks like my thermostat uses a double pole double throw latching relay (RSBL-3-902). It has pins for RC, RH, W,Y,G, so no constant power. two switches on the front allow it to run everything from a single relay (Switching heat and cooling, and a fan auto or on switch). Can't really figure out what all the transistors and other components are for. Hard to say without seeing the other side, but I"m not about to disassemble my thermostat without a working replacement.

 

The transistors are likely there to drive the relay coils.
Our thermostat is superficially identical to yours and the first two batteries lasted six years