Solar Powered Smarthings Sensors - Success!

Some of you familiar with my garage thread may recall that I’m using SmartThings to automate things. The door sensors in the garage see cold temps (-35C) and therefore their 3V lithium cells do not last long at all (like 4 months) in the cold weather. The other issue is that when they “die” the sensor just stops detecting, with no notication per se.

With my pool control project, I was able to power wireless temp sensors with a solar cell/rechargeable battery project, so I figured I’d try this out with the SmartThings sensors.

This is the older setup that powers three Fibaro door sensors that have external temperature probes attached. The three fibaro FGK-10x sensors and 3 AA cells are shielded from water using a standard outdoor electrical box. The lithium cells in the sensor (3.6 Volt) are removed, and + - leads soldered to the battery connectors inside each FGK-10x. I found reference to the 1st generation Fibaro door/window sensor (zwave) that has a connection block to allow use of an external temperature probe (DSB1820), so this is why they are being used to control my pool solar heating system.

This setup has run 100% reliable outside all year (with temps as low as -35C) since July 2019 with zero issues, resets, etc. They have always reported “low battery” but have remained powered up since installation.

Now the SmartThings multifunction sensors (I think these are gen2) use a 3 Volt lithium cell (CR-2450), so I figured 2 x AA batteries would work fine. These sensors work with an external magnet to detect open/close events but they also integrate temperature sensors and accelerometers as well. The first test was trying the sensor with 2 AA celles in series (2.9 volts) which work perfectly. They also powered up fine directly from the 5 volt solar cell in full sunlight. Once you connect the AA cells and solar panel, the voltage is maintained pretty much at 2.9 volts as this panel is only .5 watts@5V. In other words, with these very low power Zigbee sensors, the 5V panel has worked perfectly to keep 3 x AA cells charged…so we’ll see how well we fare with 2 x AA charged by the 5V panel.

The solar cell is a 5 Volt, 0.5 watt I found on Amazon.com. They have a blocking diode installed by the manufacturer, and claim to be waterproof. The older setup pictured above has been outside in the rain for 3 years.

I also ordered some holders for 2 AA cells, with an on/off switch on Amazon.

I’m just using Amazon Basics NIMH rechargeable AA batteries.

Bench testing the setup. Basically just connect up the + and - leads, and solder to the sensor. The sensor (with cover removed) is sitting between the AA batteries and solar cell.

Back of the solar cell (as provided by the manufacturer) where the diode is sealed under some silicon. This keeps the solar cell from draining the AA cells at night.

Battery holder (in series) with on/off

Here’s the panel spec.

A pretty cool way to have a wireless (temperature, contact, motion) sensor that never needs batteries and will reliably stay powered up even in very cold temps :slight_smile: I’ll post a few pics of the latest setup installed. Cost to do this for one sensor is about $8 with just a few solder connections to do.

11 Likes

place holder

So no charging or regulating circuit required?

I’ve thought about this for several outdoor sensors. Just never got around to it yet. Now I have a how to. Thanks.

1 Like

@TN_Oldman, zero regulation. The Fibaro sensors have been powered since July 2019, 24/7 continuous with the setup with zero attention required. The solar cell is very low power, .5 watts@5V, so only 100 milliamps. That said, it’s not in full sunlight 12 hours a day either.

1 Like

So going to try this! I’ve have 2x Duracell AAs running an outdoor multi-purpose sensor since the date of this post - which was just over a year ago - and still showing 3% battery!

1 Like

@Jamie_D , the Fibaro sensors use a 3.6 Volt lithium cell normally, but with 3 x AA, they have showed 1% battery since installation.

The Smarthings Multipurpose sensor uses a 3.0 V lithium, but on 2 x AA (Amazon Basics NiMH low discharge rechargeables) shows 83% battery after the freshly charged batteries dropped voltage slightly. My guess is that it will stay at 83% on solar as well.

1 Like

After a few days of testing, I’ve taken a few measurements. The 5V solar cell charges the batteries at pretty much 100mA in full sunlight, and about 5mA in shade. I’ve estimated that the power use of the sensor, combined with daily loss of the NiMH (low self discharge cells) would be about 5mAh per day, likely more in very cold temps. It makes sense then that even a few hours of outside light (even shade) should keep the cells fully charged in both of my setups. SmartThings reports the 2xAA batteries at 83% after a day in the cold garage (no sunlight) and back to 100% after 2 hours in direct winter sun (winter sun is pretty low in the sky right now).

This estimate matches up with the original sensor setup (powering three fibaro door sensors with temp probes) which uses 3 x AA cells in series and has been powered now for nearly two years with the 5V solar cell nearly 100% shaded. My suggestion would be to place the solar cell in a location that is mostly shaded as there is no need to charge at 100mAh for more than 30 minutes a day to keep the cells fully charged.

Not super pretty, but the mission of a wireless sensor with perpetual power at -35C is done. I figure the NiMH cells will be fine for 3-5 yrs. Either way, garage door open alerts due to a dead sensor should be a thing of the past. One more to do on the roll up door…

Just an update with the two sensors set up with solar charging now for a full month in brutally cold temps.

Outside temps dipped to -38 C last week, with inside temps at a low of -15 C (garage is only heated when needed).

  1. Sensor 1 (has about 1 hour of partial sun per day) has oscillated between 83% and 100% in ST GUI.
  2. Sensor 2 (solar panel in shade 100% of the day) has been showing 83% since installation.

State of charge is obviously not super accurate with these sensors, but based on voltmeter tests, the NiMH batteries have stayed at 100% despite very cold temps. I’d call that 100% success :slight_smile:

1 Like

Nice! I’m looking forward to setting up a solar powered Zigbee repeater network to reach motion sensors at different areas of my property, the furthest distance being about 300 meters.

Among other things, I was looking for a way to start video cameras, located at the house, recording earlier, as not to miss anything.

This will be helpful.

James, (@blkwll ) just keep in mind that these battery powered sensors don’t repeat. Anything designed to run off AC (and AC powered zigbee devices will work as repeaters) will of course have higher power requirements, so you’ll need to do a bit of research and make sure your battery pack and solar panel can supply enough power :slight_smile:

1 Like

I’ve done a little research already, and I know my power requirements will be considerably larger.

But thanks for reminding me to not under do it.

I’ve not seen mention of this device here, but if all goes well, It will be pretty cool if it “just works”

one reviewer stated that it connects to Smartthings, but who knows? Might also try a some LED controllers reported to be repeaters, which would also work at 5v (one of the two, mentioned is said to work as a repeater, not the Tuya one).

If those don’t work, I’ll probably go Z-wave, which would probably be better anyway, if not as inexpensive.

Do I need a over-charge protection circuit for this setup? Thanks

I don’t have any overcharge protection in my setups as the potential charging current is so low. I’ve had three of these solar charge sensor setups running now for up to three years (running six sensors)…no issues.

1 Like

It’s been 3 years now since I’ve started using the “solar” hack for my pool/garage sensors so that they have become more or less perpetual with no battery changes required (so far anyway). I’m updating the pool water temp and roof sensors to a more robust box. This waterproof box houses a Fibaro FGK-10x sensor , and a 3 AA cell holder with rechargeable NiMH batteries. The lithium cell in the sensor (3.6 Volt) is removed, and + - leads soldered to the battery connectors inside the FGK-10x. I found reference to the 1st generation Fibaro door/window sensor (zwave) that has a connection block to allow use of an external temperature probe (DSB1820), so this is why they are being used to control my pool solar heating system.
[/quote] I’ve hot glued two rare earth magnets (square, powerful) to the backside of this box so it will just attach to a pool frame vertical brace. The pool is not up yet, hence the ladder used for demo purposes :slight_smile: I have 4 Fibaro door sensors, and 3 Smarthings multipurpose sensors which have stayed powered up and happy through 3 winters (-35C and a lot of snow) using this solar setup.

3 Likes