The first sensor goes in the living area, because that is the question you bought it to answer: how hot does it get in here while we are out. Then you start wondering about the wet bay. Then the fridge, then the bedroom that is always the wrong temperature, and by the end of the season there are five of them stuck around the rig and you have opinions.
Which is worth knowing before you buy the first one. For a single sensor the honest answer is that any supported one will do, and the price difference between the cheapest and the best is not worth an evening of reading. It is the fifth and the twelfth where the differences start to bite, and they bite in a place most comparison tables never get to.
Where it can live matters more than what it reads
Everything in this category reads temperature to a degree or so, which is comfortably enough to answer “is it 96 in there with the dog”, and nobody has ever turned a truck around over a tenth of a degree. What actually separates these products is the case around the electronics and the range the manufacturer will stand behind. A sensor rated for a spare bedroom, sitting in a wet bay in February, is not a cheaper version of the right sensor. It is a sensor that stops reporting on the coldest night of the year, which is the night you fitted it for.
The three worth choosing between
1. Xiaomi LYWSD03MMC, once it is reflashed
The cheapest sensor worth owning, and it is not close. A four-pack costs less than one branded sensor, which quietly changes what you do with them: you stop rationing, and every space you are mildly curious about gets one, including the ones that were never worth a proper sensor.
The catch is real, and you want it before you order rather than after. Out of the box these broadcast encrypted MiBeacon data that we cannot read, so each one has to be reflashed with the ATC or pvvx custom firmware before it does anything useful. That is done from a web browser over Bluetooth, it is well documented, and there is a guide to it here. It is still a genuine step you have to take, and it is a step you take before the sensor works with any hub, not just ours. Fifteen minutes for the first one while you find the buttons, about two minutes each after that.
The rest of what you are trading away is physical. Indoor-rated plastic, and a little screen that is perfectly readable on a bedroom shelf and completely useless in a dim propane locker. Keep them out of anywhere damp and anywhere genuinely cold.
2. Govee H5075, if you do not want a project
The small white puck a lot of RVers already own, and the one to buy if that last section made you tired.
It broadcasts continuously, so nothing pairs and nothing gets claimed by one app at the expense of another. Your phone carries on using the Govee app while our hub reads the same broadcasts, and neither one knows the other is there. Temperature, humidity and battery, with no setup beyond putting it somewhere sensible. Several relatives work the same way if you find one cheaper: the H5074, H5100, H5101, H5177 and H5179 are all supported.
It is a consumer comfort sensor rather than an instrument, and it is fairly honest about that. Ask it whether the rig is at 96 and climbing and it will tell you, which is the decision you are actually standing in a car park trying to make. Ask it for a number you would have to defend to somebody and you have picked the wrong tool. Indoor-rated too, same as the Xiaomi.
3. RuuviTag, when the reading has to hold up
There are two reasons to spend the extra: the reading has to be right, or the sensor has to survive somewhere unpleasant. The second is the better reason.
You get an open published protocol, a sealed enclosure, and calibration that holds in a way the cheap sensors do not. It reads more as well: temperature, humidity, barometric pressure, battery voltage and a movement counter. But the sealed case is what you are really buying. A wet bay in winter is closer to outdoors than in, and this is the one on this page that belongs there.
One thing to check if you are buying used. We decode data format 5, which has been the RuuviTag default since 2019, and a tag left on an older format will not decode. New ones come out of the box on 5, so this only catches people who bought somebody else’s.
The whole of the downside is the bill, several times over per sensor. In a bedroom a RuuviTag is doing a job a five dollar sensor does perfectly adequately, and if you buy six of them for six ordinary indoor spots you have overspent on five.
What ends up in our rig
If there are already sensors stuck around the rig, start by checking the compatibility database, which will tell you whether we can read them and what the caveats are on each. Assuming you are starting from nothing, the answer is not one model. It is a mix, chosen per location, and mixing works out cheaper than standardising on the wrong thing:
- Living area, where the dog is. A Govee. Floor level, out of the sun.
- Bedroom, and anywhere else indoors you are curious about. Flashed Xiaomis. This is where the price difference pays for coverage you would not otherwise have bought.
- Wet bay. A RuuviTag. Damp, cold, and the most important reading in the rig in winter.
- Fridge or freezer. A Govee or an Inkbird. Expect a shorter battery life and a weaker signal, because you have put a radio inside a metal box.
Where to get one
Any of these three is a reasonable first sensor. Buy several cheap ones before you buy one good one.
Govee H5075
The small white puck a lot of RVers already own. It broadcasts continuously, so the hub reads it without pairing and without touching the Govee app.
We may earn a commission if you buy through this link. It does not change your price.
RuuviTag
The one to buy when the reading has to be right. Open published protocol, sealed enclosure, and it holds calibration in a way the cheap sensors do not.
We may earn a commission if you buy through this link. It does not change your price.
ThermoPro TP357
A widely stocked display hygrometer. It broadcasts on its own without any hub of its own, which is exactly what RoamVitals wants.
We may earn a commission if you buy through this link. It does not change your price.
Two things people only learn the hard way
Put an alert on the sensor battery. Every sensor here reports its own battery and hardly anybody sets a threshold on it. A sensor that quietly died three weeks ago looks exactly like a rig where nothing has gone wrong, and you carry on trusting the silence either way. Coin cells lose capacity in the cold, so the one that gives up will be the one in the wet bay, in January.
The second is placement, which beats precision every time and costs nothing. A perfectly calibrated instrument screwed to an exterior wall in afternoon sun is measuring the sun. A cheap sensor at chest height in the middle of the room, out of the light, is measuring the thing you actually wanted to know.
Which is a slightly deflating conclusion for a page about which sensor to buy, and it is still the true one. Spend properly on the two or three places that are damp, cold or genuinely important. Buy the cheap ones by the handful for everywhere else. Then take ten minutes deciding where each one goes, because that is the part that changes the readings.
