Precise temperature and humidity sensor

Hello,

I have tried using Xiaomi WSDCGQ01LM temperature and humidity sensors for monitoring of humidity firstly and temperature secondly. I am trying to grow demanding to humidity Venus’s flytrap (Dionaea) which requres control limits within 70-90%RH.

I have verified WSDCGQ01LM reading by laboratory thermohygrograph and found Xiaomi sensor to be pretty inaccrutare. When actual RH is 50% Xiaomi sensor shows 70%. Sensor tolerance equals to Dionaea humidity control range, therefore they are absolutely unaccpetable for my needs.

Are there any precise temperature and humidity sensors with tolerance let’s say 5% preferrably with ZigBee interface?

Any input would be much appreciated.

Regards

In my experience, most of these ‘off the shelf’ temperature/humidity devices seem to use one of a small set of common low cost sensors. Of the 10 or so different BLE and Zigbee devices I have tried and can still be purchased, the one that seems to yield the ‘best’ results is the Goove H5074

https://www.amazon.com/Govee-Thermometer-Hygrometer-Bluetooth-Temperature/dp/B07R586J37

It lists a humidity accuracy of ‘humidity is ±3%RH.’ However, most of the devices out there seem to say about the same. When I say this is the ‘best’ I have experimented with, that is a subjective statement, as I have never measured them against a professional calibrated device.

Not sure if it is helpful, however below are two humidity charts from a couple of my sensors. Different time scales and sensors, so it is not trying to show any comparison. The first chart shows two Goovee H5052 sensors in refrigerator and freezer. These devices are no longer sold, but I believe they may have a similar sensor to the H5074. Unfortunately, my H5074 is used to monitor stove temperatures and as such does not experience as dramatic humidity variations as the H5052’s shown. The second chart shows a shorter time scale of a person taking a shower in a bathroom. The sensors in this bathroom are Xiaomi LYWSD03MMC and Goove H5072, of the devices I have these seem to be the poorest of the humidity sensors, with the Xiaomi better and the H5072 the poorest.

Be careful of ‘close’ model numbers in the ads and pictures that show one unit, but list a different units model number.

I’m not sure what, if anything, you can take away from the charts. The H5052 sensors seem to react quickly and track well, however my example case is in very low temps of refrigerator/freezer. Not sure if high temps will cause different response.
The ‘recovery’ time for these sensors when they get in a high humidity range seem to be one of the issues, it seems that they take a while to ‘dry out’, I am guessing the unit in the freezer is fairly responsive because the water vapor in the freezer is ‘less liquid’ that these cold temps.

I have better over all experience with BLE units as opposed to Zigbee units. The ‘best’ zigbee unit I have experimented with is the Sonoff TH01, it seems to track with a Govee H5052 very closely. However, I never did put the Sonoff TH01 in a area with as wide a range of exposure as the other units.

Good hunting!

3 Likes

David,

Thank you for the comprehensive input. I will try using SonOff TH10 equipped with Si7021 sensor. Silicon Labs claim 3% accuracy for humidity. Hopefully, it will work within specified limits.

Regards

Good luck, look forward to hearing of your ‘Little Shop of Horrors’ success :grinning:
Curious how you figured out the sensor model in the Sonoff TH10? I thought about opening up my devices and trying to figure out what sensors each had, however I never got around to it.

One thing to note about zigbee sensors vs. BLE sensors. In general, you get far fewer samples per minute with any of the zigbee sensors. I have read some reports of ways to configure the TH01 to broadcast at a higher rate, but I am unsure about this. The picture below tries to show the difference between the Govee H5052 (my gold standard) and the Sonoff TH01. The Sonoff seems to increase the sampling when there is a larger temp or humidity change, but I think it is still far fewer than the Govee. In the picture, when humidity is stable the Sonoff is broadcasting about 5 samples every 10 minutes, where as the Govee is broadcasting about 6 samples per MINUTE, on the order of ten times more. Be careful to NOT equate ‘broadcasts’ with a sample reading by the units sensor. I do not have any way to determine how often each sensor is coded for sampling rate. In the link below, I did some back of the envelop musing about these topics.

1 Like

Dionaea grows rapidly after correcting humidity value. The value has been raised from 35%RH to comfortable for the plant 70%RH. New traps are being developed pretty well and it is going to bloom soon. Hopefully, I will stay alive :grinning:

Next steps are automated LED lighting and DI watering.

SonOff clearly states the sensor type in specifications: TH10 Si7021. It will be disassembled when it comes in order to assure that the sensor type is correct. Moreover, I will verify it by lab equipment and results will be posted here.

*subscribed :slight_smile:

Interesting, thanks! Sorry I do not have better example of the data between the Govee H5052 and Sonoff TH01 showing a larger set of changes. These units are in a two car garage, the example below shows two cycles of a gas clothes dryer operating. The two units track pretty well, however there is a measurable difference at times and perhaps you can see the broadcast difference better in this graph. Which unit is more accurate, that would require a calibrated pro unit to compare these against, out of my HA budget.

Your other two options are to buy a rather expensive pro unit or build something with esp8266 or esp32 and a quality sensor from adafruit or sparkfun.

Farming electronics is an interesting, ‘growing’ :wink: and useful area for both professional and home automation work. I live part time in a town with a university that is doing some interesting IoT farm work and also the area is one of the largest professional marijuana growing areas, both have offered some interesting talks on their work.

Good ‘growing’!

1 Like

David,

Just for the history.

1 Like

Wow! Interesting!

I did just realize that we are talking about two different Sonoff devices. I am using the TH01 which is a Zigbee temp/humidity sensor. You are planning on a TH10/TH16, which is a more capable wifi based unit. I just read a bit about it’s specs, an interesting and useful device.