[AllanG]

Although I've had no problems at all with the SHT21 sensors I've used since putting the WeatherDuino into active service last December, I decided last week that I'll give the 31D a try so ordered one from "Closed Cube" here in the UK. Price was GBP10 which is just about double the cost of the Chinese SHT21 but still fairly cheap I thought. It arrived yesterday - and the updated software is available today so excellent timing!!

I'm about to make some changes to the entire station so hope to have the new sensor operational sometime next week, ideally with SHT21 as a backup and comparison. Some pictures of the SHT31D:

SHT31D sensor

SHT31D sensor

You can't tell from the photo of course but the board is about twice the size of the Chinese SHT21 and seems very well made.

Allan.

[AllanG]

I have a spare TX module that I'm using for bench testing my solar sensor, so today I uploaded v2.0_b010 software and connected the SHT31 sensor. As I hoped and expected, it appears to work perfectly.

I then decided to compare all three outdoor sensors - HTU21, SHT21 and SHT31.

It certainly wasn't a "scientific" test but all three sensors were tested indoors for just a minute or so each, the whole job taking no longer than 10 minutes, with no temperature and humidity changes in the (closed) room, according to the room thermometer.

The results were quite interesting. Temperature readings were all within one degree. Humidity reading were a little different. The HTU21 showed the greatest variation, the SHT21 next and the HTU31 showing the least variation. I'm now looking forward to seeing some "real life" comparisons!

Here are the readings I took:

SHT31:
Sensor 0 - Temp: 23.96 Hum: 51.49
Sensor 0 - Temp: 23.88 Hum: 50.28
Sensor 0 - Temp: 23.89 Hum: 50.38
Sensor 0 - Temp: 23.92 Hum: 50.62
Sensor 0 - Temp: 23.88 Hum: 50.13
Sensor 0 - Temp: 23.83 Hum: 50.48

SHT21:
Sensor 0 - Temp: 23.36 Hum: 54.87
Sensor 0 - Temp: 23.22 Hum: 55.11
Sensor 0 - Temp: 23.10 Hum: 56.10
Sensor 0 - Temp: 23.13 Hum: 56.33
Sensor 0 - Temp: 23.05 Hum: 55.72
Sensor 0 - Temp: 23.04 Hum: 55.69

HTU21:
Sensor 0 - Temp: 23.59 Hum: 48.38
Sensor 0 - Temp: 23.57 Hum: 49.52
Sensor 0 - Temp: 23.50 Hum: 47.90
Sensor 0 - Temp: 23.50 Hum: 47.57
Sensor 0 - Temp: 23.53 Hum: 48.93
Sensor 0 - Temp: 23.54 Hum: 49.02

Allan.

[werk_ag]

Hi Allan,

Thanks for this test, and by your confirmation that the code to read the SHT31 is working as expected.
Before release, I tested an SHT31 during a full week, with a dedicated TX board, doing reads every 10 seconds, but I didn't any comparison with other sensors.

[hvalentim]

There's a saying about the precision of location technology which goes along something like "if you own one GPS you are sure to know where you are, if you have two you start having doubts and if you try to use three then you are lost".

I guess the same is valid for them temp./hum. sensors. The more you use, the less certain you are about the exact values...

As for the SHT31D Board, it does look like excellent value for money.

[AllanG]

There's a saying about the precision of location technology which goes along something like "if you own one GPS you are sure to know where you are, if you have two you start having doubts and if you try to use three then you are lost".

True, but only if all 3 GPS units give different locations? If all 3 give the same I would be VERY confident I know where I am....

I understand and agree with your point about T & H sensors though as I have no way of knowing the "real" room temperature and humidity levels when I did the very simple test. I was reassured that the temperature indicated by all 3 sensors was very close, so on the basis of your GPS analogy I'm confident it was quite accurate.

RH is apparently very difficult to measure accurately using this type of sensor so I wasn't surprised at a greater variation between the 3 devices. What I found interesting though was the SHT31 showed the least variation in its own readings which I think may reflect the greater accuracy according to the datasheet. So, in "real life" will it give me accurate RH figures at my location? It may or it may not, as we have agreed I have no way of knowing. I think I may be getting closer to "accurate" though....

Allan.