Inconsistent temp / humid readings across sensors


#1

I am testing my monitor / control system for the first time with multiple grove temp sensors: 2 DHT Pro and one digital temp sensor (board type, not the 2-wire). Because of the cable-length issues I’ve encountered (separate thread under this forum), I am using 20cm connections. This puts all 3 sensor devices in very close proximity. In theory, all should read within “tolerance” of one-another.

First observation: temp differed between the 2 DHTs by 1 degree; humidity differed by 4%. That seems “within”.

Second observation: I’m tracking min and max DHT readings, updating those on each cycle (about 1 minute). Min humidity values diverged sharply, suggesting at least one “odd” reading. The outlier was ~ 10% or more lower.

The digital temp board is giving me readings of 70F, while the DHT’s are reading ~80F. That’s way out IMO.

Anybody used multiple temp / humidity sensors with the GrovePi? Is this similar to what you’ve experienced?


#2

Hi,
The DHT sensors are used universally everywhere and should give you pretty consistent values. We don’t have multiple sensors with us so cannot say much about the behavior. Even Adafruit sells them and you should find some information about them here https://learn.adafruit.com/dht?view=all. You should try searching on the Arduino or the Seeedstudio forums for if someone has faced a similar problem.

If you think that the problem is with the sensor itself, you should ask Seeedstudio about the problem.
-Karan


#3

Over the couple of weeks that these sensors have been running (still side by side due to the problems discussed separately with cable lengths over 20 cm), the humidity readings have converged. Temp and humidity readings from both sensors now are identical or differ by a small amount that makes sense given the tolerance specs of the devices. Although I do not have a benchmark device against which to verify these sensors, it’s looking like there is a burn in period - esp on the humidity side - before the readings stabilize.


#4

Further update: When I posted my previous update, the environment was very humid (99.1%). I probably concluded “reading the same” under outside of tolerance conditions. I think accuracy declines at very high and low limits. Over subsequent days of lower humidity(60% range), I’m observing about a 4% difference in humidity between the two sensors, while temp readings are the same or 1deg F different.

1deg F difference seems to be within sensor tolerance as I understand the DHT Pro specs. 4% difference in humidity seems a little outside of specs, but it is much better than the original 10% difference that I reported.

General question for anybody who has used the DHT in “dirty” environments (i.e. other than a clean lab bench, so basically like reading outside where there is normal dust, not seriously messy or excessively dusty): have you noticed changes over time in the DHT readings vs any other benchmark device? I’ll raise this on other forums (Adafruit, etc.) also.

My 2 sensors went into service a few weeks apart. Doesn’t seem that environmental factors like normal dust should impact sensor readings so quickly. Looking for insights. At some point I’ll also get some additional DHT sensors and swap out both as a test.


#5

Hi,
We have not done a stress testing like you are doing on the DHT sensors. Let us know what you see when you swap the sensors.
-Karan