Hello All,
I have had my emonTH unit up and running for a couple days and decided to put out a hobo data logger I have to compare accuracy for the temp and humidity readings in my house. Both are taking samples on a 60sec basis. Over the couple hours I had the hobo running it captured the thermostat cycle pretty well from 66.8F to 65.8F and read the humidity level at an average of about 42%RH (Im in heating mode at my house. Outside is about 52F, 96%RH). The emonTH unit has been reading between 64.8F and 65.8F, fluctuating in a similar manner to the hobo readings with the thermostat cycle, but not in quite as much detail (somewhat hard to see without zooming in closer on the multi graph). It is reading the humidity at an average of about 38%RH.
So my question is, is this pretty good for the emonTH unit? Screen shots of the hobo readout and emonTH multigraph below.
Also, because my brain thinks in Fahrenheit, the process for my temperature input data is (X 0.1) (X 9) (X 0.2) (+ 32) then log to feed. For humidity just (X 0.1) and log to feed.
Dashboard with multigraph: http://emoncms.org/pksmith
Re: EmonTh Accuracy
My experience with the EmonTh is that is registers the same temperature as my heating thermostat mounted besides it. Although the thermostat only shows steps of halve degrees Celsius.
Re: EmonTh Accuracy
Update: I continued to let the HOBO logger run for the rest of the week. Looking at the plots of the HOBO data and the emonTH temp and RH readings it seems that they both have the same profile, the emonTH appears to be reading a few degrees and RH% values lower than the HOBO. I have attached the plots for comparison. I would tend to believe the HOBO logger has the 'more accurate' readings, but I am pleased with the emonTH's performance. Not to far off from the HOBO.
Also, both the emonTH and HOBO loggers were placed within a few inches of each other by my thermostat.
Re: EmonTh Accuracy
FWIW, I've been running an emonTH and an emonTX at the same position for a while and both were really close (about 0.1 or 0.2°C difference).
Then I moved the emonTH to the other side of the room, about 2 meters high, while the TX was close to the floor, and there was a huge difference. The temperature on the floor has much less variations than the temperature under the ceiling. The effect of the position is much more important than I expected. And it is not just an offset. There was much more dynamics at ceiling level. The ground is not affected by the heater that much.
I know you put all your sensors at the same place, but I'm mentioning this anyway for anyone wanting to run comparisons.
Re: EmonTh Accuracy
Interesting observation. We I do industrial energy audits we often check the temperature gradient from the ceiling to the shop floor and recommend destratification fans to take advantage of the warm air at the ceiling during winter months. I have also found that the ceiling fans in my house have winter and summer modes (the direction the fan spins can be reversed in the winter to drive some of the warmer air down).
Re: EmonTh Accuracy
Update: I have my second emonTH up and running now. I have noticed that my original unit (used to make the graphs above) is reading about 2-3 deg. lower than my new unit. The new unit is reading exactly what my thermostat reads, the original has always been 2-3 deg. lower. I decided to just add a process adding 2 degrees to the raw data in the input config setup to account for this. Both units seem to bee reading RH within 1 degree of eachother.
Re: EmonTh Accuracy
Out of curiosity, what happens if you exchange the RH/temp sensors on both modules ?