Discrepancies in readings

[clippermiami]

Attached are two images of sensors attached in a Bridged Vera environment. The first image is Vera 2 the Secondary Vera and the second image is Vera 2, the Secondary Vera. The Sensor Gateway is directly attached to Vera 2.

The readings are clearly inconsistent. As can be seen the temperature reading on Vera 2 are all the same which is not possible, one of the sensors is outside the others inside and they do not match the comparable values displayed on Vera 1. The humidity readings on Vera 2 are consistent with reality but do not track with the corresponding values on Vera. The Last Update times are consistent with reality and clearly show that the updates are occurring and being transferred across the interface.

Any ideas why this would occur?

[hek]

@clippermiami

Hmm.. I have actually no idea how a bridged setup works behind the scene. Not much documented on micasaverde. Really strange that some values gets updated and others not.
Are you sure it's not just UI that don't get updated? Are you using the values on the bridged device in scenes or PLEG?

[clippermiami]

Its not only the bridged environment that's causing me problems. Today the two UIs seem to be tracking but the numbers are impossible.

In the attached images note the humidity numbers. Node 1 is in the Study, air conditioned space so the 39% humidity while probably wrong is at least plausible. Node 3 is in the garage, not A/C space but the number is very high (as I type this I've looked at it again and now its 90% --- if that was true it would be raining in my garage

Node 6 is in the Kitchen, again air conditioned space and the chance that there is 78% humidity is impossible.

And again as I type this the temperature in the garage and study are both exactly 78.8 again, not very likely, one is inside in A/C the other essentially outside non A/C space.

The bottom line, the numbers just make no sense.

[UPDATE]
The Humidity in the kitchen sensor seems to have suddenly "popped" back to something reasonable at 43%. I say "popped" because in the space of two updates (120 seconds) it went from 78% to 62% to 43%. The garage is still sitting on 91% ... yesterday it was at 65% most f the day and evening. The study and garage temps are still identical at 78.8, again not very likely.

Very odd indeed.

[clippermiami]

@hek re: the Bridged discrepancies.

Ultimately my intent would be to use the values as input to Program Logic Event Generators (PLEG) running on the Primary Vera to assist in control of lights, A/C, etc. All the PLEG house management controls run in Vera 1; just a few Plug Ins, smoke detectors, the Arduino sub-system on Vera 2 right now. So obviously the numbers need to update correctly.

I've updated this post sith some information from this morning that reflect what appear to be impossible numbers, such as 91% humidity in the garage. Its South Florida but it isn't THAT humid at 9AM unless it pouring down rain

I've done some experimenting with other things in the bridged environment and there are a few APP Plug ins that don't work properly at all. As you note there is not a lot of information on the details of the bridging interface so its difficult to discern what's correct or not.

As of this morning the numbers on the two UIs seem to be tracking ...some of them are wrong but they are at least wrong on both UIs this morning

[BulldogLowell]

@clippermiami said:

I've updated this post sith some information from this morning that reflect what appear to be impossible numbers, such as 91% humidity in the garage. Its South Florida but it isn't THAT humid at 9AM unless it pouring down rain

Remember regarding humidity, it is the Relative Humidity or more accurately "how close is the current temperature to the dew point temperature."

Living also in paradise, we experience very high humidity levels early in the morning when the temperature has dropped, but the moisture in the air has not really changed. I can see temperatures of 78F with 90% plus humidity every morning. Then as the temperature rises, the RH drops. I would show you my RH logs but I don't have vpn access to my network right now.

Also, indoors... if the temperature spikes or lowers, you will see a corresponding change in the air's RH, even though its moisture content of the air is unchanged.

Finally, don't underestimate the humidity sensor's sensitivity, it can detect a sudden drop or increase very quickly, in my experience. It makes it useful for detecting if the shower is on and the exhaust needs to be turned on.

[clippermiami]

@BulldogLowell First, I did not realize they were that fast ... mine are not located whee they would experience a rapid local change do I poisoned they were more allow acting devices. As to the very high levels I saw they were effectively sudden changes, it hit 90% in a matter of minutes.

I still have a hard time accepting that the inside temperature and the temperature the garage were exactly same, to decimal point. Several hours later I saw exactly the same values for the study and the kitchen. I just find it hard to believe that these spaces are experiencing same temp the same time.

But I have nothing other that feeling on which to base this so I guess I'll have to go back to sleep until I can gather more data

[BulldogLowell]

@clippermiami said:

@BulldogLowell First, I did not realize they were that fast ... mine are not located whee they would experience a rapid local change do I poisoned they were more allow acting devices. As to the very high levels I saw they were effectively sudden changes, it hit 90% in a matter of minutes.

I still have a hard time accepting that the inside temperature and the temperature the garage were exactly same, to decimal point. Several hours later I saw exactly the same values for the study and the kitchen. I just find it hard to believe that these spaces are experiencing same temp the same time.

But I have nothing other that feeling on which to base this so I guess I'll have to go back to sleep until I can gather more data

I'm not saying that your temp readings are correct, I am trying to expelling the high humidity/temperature relationship.

You may want to consider logging the temps and looking at them.

Also... how often are you transmitting temperatures? I remember reviewing that there was a comparison of two floats, I wasn't sure if it was reporting changes more frequently because float math several decimal points out lacks precision. So a comparison of:

if (myOldFloat != myNewFloat)  // I think the precision of this cause a lot more not equal results and we don't really care beyond 10^-1

may yield a lot more radio traffic than if you multiply them by 10, cast them to integers, and then compare them.

I'm not sure on this but I suspect. I'm interested in opinions here.

I have my temperature/humidity sensors updating on a timer, recognizing th

[clippermiami]

@BulldogLowell Interesting point. I soups they COULD be the same but it just makes me uncomfortable for a room in A/C space to be 80.6F and the garage to be exactly the same all day with an outside temperature of 87F all day It just doesn't seem rational

[Yveaux]

@BulldogLowell As you don't have access to your vpn I decided to share a screenshot of my weather station data to illustrate the relationship between temperature & humidity:

This is one week of data (sorry, the axis labels are unreadable). Top one is temperature (in C), light blue is RH.
You can see it's almost the inverse of temperature. It goes up every night to go down during the day again.

Agree that these things (this is an SHT11, if I remember right) are very sensitive. If you breathe over it, it will show directly and when it starts to rain it usually responds faster than my rain sensor (e.g. the small 'hump' in the light blue RH curve at approx. 20% of the curve is caused by some showers)

[BulldogLowell]

@Yveaux said:

You can see it's almost the inverse of temperature.

yes, sort of its negative slope...

thanks for posting that.

[clippermiami]

I understand the significant "inverse" relationship between temperature and RH, that's not the problem.

The problem is the numbers being displayed in Vera and the fact that the RH could not change over the ranges I'm seeing that fast (a plausible 60pct to an implausible 90 pct in a few minutes) unless you dumped the sensor in a bucket of water ; and it is not rational that temperature numbers can be EXACTLY the same (80.6 F) for 14 hours in two entirely different kind of spaces, one inside, insulated and under A/C (kitchen), the other un-insulated, essentially outside (garage)

What does make sense is that for some reason it appears the Vera Plugin is not assigning incoming values to the correct sensor all the time. Unfortunately it is beyond my meager understanding of Vera and its code and MySensors plug-in and its code to debug the phenomenon.

[BulldogLowell]

@clippermiami

that may be a bridged vera problem.

I have bridged devices but have mySensors on the parent not the child.

[clippermiami]

@BulldogLowell that's a rational possibility. I have the sensors on the secondary Vera. Both Vera (what's the plural of Vera ) show the same illogical numbers however.

This morning I've managed to get then all reporting different numbers now. I reset the three operating nodes at the same time (they are all battery operated so I brought then all to one place) and at the same time reloaded the secondary Vera. When I moved the nodes back where they belong on the next reporting cycle everything appears to make sense ... I'm not getting the same numbers any longer. It looks like "things got wound up around the axle" somehow and a little electroshock therapy was in order.