Haiku Temp/Humidity Sensor hiccups

[pct88]

I recently turned on temperature and humidity sensor logging in HaikuHelper and find that there are occasional data dropouts from my HAI Extended Range auxiliary sensor.  This causes the temperature logged to suddenly drop to very low values, then just as rapidly return to where it was.  Is this a typical failure mode for these sensors?  Is there a way to configure HaikuHelper to average across such intermittent failures?
 
I am only seeing this on one of two Extended Range sensors.  The four OmniStats are also logging properly without spurious jumps in data.
 
I would post an example of the data dropouts but I am not allowed to paste into my messages apparently.
 
Thanks

 

[lupinglade]

How long does the reading change for? What value does it drop to?

I would check wiring also.

 

[pct88]

Is there a way for me to view the raw temperature/humidity data on the Mac Mini to see?
 
I have swapped out the pairs to the sensor without fixing the issue.  Next I guess I have to replace the sensor.

 

[lupinglade]

Go to the Web > Weather & Graphing tab in Haiku and turn off the Ramer-Douglas-Peucker option. This will show the exact values on the graph.

 

[pct88]

I have e-mailed you sample graphs with the Ramer-Douglas-Peucker option on and off.

 

[pct88]

I continue to see occasional large spikes in the data.  I did replace one of the problematic HAI sensors which decreased the problem, but it does still show drops to zero or spikes to high values.  
 
Is anyone else seeing this sort of problem?  Perhaps I just have a batch of faulty sensors.  
 
Given that these HAI sensors are not designed to detect fires, etc...  is there a way to have the HH software ignore or smooth sudden spikes in temperature and humidity data?  I would like to use these sensors to control cooling/dehumidification fan in my attic, but I don't want the fan kicking on every time one of these spurious spikes comes by.  I have confirmed that the wiring is solid to the sensors.  Other problems I had seen a year ago were resolved when I replaced the OPII board.
 
Here is a sample from the last week:
[sharedmedia=gallery:images:736]

 

[ano]

It is very possible that your sensors are picking up noise. Did you use a twisted cable to connect to your sensors?  Also make sure the sensor wires DO NOT run parallel to 120V/220V cables that could induce large noise spikes. Always only run the sensor cables perpendicular to these high-voltage lines, and separate the two cables if you can.

 

[lupinglade]

Yes, I think the issue is noise. We do filter extreme spikes, but smaller spikes like on the graph you attached would be harder to filter out without losing accuracy.