Crazy rain gauge

[kas]

Michael,

Everything has been OK for months
Since 2 weeks, my rain gauge is going crazy
On August 29th, I made a rough measurement with a graduated glass and found 35mm rain
During the same period (24 hours), HomeSeer indicates 3700mm rain !!!

This is an excerpt of the log file:
(normal interval 30 seconds as been raised to 300s for this test)

29/08/2007 21:28:42 ~!~mcsTemperature Debug~!~DS9x9x Port USB2 Device = DS2450 ED00000000D5DD20
29/08/2007 21:28:42 ~!~mcsTemperature Debug~!~Calibrate 68 with bias 0
29/08/2007 21:28:42 ~!~mcsTemperature Debug~!~DS9x9x Port USB2 Device = DS1920 7A0008001417B510
29/08/2007 21:28:43 ~!~mcsTemperature Debug~!~Calibrate 14.5625 with bias 0
29/08/2007 21:28:43 ~!~mcsTemperature Debug~!~DS9x9x Port USB2 Device = DS1920 B4000800BB727D10
29/08/2007 21:28:44 ~!~mcsTemperature Debug~!~Calibrate 16.625 with bias -0.3
29/08/2007 21:28:44 ~!~mcsTemperature Debug~!~DS9x9x Port USB2 Device = DS2438 430000009E71D626
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~DS9x9x Port USB2 Device = DS1990A CB0000002645BF81
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~DS9x9x Port USB2 Device = DS2423 D1000000019FF11D
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~Wind Speed Counter=755552, Prior Counter=753318, Interval=304 seconds
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~DS9x9x Port USB2 Device = DS2423 2100000007E3471D
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~ComputeWater Count=15075, Type=29, Device1=R5, Device2=R9, Interval = 300
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~Water Counter=15075, Prior Counter=15062, Interval=304 seconds
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~R5 Water Delta Seconds=304, Prior Value=15062, Current Value=15075, Rate=70.10905
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~ComputeWater Count=18589, Type=4, Device1=R10, Device2=R11, Interval = 300
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~Water Counter=18589, Prior Counter=18576, Interval=304 seconds
29/08/2007 21:28:45 ~!~mcsTemperature Debug~!~R10 Water Delta Seconds=304, Prior Value=18576, Current Value=18589, Rate=1018.574

Please let me know if having two water counters (R5, R10) is a normal situation
Should I delete devices and reinstall
Any other clues?
Possible hardware failure?

I enclose a longer portion of the log file

Thanks for your help

yves

[Michael McSharry]

A DS2423 contains 2 independent counters. If you have two different sources then each can be configured. I suspect most with a rain gauge have only one source and they may have both inputs connected to that same rain bucket source. In HS I would only use one of the two if that is the case.

Your output data shows a 3514 counter input delta between the two measurements. This would correspond to 3514 mm in Homeseer plus whatever had been accumulated since midnight.

If you rain bucket counts at a different rate than 1 mm/count then use the rain counter calibration setup in mcsTemperature to indicate the volume/count relationship.

[kas]

Oops... I missed something
3514 is the difference between R10 and R5 (18576-15062)
does it make sense ??

Again, everything was OK last month, and I didn't modify the rain counter calibration...

[Michael McSharry]

I missed that these were different devices. You cannot have the same type of device for both counters as the .ini setting that maintains history is keyed only to the serial number of the sensor. Change one of the two input types so there are different.

[kas]

Michael,

R10 (Rain today) and R5 (water flow) are not same devices
Should I change R5 to an other type?

[Michael McSharry]

The log shows that the rain counters are being processed correctly for the period of the log. It looks as if the smoothing rate is weighted toward the history so it will be slow to respond as shown at 21:54 where the counter values did not change over 5 minutes, but the rate was still at 578 mm/hr.

The mcsTemperature.ini contains "history" keys that contain the midnight reading, last reading and a few other entries related to the counters. It is the last of these values is the counter reading at midnight. If this value is very low then it will show a large daily reading.

There is also the rain counter calibration value that is multiplied by the delta in the sensor readings to arrive at a value that is placed in the HS device string. This value multiplied by 100 is the number placed in the HS device value. A calibration value of other than 1 will can cause the HS device to be larger than expected.