Announcement

Collapse
No announcement yet.

mcsTemp stops recording Temp08 readings

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    mcsTemp stops recording Temp08 readings

    Intermittently and it seems after my system has been running without a restart for a close to a week, mcsTemp stops recording Temp08 readings (in devices and in database) and I need to stop and start Homeseer to get it working again. I checked the IO Stream prior to restart and normal readings were being delivered.

    I went back through my logs and found 3 occurances since Dec 18th (not enough patience to keep looking), lasting from several hours to over 3 days before I noticed it or did a regular maintenance restart, which I try to do every 7 - 10 days whether I need it or not.

    It happened before too but I haven't had time to investigate the various snags I've been running into over the past couple of months.

    #2
    Do you have anything in the logs that might indicate a problem? I can log heartbeat type info at some stages, otherwise I do not know how to look into it.

    Comment


      #3
      I noticed this morning readings stopped being recorded nearly 4 days ago. Here are the DB records:

      112070 06/01/2005 10:58:16 AM 3400 7400 3400 2010 2080 2090 -1210 -1220 2180 2180
      112071 06/01/2005 10:59:25 AM 3400 7400 3400 2000 2080 2090 -1210 -1220 2180 2180
      112072 10/01/2005 7:42:42 AM 3200 7400 3400 2060 2270 2270 -160 -240 2230 2210
      112073 10/01/2005 7:43:42 AM 3100 9700 3200 2070 2270 2280 -160 -240 2210 2220

      Restart at 7:42 am today occured because I restarted HS.

      Here's the HS Log around the time readings stopped being recorded. All that was happening at the time was someone hanging around our front door for a minute or so, then leaving the area...

      06/01/2005 10:59:04 AM~!~X10 Received~!~A15 (Front Door Motion Sensor) A On
      06/01/2005 10:59:14 AM~!~X10 Received~!~A15 (Front Door Motion Sensor) A On
      06/01/2005 10:59:32 AM~!~X10 Received~!~A15 (Front Door Motion Sensor) A On
      06/01/2005 11:02:02 AM~!~X10 Received~!~A15 (Front Door Motion Sensor) A Off

      Comment


        #4
        Running version 4.32.3

        In the excerpt of DB records below, notice the "ThreeHum" value changes. From the IO Stream the value 61 is actually the temp08 reading for "TwoHum", not "ThreeHum". TwoHum has not changed in device/DB from 52 since "Yesterday at 1:05:11 PM". Note that while I do get a few "???", these entries reflect temp08 readings of 32, 61, 31 consistently.


        ID SampleDate OneHum TwoHum ThreeHum
        122836 17/01/2005 8:43:14 PM 3200 5200 3100
        122837 17/01/2005 8:44:14 PM 3200 5200 6100
        122838 17/01/2005 8:45:14 PM 3200 5200 3100
        122839 17/01/2005 8:46:14 PM 3200 5200 6100


        More examples (the IO stream data for these are long gone). This one shows either OneHum or ThreeHum being recorded in the TwoHum column once in a while. (I'm pretty sure TwoHum should be in the 5000 range throughout)

        ID SampleDate OneHum TwoHum ThreeHum
        120907 16/01/2005 12:05:21 PM 3400 3400 3400
        120908 16/01/2005 12:06:19 PM 3400 3400 3400
        120909 16/01/2005 12:07:19 PM 3400 5300 3400
        120910 16/01/2005 12:08:21 PM 3400 5300 3400
        120911 16/01/2005 12:09:22 PM 3400 3400 3400
        120912 16/01/2005 12:11:32 PM 3400 3400 3400
        120913 16/01/2005 12:12:23 PM 3400 3400 3400
        120914 16/01/2005 12:13:19 PM 3400 5200 3400
        120915 16/01/2005 12:14:19 PM 3400 5200 3400
        120916 16/01/2005 12:15:25 PM 3400 3400 3400
        120917 16/01/2005 12:16:27 PM 3400 3400 3400
        120918 16/01/2005 12:17:27 PM 3400 3400 3400
        120919 16/01/2005 12:18:20 PM 3400 3400 3400
        120920 16/01/2005 12:19:20 PM 3400 3400 3400
        120921 16/01/2005 12:22:06 PM 3400 3400 3400
        120922 16/01/2005 12:22:18 PM 3400 5200 3400
        120923 16/01/2005 12:23:20 PM 3400 5000 3400

        Comment


          #5
          I'll look to see what I can figure out. I know that some versions of the Temp0? place multiple humidity readings on the same line and other versions deliver them with one per line. I have two humidity sensors, but only one connected to the Temp05.

          Comment


            #6
            Attached is a version that tracks a humidity sensors through the various processing stages. From this it should be possible to see what is lost along the way. The output will go into the \Data folder. It also has the correction for the second trigger 1 minute after the first.
            Last edited by Michael McSharry; January 29, 2005, 04:44 PM.

            Comment


              #7
              here's the log file after a handful of reading cycles
              Attached Files

              Comment


                #8
                The three humidity sensors all appeared to consistenty written to the correct database location. Let it run for awhile and look into the database to find a time such as you saw before when the values appeared to be mapped into the wrong sensor.

                I did not look back at the source, but I suspect the calibration is not happening because of the "%" in the humidity number.

                Comment


                  #9
                  Right now the first sensor on my IO stream is not being updated in device and db (situation started a couple of hours ago according to last device update).

                  IO Stream:

                  FRI 15:10:27
                  15:07:28 | Reading Sensors...
                  15:07:28 | Humidity #01[410000004FEE3A26]=19%
                  Humidity #02[800000004FE5AA26]=47%
                  Humidity #03[820000004FFCE126]=21%
                  Temp #01[D60000004FFF2626]=21.31C
                  Temp #02[3D000800412B7A10]=22.50C
                  Temp #03[410000004FEE3A26]=22.12C
                  Temp #04[DF00080046A0E810]=-05.32C
                  Temp #05[800000004FE5AA26]=-06.12C
                  Temp #06[820000004FFCE126]=22.00C
                  Temp #07[7D00080048F1CB10]=22.37C
                  >

                  matching mcsTempHumidity.txt log entry. Value of 104 for sensor 1 has been stuck at that for a few hours now (s/b 19).


                  28/01/2005 3:07:28 PM | Calibrate Device=R1,Value=100,Bias=4
                  28/01/2005 3:07:28 PM | Device=R1,SensorID=01,Code=410000004FEE3A26,Value=104
                  28/01/2005 3:07:29 PM | Calibrate Device=R2,Value=47%,Bias=4
                  28/01/2005 3:07:29 PM | Device=R2,SensorID=02,Code=800000004FE5AA26,Value=47
                  28/01/2005 3:07:35 PM | Calibrate Device=R3,Value=21%,Bias=2
                  28/01/2005 3:07:35 PM | Device=R3,SensorID=03,Code=820000004FFCE126,Value=21
                  28/01/2005 3:07:37 PM | Database Field=OneHum,Device=R1,Value=10400
                  28/01/2005 3:07:37 PM | Database Field=TwoHum,Device=R2,Value=4700
                  28/01/2005 3:07:37 PM | Database Field=ThreeHum,Device=R3,Value=2100
                  28/01/2005 3:07:37 PM | Calibrate Device=R4,Value=21.31,Bias=-1.5
                  28/01/2005 3:07:39 PM | Calibrate Device=R5,Value=22.50,Bias=-0.4
                  28/01/2005 3:07:43 PM | Calibrate Device=R7,Value=-05.32,Bias=1
                  28/01/2005 3:07:45 PM | Calibrate Device=R8,Value=-06.12,Bias=1
                  28/01/2005 3:07:47 PM | Calibrate Device=R9,Value=22.00,Bias=0.45


                  Here's another IO stream - logfile pair. Not a pretty stream. Is formatting inconsistency what's causing my update problems?


                  15:16:28 | Reading Sensors...
                  Humidity #01[410000004FEE3A26]=19%
                  Humidity #02[800000004FE5AA26]= ???
                  15:16:35 | Humidity #03[820000004FFCE126]=21%
                  15:16:35 | Temp05 Error= ??? Response
                  Temp #01[D60000004FFF2626]=21.43C
                  Temp #02[3D000800412B7A10]=22.50C
                  15:16:40 | Temp #03[410000004FEE3A26]=22.18C
                  Temp #04[DF00080046A0E810]=-05.57C
                  Temp #05[800000004FE5AA26]=-06.50C
                  15:16:55 | Temp #06[820000004FFCE126]=21.81C
                  15:16:55 | Temp #07[7D00080048F1CB10]=22.25C
                  15:16:55 | >


                  28/01/2005 3:16:29 PM | Calibrate Device=R1,Value=100,Bias=4
                  28/01/2005 3:16:35 PM | Device=R1,SensorID=01,Code=410000004FEE3A26,Value=104
                  28/01/2005 3:16:36 PM | Calibrate Device=R3,Value=21%,Bias=2
                  28/01/2005 3:16:36 PM | Device=R3,SensorID=03,Code=820000004FFCE126,Value=21
                  28/01/2005 3:16:38 PM | Database Field=OneHum,Device=R1,Value=10400
                  28/01/2005 3:16:38 PM | Database Field=TwoHum,Device=R2,Value=4700
                  28/01/2005 3:16:38 PM | Database Field=ThreeHum,Device=R3,Value=2100
                  28/01/2005 3:16:38 PM | Calibrate Device=R4,Value=21.43,Bias=-1.5
                  28/01/2005 3:16:40 PM | Calibrate Device=R5,Value=22.50,Bias=-0.4
                  28/01/2005 3:16:44 PM | Calibrate Device=R7,Value=-05.57,Bias=1
                  28/01/2005 3:16:55 PM | Calibrate Device=R8,Value=-06.50,Bias=1
                  28/01/2005 3:17:28 PM | Calibrate Device=R9,Value=21.81,Bias=0.45

                  Comment


                    #10
                    % in humidity number

                    you suggested the % symbol in IO stream is preventing calibration but you don't say whether it's something you plan to address in mcsTemp, or something I would need to bring up with Midon Design. can you pls clarify?

                    Comment


                      #11
                      I had through I took care of it already. I'll look again.

                      What is the value you have for minimum valid humidity? If it something like 20 then the reading of 19 will be treated as 119 and then limited to 100.

                      Comment


                        #12
                        min humidity has been at 20. will change it to 15. although I probably need to raise the humidity in my house a bit, it appears 20 may be a little high as a minimum in my neck of the world.

                        it looks to me as though the bias is added when the humidity becomes 100 due to min humidity value.

                        Can you set it up to ignore the bias in this case?

                        DB entries below show humidity in my house (OneHum and Threehum) are hovering around 20, and 20 gets converted to 104 (bias is +4) for OneHum and 102 (bias is +2) for ThreeHum.

                        ID SampleDate OneHum TwoHum ThreeHum
                        139092 29/01/2005 7:13:39 AM 2100 7600 10200
                        139093 29/01/2005 7:14:35 AM 10400 7600 2100


                        Comment


                          #13
                          Just to clarify, bias is only added when humidity is below 20 (and changed to 100). It does not get added when humidity is above 20

                          IO stream:

                          07:29:25 | Reading Sensors...
                          Humidity #01[410000004FEE3A26]=21%
                          Humidity #02[800000004FE5AA26]=76%
                          Humidity #03[820000004FFCE126]=21%

                          Logfile trace

                          29/01/2005 7:29:25 AM | Calibrate Device=R1,Value=21%,Bias=4
                          29/01/2005 7:29:25 AM | Device=R1,SensorID=01,Code=410000004FEE3A26,Value=21
                          29/01/2005 7:29:26 AM | Calibrate Device=R2,Value=76%,Bias=4
                          29/01/2005 7:29:26 AM | Device=R2,SensorID=02,Code=800000004FE5AA26,Value=76
                          29/01/2005 7:29:35 AM | Database Field=OneHum,Device=R1,Value=2100
                          29/01/2005 7:29:47 AM | Database Field=TwoHum,Device=R2,Value=7600
                          29/01/2005 7:29:47 AM | Database Field=ThreeHum,Device=R3,Value=2100
                          29/01/2005 7:29:47 AM | Calibrate Device=R4,Value=21.75,Bias=-1.5
                          29/01/2005 7:29:47 AM | Calibrate Device=R5,Value=22.87,Bias=-0.4


                          DB entries (time didn't match up so I provided entries around the above timestamp):

                          ID SampleDate OneHum TwoHum ThreeHum
                          139107 29/01/2005 7:28:35 AM 2100 7600 2100
                          139108 29/01/2005 7:29:35 AM 2100 7600 2100
                          139109 29/01/2005 7:30:39 AM 2100 7600 2100

                          Comment


                            #14
                            I do not understand why the % is not removed. The following is the code applicable code segment. I did just add the first logtofile for greater visibility.

                            Function Calibrate(device As String, sValue As String, iType As Integer) As String
                            10 On Error GoTo Error_Exit
                            Dim i As Integer
                            Dim j As Integer
                            Dim devicecode As String
                            Dim units As String
                            Dim Value As String
                            Dim newValue As String
                            Dim nValue As Single
                            Dim strValue As String

                            20 strValue = Trim(Replace(sValue, "%", ""))
                            30 LogToFile "Calibrate OriginalValue=" & sValue & ",NewValue=" & strValue & ",EndAscii=" & CStr(Asc(Right(sValue, 1))) & "/" & CStr(Asc("%"))
                            40 devicecode = UCase(device)
                            50 newValue = strValue
                            60 For j = 0 To Calibrations - 1
                            70 If devicecode = BiasDevice(j) Then
                            110 LogToFile "Calibrate Device=" & device & ",Value=" & strValue & ",Bias=" & CStr(BiasValue(j))

                            I also changed the order of bias vs limit per your request. At this point I also removed the % before the calibration funciton is called.

                            In most cases I think a value of around 10 for min humidity would be reasonable. The sensor reporting should not be off by 10% and 10% is even dry for anywhere. The update is attached for your evaluation.
                            Last edited by Michael McSharry; February 1, 2005, 11:55 PM.

                            Comment


                              #15
                              Am a VB Script lightweight at best and am not sure what to suggest.

                              I did notice an error logged in ah.log tonight that may or may not be related to this and filtered my ah.log file to look for other abnormal entries.

                              Here's what has been logged since Jan 1st. (There was more of the first error before that). Not sure how long I've been running 4.32.4 but I think it may have fixed the first batch of errors.

                              01/01/2005 0:41 mcsTemperature UpdateForecastDevices 4 on line 230 Type mismatch
                              03/01/2005 6:33 mcsTemperature UpdateForecastDevices 4 on line 230 Type mismatch
                              03/01/2005 12:42 mcsTemperature UpdateForecastDevices 4 on line 230 Type mismatch
                              04/01/2005 10:43 mcsTemperature UpdateForecastDevices 4 on line 230 Type mismatch
                              07/01/2005 7:50 mcsTemperature GetPluginTriggers with Progress 275 on line 110 Object variable or With block variable not set
                              10/01/2005 20:55 mcsTemperature GetPluginTriggers with Progress 278 on line 110 Object variable or With block variable not set
                              15/01/2005 16:06 mcsTemperature GetPluginTriggers with Progress 275 on line 110 Object variable or With block variable not set
                              21/01/2005 6:32 mcsTemperature GetPluginTriggers with Progress 280 on line 110 Object variable or With block variable not set
                              30/01/2005 19:25 mcsTemperature GetPluginTriggers with Progress 279 on line 110 Object variable or With block variable not set


                              NOTE: haven't yet tried the version you provided (4.32.6). will let you know what happens after I do.

                              Comment

                              Working...
                              X