No announcement yet.

moisture out of sync after cycle

  • Filter
  • Time
  • Show
Clear All
new posts

    I see two issues with your debug data. There is no response from the Rain8Net. This should be showing up as Messages as well as a Fail status for the zones that use the Rain8Net. In the debug it is " Zone R1 has been declared failed after multiple attempts due to timeout in expected response". My experience with the Rain8Net is that this is usually caused by using COM adapters that do not drive the voltage all the way down to 0V, but only down to 0.5V or so. What I do not understand it why this affects the status request, but does not affect the on/off commands. Could be the bit pattern sensitivity. Without a hardware status mcsSprinklers needs to guess the actual state and develop control heuristics.

    I looked at R1, but likely the same is occurring with other zones. Your setup option selected is to increase the moisture level by the delta from the current level and the trigger level. Your trigger is 0.5. It appears irrigation has not occurred for awhile and the moisture level had gone down to -0.01 by the time the R1 was allowed to run at 3:45 AM. When R1 completed at 4:20 the moisture level was adjusted by 0.5 to 0.49. Note that 0.49 is still below the threshold of 0.5. When R4 completed at 4:35 AM R1 again started because of the threshold being below 0.5.

    The failure mode was introduced at 5:10 when mcsSprinklers heuristic did not recognize that on the second cycle of the day it needed to turn OFF R1 after 35 minutes. It only handled the failure mode for the first irrigation cycle of the day. This means that the moisture level was not adjusted at 5:10 and another cycle started when other zones finished.

    If the Rain8Net reported status, the moisture level not decreased so much so two cycles ran in one day or if the moisture was set to 100% on the first cycle then this scenario would not have been observed.

    I updated the failure mode control heuristic in Either extract from the zip or use the Other Page facility to install version updates.


      Thank you Michael, I will try v.29.

      Yes I use a Digi Edgeport USB/RS232 converter and get status failure message almost daily typically at midnight when mcsSprinklers restarts but never have problem turning zones on/off. I did some test using config utility from rain8, it got status ok. Perhaps the status problem is intermittent. When I click on “General” tab, will it force mcsSprinklers to refresh status from rain8net?

      In my setup I found if forecasted rains don’t materialize it gets too dry, so I chose to increase moisture by delta trying to cause it to water more in subsequent days to compensate. Do you suggest a different approach? With my current configuration, it doesn’t water enough in hot sunny days. How should I adjust settings to make it water more often in hot sunny days and less in cooler cloudy days? I use UV index from Accuweather.


        The communications test on the page where the Rain8Net is setup will run each valve in sequence continuously until stopped and display error counts. I developed a simulation of the the Rain8Net where I could tweak communication parameters. The mcsSprinklers timing has considerable margin. I can run the communication test all day long with the simulation and have no errors. I had an old computer with real RS232 voltage swings and did not have problems then either. I have had variable success with IP/serial and USB/serial interfaces. I did have a mod from Warren (WGL) to improve the zero level. I don’t remember if it helped or not.

        There are two dials that you can turn to change the moisture loss rates. On the top of Weather Setup is a solar intensity factor. Nominal is 1.0. A value of 1.1 will increase the sensitivity of the solar source which is UV in your case. This will speed up moisture loss more on sunny days and increase it at a lower rate on cloudy days.

        Zone setup has a soil factor which is a composite of the ability of the soil to hold moisture and the propensity of the plant to loose moisture through its leaves. The default is 0.75. If it is decreased the the moisture loss rates will decrease for the zone.

        The threshold on the Area page is another. Raising the threshold increases the irrigation frequency. The caution with this one is that you want roots to grow deep and if they always have surface water they will be lazy and not develop a tolerance to short term weather variations. Of course if you have sand for soil then water will not be retained anyway.


          I can pull status from rain8net using Warren's config utility. Also run communication test on mcsSprinklers for an hour without any error, but almost daily at midnight after mcsSprinklers auto restart, I'd get an email about a valve failure (typically only a single valve failure but on different days the email alert may show different zone failing). Is the zone status test at startup more stringent than those run in communication test?


            There is no specific test done at startup. There may be a status polling to get the initial status. The Serial IO page will show what actually was communicated. Typical protocol used by mcsSprinklers is four attempts before declaring a failure.


              Earliest entry on Serial IO page starts at 4am when my normal watering cycle starts, but I always get email alert about valve failure at midnight when mcsSprinklers auto restarts.

              Below is earliest entries on Serial IO page:

              7/28/2020 4:26:01 AM | 94015781 | NET(0) OUT | 40, 01, 31
              7/28/2020 4:26:06 AM | 94020796 | Net Send Timeout
              7/28/2020 4:26:06 AM | 94020796 | NET(0) OUT | 40, 01, 31
              7/28/2020 4:26:06 AM | 94020812 | NET IN | 40, 01, 31 Valve Command Acknowledge
              7/28/2020 4:26:09 AM | 94023328 | NET(0) OUT | 40, 01, 36
              7/28/2020 4:26:09 AM | 94023343 | NET IN | 40, 01, 36 Valve Command Acknowledge
              7/28/2020 4:26:11 AM | 94025875 | NET(0) OUT | 40, 01, 38
              7/28/2020 4:26:11 AM | 94025890 | NET IN | 40, 01, 38 Valve Command Acknowledge
              7/28/2020 4:26:14 AM | 94028421 | NET(0) OUT | 40, 01, F0
              7/28/2020 4:26:14 AM | 94028437 | NET IN | 40, 01, A1 Status Response
              7/28/2020 4:27:03 AM | 94077937 | NET(0) OUT | 40, 01, 38
              7/28/2020 4:27:03 AM | 94077937 | NET IN | 40, 01, 38 Valve Command Acknowledge
              7/28/2020 4:27:06 AM | 94080484 | NET(0) OUT | 40, 01, F0
              7/28/2020 4:27:06 AM | 94080500 | NET IN | 40, 01, A1 Status Response
              7/28/2020 4:28:03 AM | 94137953 | NET(0) OUT | 40, 01, 38
              7/28/2020 4:28:03 AM | 94137968 | NET IN | 40, 01, 38 Valve Command Acknowledge
              7/28/2020 4:28:06 AM | 94140531 | NET(0) OUT | 40, 01, F0
              7/28/2020 4:28:06 AM | 94140546 | NET IN | 40, 01, A1 Status Response
              7/28/2020 4:29:03 AM | 94197734 | NET(0) OUT | 40, 01, 38
              7/28/2020 4:29:03 AM | 94197734 | NET IN | 40, 01, 38 Valve Command Acknowledge


                What is the specific content of the email? It may be during shutdown rather than during startup.


                  subject line of email shows: mcsSprinklers detected value failure for Zone _25

                  body of email shows: NET Module 1 Valve 1 failed
                  On different days, Valve number may change from 1 to 8. I only use 1 module, not using any extension.


                    The debug should show this as well. Is there something in the debug related to valve control in the last minute prior to shutdown/restart? Does Zone_25 make any sense to you?


                      Michael, your question about Zone_25 helped me figured out the problem and I feel quite silly! I turned my old HS2 system back on at one point to compare settings in mcsSprinklers between HS2 and HS3, then forgot to shut it down. Since that machine is not connected to rain8net any more, I think it has been sending alert emails about connection problems everyday...