Soil moisture opinion

[Michael McSharry]

I used your data and stepped through the computations and correct values were being computed. It looks like it may be a cache type thing where the data has not yet been stored at the time it is being read to compute the change in the last hour.

What I will do is add debug where it is being computed. What database type are you using?

[cjin]

I used your data and stepped through the computations and correct values were being computed. It looks like it may be a cache type thing where the data has not yet been stored at the time it is being read to compute the change in the last hour.

What I will do is add debug where it is being computed. What database type are you using?

I used Access. I am still running .Net2, does that make a difference?

[cjin]

I found another strange behavior: had a watering cycle last night/this morning. In one area, some zones finished the entire cycle (over 2 round robin cycles), one zone only had 1 of 2 round robin cycle before time restriction kicked in at 7am. At end of the day (10pm), zones finished the full cycle moisture stuck at 100%, the one zone had only 1 of 2 round robin cycle moisture stuck at 50%. Calculated ET for the day is 0.2 inch but moisture for all the zones in this area didn't change.

I have two other areas, one watered also last night, one didn't water. Moisture for both these two zones acted normally.

[Michael McSharry]

Round robin is a tough situation because basic scheduling is the full time allocated to a zone. If I recall correctly moisture update is calculated at the completion of the cycle which would be the completion of the last round robin. If you have the debug then post it so I can assess.

.Net 2 is not a problem, but this implies to me that you are not running a current version of mcsSprinklers. If I do a build for HS3 it will be .Net 4. If I do a build for standalone which share the same .dll as HS2 then it will be .Net 3.5.

[cjin]

I am still running HS2PRO. The version is V2.16.0.5/2.16.0.5 (Professional).

I can turn off round robin, so it doesn't complicate the issue for now. I am guessing you want to hold off ET calculation until a cycle is complete in case it is only temporary restricted and will resume the cycle in a short time. Maybe this is a good reason to select setting to complete already started cycle when restriction becomes effective (currently I immediately stop the cycle when restricted).

[Michael McSharry]

The attached is a current 2017 build for standalone that should work with HS2. It is a 3.5 .NET. Are you on .NET 2 because of an embedded XP restriction? There are some hoops I need to go through for get back to ability to generate a .NET 2. The charting uses the 3.5 libraries.

For scheduling issues the debug is what I would need to see to understand why something did not work as expected.

[cjin]

The attached is a current 2017 build for standalone that should work with HS2. It is a 3.5 .NET. Are you on .NET 2 because of an embedded XP restriction? There are some hoops I need to go through for get back to ability to generate a .NET 2. The charting uses the 3.5 libraries.

For scheduling issues the debug is what I would need to see to understand why something did not work as expected.

Yes embedded XP is exactly the reason I am still on .NET 2.

Here is another observation, hope it helps to identify if it is a cache problem: it rained for 0.5 yesterday afternoon. Rainfall amount was reflected correctly on General tab right away, but moisture % for all zones didn't reflect the rainfall till a few hours later.

[Michael McSharry]

The .NET 2 version is attached

[cjin]

Michael, thanks for posting .NET 2 version. I will pour some water in the rain sensor to test. How frequent is moisture % calculated (or over how long a period of time I need to do this test)?

My rain data is from an Oregon PCR800 read through RFXCOM plugin. The plugin has a virtual device that keeps track of accumulate rainfall for the day then around midnight resets it to zero for the new day. In debug file, I see the rainfall (device ^68) retained prior day value for the first two entries then reset to zero around 12:01am. Does this create a problem? In Weather Setup I selected Daily Total.

[Michael McSharry]

Your setup seems to be correct.

The rain sensor is updated on the one minute cycle for HS devices with the database being updated when the device value changes. Soil moisture is updated hourly or when a cycle completed. The ET is computed hourly and at that time the rainfall over the last hour is determined by looking in the database over the period -30 minutes to -150 minutes from the current time. The skew is primarily to support the variety of rain sources that tend to update somewhere around the top of the hour. mcsSprinklers hour start at 7 minutes after the actual hour to let other sources stabilize.

mcsSprinklers is looking for a time delta of at least 30 minutes, looking backward, and then uses the delta in daily rainfall as the amount of rainfall in the last hour. The algorithm is not perfect and it is possible that there is a small difference at the end of the day for the total rainfall shown by the sensor and the sum of the 24 hourly calculations. This variance in "rounding", however is not as significant as you are seeing.

[cjin]

I noticed some problems in v2.17.0.11 for .NET 2 you posted, although they are not directly related to the rainfall data issue:

1) on Other Setup tab, it only show Basic and Standard version. The pro version is no longer visible.

2) on Area setup tab, most of the settings are no longer visible.

3) On General Status tab, weather forecast/UV etc no longer get updated although on Weather Setup tab the I can display correct forecast info when click on Evaluate Internet Download button.

[Michael McSharry]

Look at Other Page. Is P:rofessional option available? At the bottom is expert mode available? If so select both of these.

[cjin]

Look at Other Page. Is P:rofessional option available? At the bottom is expert mode available? If so select both of these.

Professional options isn't available. At bottom of the Other page, only expert mode context help is available and already selected.

[Michael McSharry]

All we are trying to do is look at the hourly moisture calculation which queries the rain information from the database. The license key is HST magic so not able to easily investigate, but should not matter. If the test does not provide the data we need then other investigation of the variance in operation can be researched.

[cjin]

All we are trying to do is look at the hourly moisture calculation which queries the rain information from the database. The license key is HST magic so not able to easily investigate, but should not matter. If the test does not provide the data we need then other investigation of the variance in operation can be researched.

Understood. Please see attached debug file. Some of the RainLastHouse numbers don't look right. Thank you Michael.