Your help is needed - question regarding rules and variables
so the rule flow could be
while $humidity-diff > 0
regenanlage on
wait 1 minute (shorter than max, to prevent humidity overshoot)
regenanlage off
wait 1 minute (waiting time of sprinkler)
$humm-diff = new value
If you agree, i can translated it into rules, variables and 2 timers
@bertreb Thanks for your help so far! I would like to give it a try 
My suggestion for the vietnamse-turtle-humidity-controller
Create a variable called ‘sprinkler-state’ with value 0 and create a timer device called 'sprinkler-timer’
Create the following 3 rules:
The stop, reset, start sequence of the timer is to be sure (there are/were some issues with timers not resetting). I turn the sprinklers on for 60 seconds, but you can increase that to max of 120
Let me know if it works and you can tweak the values as you like.
@bertreb Thanks for your time and fantastic support!
It worked out well in my simulation. Go live will be tonight as the DHT sensor is not wired yet.
Regarding my DS18B20 sensor: I activated a manual interval of 2 minutes for every request.
Before it was at 10 seconds and thats not necessary and this could have caused the issues.
Could you guys help me out with another problem? I´m sure that this one will not be easy to solve but lets see…
As you can see in my initial post i´m pulling live weather data from openweather plugin using a “OpenWeatherDevice”. The city is Hanoi in Vietnam: https://openweathermap.org/find?q=hanoi
The local time in hanoi is 22:04 Friday, 24. Januar 2020 (GMT+7) and this means that the humidity level is not correct for my automation. We are now applying with the new rules from @bertreb the realtime humidity level from 10 PM (local time vietnam) but it should be the humidity level from 7 hours (GMT+7) in the past.
I had this discussion two years ago with @SaxnPaule and some other guys
BR,
André
@bertreb said in Your help is needed - question regarding rules and variables:
If you assume the humidity sensor is as fast follower of the humidity increase, i could give you some rules.
The DHT22 respond in approx 6-10 seconds to a humidity change, if there’s enough air flowing thru the sensor. Thats seems fast enough.
just to understand you correctly. You want every 15 minute the ‘actual’ humidity for Hanoi like it was 7 hours earlier?
Yes, I think thats right, but could also be that im totally confused ^^
I cant remember well but we need the 15 minutes because of the api limitations from openweather.
This is the last missing piece to complete my goal of a almost 100% automation.
I dont know if there is a solution but if there is one, then we need to apply it also for the temperature.
BR,
André
the solution is to buffer the 15 minutes humidity values for 7 hours and read the oldest value out of this fifo buffer.
With ‘only’ rules and variables available this is difficult. In coffee or javascript this is easy.
I could write a small swiss-knife pimatic plugin to support this as a first tool.
This Sounds like a plan. Let’s give it a try
And thanks again for your time and help!
@Andre3000 I released a first version of the pimatic-tools plugin.
It has 1 device type called VariableDelayDevice. You can imagine what its for.
It can be installed via the plugin page (browse tab).
The readme should explain how to use it.
Let me know what the results are and if anything should change.
I’m also testing it, but because of the delay it takes some time 
@bertreb Thanks again for your time and help!
The installation of the plugin was easily possible without any issues.
I´m still testing the plugin and I will come back to you as soon as possible.
BR,
André
@bertreb I´m not sure if the plugin is working as intended. The variable is sampled every 15 minutes but the value is not changing. It stays at the same value as 7h before.
Could you double check with your environment?
Br,
André
Hi andre,
What is the delay time you use? because after the startup during the delay time the value stays the same. And that is because after startup there’s no info on the past, so i have to build up the delay buffer.
I have a test running with 60 minutes delay and sampleRate of 15
As you can see in the graph the value is changing with a delay of 60 minutes
Thanks for your fast reply.
The delay time is 420 minutes. I knew that there can´t be values for the first 7h but the device is now running for longer than 7h.
BR, André
The timestamp is not changing?
The screenshot is only covering 3 hours.
I mean the timestamp of the value. You see when you click on the value T 15% or RH 58%
A few questions to isolate the problem:
I´m not sure if this is the information you are asking for…
Please forget the values before 2 p.m. <- wrong setup of the device.
The timestamp after 2 p.m is changing every 15 minutes as expected.
A few questions to isolate the problem:
Is the timestamp of the value also the same or does it increases after 7 hours?
Are you sure that the actual value is changing
What is the result if you take a shorter periode like i did (60 minutes delay with 15 minutes timeSample)
Did you use the latest version 0.0.5. Because earlier version had a wrong delay periode (like times 10)
I will test now the newest version and report again.
BR,
André
@andre3000 thanks for the info
If you were on 0.0.2 the error in the delay time was probably causing the problem.
In your setting the time before any change would have happened was 420 * 15 = 6300 minutes = 105 hours = 4,375 days 