You are right in that these will also trigger flows a lot if ill designed. With these that is unavoidable and needed, so there the things @B3rtmentioned for protection are right:
Make sure you test the for the desired range a.s.a.p to make sure the flow is stopped when the value is outside the range you want to react on.
Make use of booleans, or device states or even better turn flows off that are not supposed to work at that time
If present, favour “value becomes greater/less over the above bullet(s) as they will trigger way way less.
This is what I do understand; make sure the flow decides it’s not needed ASAP, in other words disabling it in the IF gives less load then in the THEN? Right?
But on sunny moments and the sensor is going up and down every second, is this than still more efficient than running it once a minute?
Looks to me we explained the once per minute trigger will trigger 1440 times a day and the WHEN value becomes lower then 500 will only trigger a few times a day. But if u want a second, or 3th opinion that’s ok. But don’t say we told u to use this just because.
Sorry m8, i won’t annoy u again!
I am not sure if you do not read my entire question or maybe I am not great in explainig what I am asking.
I am not looking for second or third opinions, I am looking for the answer what is giving the lowest load on my Homeys processor.
The discussion mentioned above is giving a answer to my question, but I do not understand the answer. And my head likes to understand why instead of just doing something.
I DO understand your XXX triggers per day statement and I agree that in this usecase my Homey is doing this and it’s not a good thing. But building the way you’re telling me gives my Homey a trigger every second when the sun is out. My sensor really does change every second! Admitting your opinion in 1440 triggers a day, this triggers 60 events a minute. How is this better?
Again, it’s not to nag or have doubts about your answer, it just feels there’s more to it and I wish to understand the why.
One of the flows below will always be triggered once the sensor is updated. If the sensor is updated almost every second, will this still ask less load from Homey than the ‘1 minute’ flow? That is my question!
The flow with helderheid becomes lower then 500 will only trigger when it goes from more then 500 to less then 500. It will not trigger when helderheid goes from 499 to 498 because it’s not coming from higher then 500 to lower then 500.
So the trigger will be:
When helderheid was more then 500 and now it goes below 500…
Ur 2nd flow in the last examples will fire every time it goes below 22000 AND is over 500, bad idea.
The other 2 flows will only trigger when 1) helderheid was over 500 and now goes below 500 or 2) helderheid was below 40000 and now goes over 40000
What is for sure, is that the results in the evening are 100% better with the ‘updated sensor’ then the minute trigger. Have been tracking for the last few hours and it became clear that in the evening, when there is no direct sunlight hitting the sensor, it only updates anout once every 30 minutes.
I am very curious about tomorrow morning, the first hours of the day the sun is bright and continuesly hitting the sensor. Will it go all crazy and catch up for the evening hours or will it not catch up?
Curious about your results as well, we’ll talk tomorrow Nightynight
The “becomes greater/less will do the trick for all flows, as they will only trigger when the bound is passed. the chances of it going from very high to very low in one step are close to zero.
There is a slight chance the flow will trigger more when the light value outside is varying slightly above and below the trigger value. This is not that likely in this case, but to prevent it you can use a “Schmit trigger”. It simply means you should make the “becomes higher than” and ”becomes lower than” different, where the first is higher than the second. The difference should be higher than the slight variances in light. This will stop any yo-yo effect.
Last but not least: setting the blind hight in a variable is a good idea. To make things simpler, you can create a flow: When variable rolgordijn changes, set blind to corresponding position. This will prevent sending the same value to the blind over and over. Because if you set the same value to the variable repeatedly, the variable does not change and the command is not sent to the curtain.
First; de sun stays away today and then it’s becoming more and more clear that the counter of the variable change is winning this battle. Yes, there might be a chance that in the hot summer this will trigger more and more often but this will never win from the 1 minute battle. I will change my light flows as well, keeping the 1 minute trigger out of homey.
That said, looking at the yo-yo boundries that are mentioned will take the ‘more triggers in summertime’ away completely. Sounds like a very good solution. I will read the documentation of this and H.O.O.P. in the upcoming dats and implement this in my flows.
Thank you so much!! After reading your post I went to buy two lux-sensors (for outdoor use) and mounted them on the outside.
Followed your instructions and made som adjustments on my own to suit my needs and it turned out great. Looking forward to tomorrow so I can see how it worked during a full day.
One suggestion that I did myself. I calculated the lux value as an average for the last 5/10/15 minutes and used that value to trigger the flow. This due to the flows reacted a bit too fast when dark clouds came by. Worked great!
I think your average value is a good method, but how did you do that?
In Homey, in the lux-sensor or in some other external program?
Should it be possible for me, without any “Programming skills” to do the same?
That was the thing I was searching for. Cool, i’m going to build both and run them for 24 hours 
