why didn't I get a trigger?

[bart1]

this is the trigger should go off at 4% change in humidity...not sure you sample rate. tneres nothing in the log....is my pause causing grief?

[jvm]

The issue may relate to "when" the OLDVALUE is determined and the step-size between changes. Your looking for a 4% change; however if the change occurs slowly - say it goes up by 0.5% then another 0.5% then another 0.5% so that, after a while the "total" change is more than 4%, you would still not trigger because each step was only 0.5% so NEWVALUE - OLDVALUE = 0.5 even though, over time, you did a larger change.

I note that even though your graph looks like it is doing big jumps of > 4%, maybe that graph is a minute-by-minute graph and within a given minute, there are actually several reports from the humidity device that add up to a big jump, but each step is smaller than is apparent. (Just guessing here).

One way you might achieve what you are looking for is to set up a virtual device or a counter, call it "Old_Value". You would use this "Old_Value" device to store the value of humidity whenever your event executes so your event would look like

IF Humidity sensor changes
And IF Easy Trigger Comparison Between Two Devices [Humidity Sensor] > [Old_Value Device] + 4

OR if Humidity Sensor Changes
And If Easy Trigger Comparison Between Two Devices [Humidity Sensor] < [Old_Value Device] + 4

Then do your processing
And finish by setting the OLD_Value device to the current value of the Humidity Sensor

[bart1]

yes...sample rate is what your trying to say...
the chart I provided looks like I had a rate of change of about 38 pts of humidity change in 7 mins I changed to a trigger test that worked in the past and it worked again today roughly at the same temperature (the sensor as a significant variability) and got a similar response time 40 in 7 mins. The difference was I put a wait in rather then a pause (see below)
would like to know what the sample rate is (e.g. once a second)..yes I can zoom in to infer but it would be nice to know from the designer and to know if its modifiable.

[bart1]

I think I figured it out...the delay you can added to the event trigger runs first which means it doesnt run for 20 minutes (in this case) not the effect I want...its a bit deceiving since its at the end of the task.

[bart1]

I understand how to fix it...the delay is on the next event that I run...

[bart1]

so does anyone know what the sample rate of easy trigger is? for reading a device

[spud]

so does anyone know what the sample rate of easy trigger is? for reading a device

There is no sample rate, EasyTrigger is notified by HS of any device value change, and then check if some triggers need to be fired.

[Uncle Michael]

so does anyone know what the sample rate of easy trigger is? for reading a device

I think I may be starting to understand what you want to do.
If I do, then what I think you will need to do is create a virtual device and a recurring event.
The recurrence time of your event will be the "sample rate". The event action will be do assign the relative humidity reading to the device value of the virtual device.
Then changes to the value of that virtual device can be the used to trigger the event.
So, if you sample the humidity every x seconds, and the value of your sample device changes by more than 4 (%) in that interval, then you'd initiate the ventilation sequence.
Is that what you have in mind?

[bart1]

sounds like a windows based trigger action limited by the OS and HS cpu load...I have 300 devices (per the device management page) so guessing hundreds of milliseconds per full update. I'm using the device change in percentage and would like to see the change over a period of time (without writing a script). e.g. 4% change within 30 seconds rather then hundreds of milliseconds.

[Uncle Michael]

I'm using the device change in percentage and would like to see the change over a period of time (without writing a script). e.g. 4% change within 30 seconds rather then hundreds of milliseconds.

I describe how to do that in my previous post. Do you have further questions?

[bart1]

sorry was responding to spud...I will try your idea...like it a lot...again without writing a script (will do it but want to get things back to where I was before at some even crude level). The issue I have is I have a relationship between the HAI temp/humdity sensor with humidity and temperature either because of the device or because the heat not allowing the relative humidity to change as fast. e.g. for a given shower temperature and duration I get something like 4% change at less then 80, 2.5 80 to 90 and 1 at >90. I need to do a math function with humdity/temp and time and it sounds like the virtual device might help that (if there's an example you know of? albeit I can search)

[Uncle Michael]

I think you are over complicating it. Since you can create a graphical representation of the function, it should be relatively straightforward to find a time interval and RH change that will occur at all reasonable temperatures. (The issue you are describing sounds like relative humidity. The air has more capacity for holding water as temperature increases. The equations that describe the phenomenon must be readily available, but unless the exercise looks like a fun project, why bother?)

[bart1]

its fun for me fyi I had a perl script that ran and I had dialed this in to within 20sec of opening every morning..now that I'm retired (electrical eng..35 yrs aircraft integration (mil)).

[bart1]

having said that...the consistency was taking a shower at 5am everyday..now it could be 11 so didn't have the temperature issue (sun can be shining directly on the sensor)...I might move the sensor to be more predictive..I think your right about the humidity and temp. but I have a feeling the relationship to aging and quality of the sensor has a 1st order effect as well.

[Uncle Michael]

its fun for me ...now that I'm retired.

Go for it! That's what retirement should be about.