Easy Trigger event does not trigger

[alphatech]

Hi Spud,
Perhaps I do not understand this trigger. I have my washer that when is in idle status is around 1.34 W. When it starts, it goes between these values of 8.40 W to 823 W. when then it is done, goes back to 1.33, 1.35. Why this event is not working?

Thanks,
Aldo

[logman]

Perhaps your range is too narrow... Maybe there are some digits to the right of 1.34 you are not accounting for (which would make the value out of range). Also you mention the washer occasionally goes to 1.35, which would also be out of range of the lower band. Maybe broaden your lower range out to 0 - 4.999 watts and your upper range from 5.0 - 920.

Or use this ET trigger instead, which will fire once the washer has been off for exactly 10 seconds. This way you are only monitoring an exit from the upper band and not both bands.

--Barry






--Barry

[alphatech]

Hi Barry,
I may try your last suggestion but please take a look at this attachment, based on those values why my query is not working?
Thanks,
Aldo

[ewkearns]

Do the values need to say "Watts?"

[alphatech]

No. the device value is numeric, the query that you see is from the database that capture these values every time the device changes.

[spud]

How does this device value is updated? using a plugin or a script?
If you set the log level to Debug in the EasyTrigger config page, do you see some log lines similar to this for your device:

Apr-20 10:52:56 DEBUG Device 4373 value changed from 43 to 3

just tested again this trigger with a virtual device, and it worked for me.

[alphatech]

Spud, sorry for the confusion, I use the the z-wave devices and I collect this info into a database. The pdf is the collection of these values from the database so you can see each value. Can you please suggest the right trigger based on those values since it is not working for me?
guess I do not fully understand that trigger myself, it will be great if you can give some examples on how I would correctly use it.

This is what I see in the debug for that device

Apr-20 12:06:07 PM

EasyTrigger

DEBUG Device 2198 value changed from 1.128 to 1.119

Thanks,
Aldo

[alphatech]

I'm thinking about it, I think these triggers would not work, I think another trigger that has a trigger that looks like the picture may work, what do you think?

[spud]

According to your first screenshot and your pdf, the event should have been fired at 9:33 when the device value changed from 2.33 to 1.33, unless you do not collect the values to your database everytime it changes? For example if the value changes first from 2.33 to 1.80 and then from 1.80 to 1.33, the event would not be fired because 1.80 is not in any of the ranges.

So to solve your original problem I think an event like the one suggested by Barry would be a better approach.

[alphatech]

I tried it but is not working either. What about a new trigger as I suggested, I think this approach should work, would you think? As for washing on, I could check if is from 0 to 10 and if great than 10 means is on. I think it will work well for the off as well by using the less than. Is this visible?

Sent from my SM-G935V using Tapatalk

[spud]

I don't think there is a need for a new trigger, if you want a trigger for

a device's value was less than X and becomes greater than x
just create it as

device's value was in range [-1000000, X] and becomes in range [X, 1000000]

it's strictly equivalent assuming your device value can never be less than -1000000 or greater than 1000000

[logman]

I think Spud is right: Your database is missing some values or collecting values based on different reporting rules. Leave the ET debug logging on through an entire washing machine cycle and then filter the log down to just the device. Then compare the values that debug collected to the values that your database collected and see if they are an exact match (same change count and same values collected with each change). I suspect they are not.

First thing I notice is your database is collecting values 2 digits past the decimal and the debug log is showing 3 digits past the decimal. Unless you have the database set to round the data that way specifically, it should be collecting the data in the same format as ET is seeing.

Wattage is never a "static" value. And in your case it is not an average of values over time. Rather it is a measurement at a specific millisecond in time when the z-wave device reports the value. If you check it exactly 1 millisecond later, it is likely to be slightly different... Especially when you are measuring thousandths of a watt. On a Z-Wave device it will report a wattage change when the change is n% <> last value. In this way the device is not constantly reporting changes.

In the screenshot you provided it shows a last change time for Watts to be 11:28am vs last change for Amps at 11:01 am. Now in reality this is not possible, because a change in watts means a change in amps (and vice-versa) because each are included in the others calculation formula...

Amps = Watts / Volts
Watts = Volts * Amps
Volts = Watts / Amps

You could try firing your event based on a change in Amps instead of Watts. It appears your Z-Wave device (as most do) has a larger change threshold for Amps than it does for Watts required to report a change. Your database may be collecting data based on a reported change in Amps and then converting to watts. If so, there is your discrepancy.

No matter, you'll have better success if you fire your event based on either watts or amps being less (or greater) than a given value for x number of seconds. That gives the effect of measuring the reported values over time, which is more accurate for your case than a millisecond snapshot.

--Barry