I wondering if anyone in the community has had to solve a power monitoring of an outlet where the device is very high load and very short timeframe. 240v 5sec 3000w around 60 / 100 cycles per day. My current Tuya solution is hopeless, capturing less than 50% of the actual power draw of the device.
Also, my water softener draws a very small load 3w, just to power the board to accumulate consumption. When it regenerates I can see it draws about 7w to rotate the valve. Also this draw is just to rotate the valve ~5secs again. Again the Tuya device isn’t catching this load and so I am not getting the regeneration data out of it.
So I need something with a high sample rate I guess? Does anyone have any real world devices solving similar issues successfully?
Hi, I converted some of my wifi power monitoring plugs to zigbee, and then in Zigbee2Mqtt, I was able to adjust the reporting time by setting the poll_interval option to 1 second. So just make sure to check before you buy, that the model support changing the poll_interval.
The measurement_poll_interval is specific to that particular device. Chances are you will not have it available.
However, all devices in Z2M can be set up to configure reporting. Setting this to 0 for minimum reporting interval and something reasonable (in seconds) for maximum will ensure that any changes get reported immediately, while not flooding the network.
Do not be scared of the code in that link - this is all configurable in the Reporting tab of the device in the UI. Here’s mine, just so you get an idea:
High sample rate or high update rate?
My Shelly EM can read phase cut power very accurately, I expect that the sample rate is at least 1000/second, probably more…
Is the sampling rate on the Shelly device documented anywhere? I’ve never been able to find a device with a high enough sampling rate and reasonable price to give useful data from bang-bang controlled devices, the vast majority of home appliance devices… Thx.
But I can confirm that Shelly EM can read phase cut load at high accuracy. I haven’t done research for the sampling rate but it has to be more than 1000/s . Probably much more.
Any idea how many data points the Shelly device sends out over WiFi per second or minute? And is that data some type of average over a period of time or single point in time? Getting accurate power usage figures for home appliance devices has always been interesting to me, however ‘over my pay grade’ of knowledge. Most of the power monitoring devices in home automation seem from my experience to sample and send out data a most around one data point per second and just a point in time measurement value. For a light bulb or perhaps a phase controlled dimmer such as you have, I am guessing this might give accurate measures. But for things like the average coffee maker that controls heating via toggling a square wave on and off, getting an accurate answer to the question ‘what did that pot of coffee cost to make’ does not seem to be ‘easy’ to find.
I’m repeating myself here.
To measure power of phase cut load (dimmered) you need to sample within one AC half wave; 0.01s. To do it accurately, you need a lot of samples. While I don’t know sample rate of Shelly EM, I could imagine it to be in range 10,000/s to give results what I get.
I update only 1/s, no idea how fast it could go and where my esp limits would be.
