We can configure the threshold (level, time) between standby and ON, but we cannot do the same for the transition between standby and OFF. Many modern devices have a standby power that is close to zero so the distinction between off and standby is tricky. For example my TV is typically well below 1W (as reported by Kasa plug), but occasionally might go a hair above that. Unfortunately, this causes constant log entries when the TV is always on standby. Here’s todays:
I understand that part of the problem is that Sense seems to drop fractional digits. Kasa can tell the difference an reports the average standby power as somewhere in the range of 0.3-0.6W, which is clearly different from off, which is a true 0W. So why can’t Sense also read the fractional digits to distinguish between Off and standby, even though it only displays integers. This would quiet the logs quite a bit.
Another option would be to allow configuring the off/standby thresholds. Is this case I would set the threshold to 0W, meaning that 0W is still considered standby. Would be more realistic.
I get this for my Smart TV that is plugged into a KASA plug.
You can set the threshold in the device properties. I haven’t managed to get it to work 100% at the moment.
I may need to set the time to be a but longer.
That’s the threshold to distinguish between standby/off and ON.
I need to distinguish between standby and off. Currently the standby power is so low that Sense cannot tell the difference and constantly flickers between standby and off on the device log, even though the device is always in standby.
(Quite a different problem, unless I don’t understand your reply.)
@altenbach, I had that same exact question with respect to some HS110 / KP115 experiments I was doing a while back. @JustinAtSense got back to me. Today the Standby to Off threshold is a preset 0.5W though Justin indicated that they were looking at that.
Thanks, I will submit a ticket for tracking. I understand that for “Sense detected” devices it would be impossible to tell the difference between off and standby.
For Kasa plugs it’s a different story: Only a true zero should be “off” and any detectable current should be considered standby. (I make the assumption that the Kasa hardware is accurate and does not have an offset error. No way for me to tell if this is a valid assumption). This is independent from the fact that the Kasa plug itself consumes some power, because Kasa does not report that value.
I think many new devices are intentionally optimized for low standby power and since hardware is now available to do full standby processing at <<1W this will become more pressing in the future.
We don’t really care if it is off or on standby, except for the excessive meaningless log entries.
@altenbach, your comment that many new devices are optimized for low standby power rings a bell. Back when Sense began that was not the case. Perhaps it is time for them to adjust how they report. I know this might be a big deal for them as they have been rounding things to the nearest watt for a long time, but they are good at innovating. I would like to see more precision displayed for small values. In math class they taught us about significant figures. If Sense were to set the number of significant figures to two, then a measurement of 12W would not have a decimal, but a measurement of 9.8W would.
Kasa plug measurements are very precise, per this research by kevin1
When I got my first Kasa HS300, I hooked the whole thing up to a Kill-A-Watt out of curiosity. I found good correlation between measurements, except that as you note the Kasa plug itself consumes some power. If all six plugs were ON with nothing plugged in, the Kill-A-Watt measured 3.6W. If all were OFF, it measured 0W. Turning them on one by one increased the measurement by 0.6W each.
I would think that it should be more than 0W if all outlets are off. There is still the main LED ring, the processor/microcontroller, and the WIFI radio running. (There will also be some drain if the USB ports are in use).
Does that also mean that all currently unused plugs should be turned off? (I have an old kill-a-watt, but the HS300 is currently plugged in where it is inconvenient to easily test this)
Yes, I do turn unused plugs to OFF to avoid the overhead of 0.6W.
The Kill-A-Watt I have displays data only to the nearest 0.1W, so the LED ring and WiFi radio together probably did not create enough usage to register. I know you like your zeros to be nothing, but my technology wasn’t that precise. I guess I could repeat the test by plugging the HS300 into an HS110 or KP115, both of which display to 0.01W in the Kasa app. Like you, however, my HS300 is at an inconvenient location. I made my measurements just before screwing the power strip to the wall.
