FYI… I’ve seen a lot of discussion about “plasma” TVs being noisy… I have to say: OLEDs are exactly the same or “worse” (as in: OLEDs consume power directly proportional to the brightness of the image being displayed).
Both plasma and OLED obviously consume some power “constantly” for image processing and other circuitry (when the unit is “on”). However, the primary power consumption is directly based on the image being shown. A bright white hockey rink will consume lots of power; a dark Game of Thrones scene will consume very little.
When fast-forwarding in particular, the usage will change extremely rapidly for both of these types of devices.
Contrast this with an LCD: for the LCD, if you disable energy-saving, and have minimal “local dimming”, you can get effectively constant power usage (when on).
These variable loads need decoding at the point of use and better internal dc power supplies that average out the power consumption. Maybe some day our power sources will include a DC power buss so that PV solar and battery storage will be more efficient and the variation in power will be more filtered. Computers have the same issues with bright vs dark displays which makes it difficult for sense to detect power on and nearly impossible to monitor variable power levels. As for following the programming stream for dark vs bright screens there are delays in the data stream, data buffers and if On demand streaming is used syncing seems unlikely.
I have this same problem with a similar Panasonic plasma. I was hoping Sense would pick it up but I can understand now that it is too complicated to recognize it. Maybe I’ll use a KillAWatt to track it instead… or maybe a smart plug.
@Larkwoodhome, unless a Kill-A-Watt can now show waveforms, I would recommend picking up a 22$ HS110 and maybe a few short (9”) extender cords, and configure as your “Explorer” smart plug with Sense. You can try plugging it into many different devices and capture their usage over several days.
None of the Kill-a-Watt meters I know of capture waveforms, but they are very accurate. I’ve used the HS110 (which you can find for as little as $17) just as described, but do note that the waveform captured by the HS110 is quite different than that captured by Sense. It’s orders of magnitude lower frequency and thus misses a lot of the “sharpness” that Sense seems to depend on. And, of course the HS110 is capturing the whole power signature coming thru to the de vice, where Sense captures what can recognize as device components based on their transients and has zero capability to capture individual steady state load signatures…
@andy,
All good points. I will point out though, that displayed resolution is quite different than measurement resolution with both Sense and HS110. The displayed waveforms you would see, assuming both Sense and the HS110 were measuring the same single device, wouldn’t be all that different.
Sense samples at a max of 1MHz and does a bunch of it’s magic at high speed, but for purposes of display to mere humans, it displays RMS power at 1/2 second resolution, so you never see the peakity peaks of power/current spikes. The HS110 samples at something like 3kHz and outputs new RMS data every second. Sense only asks for data from the HS110 every two seconds. So there are 3 orders of magnitude difference in the raw sampling, but the aggregated data you see is not all that different, RMS power at 2Hz and 1/2Hz respectively…
Thought I’d add my “Dirty Device”. It’s a Lennox XP25 variable speed heat pump. The compressor, cooling fan, and air handler are all variable in 100 steps. Sense has detected none of these. Good to know the heat pump may be hiding other devices…it rarely runs in the spring, perhaps Sense can do some profiling a few weeks from now.
I have a similar heat pump HVAC system (Day & Night); fortunately the thermostat (smart device) is capable of determining energy consumption, including gas usage. The thermostat dictates fan speed, compressor and TXV position. I can’t complain; knowing the ever changing behavior of the HVAC system is sometimes masking accurate device detection.
Overall, I am really happy with the Sense system, it is accurate and very sensitive.