How to contribute with my TV?

[knudsvik]

I did some measurements (with a TP-Link HS110) for my Samsung TV. It seems to be drawing quite consistent power when in use. I think this add-in is great, and I would like to contribute with my TV readings so everyone else with similar TV gets to enjoy the power reading without any setup. But the wiki only covers lights. Do you want us to add whatever device, so the database grows to cover all sorts of devices? I could potentially also measure an IKEA speaker, a Sonos One speaker, an LG TV etc as well to further expand. What do you think?

[bramstroker]

Hi, lately the library has already been expanded with smart speakers and self usage of smart plugs / switches.
Also see the supported models list where you can select device type now.
The WIKI has also been extended with procedure how to do smart speakers. So you can use that for the IKEA and sonos speakers. See the library files for example how to configure.

How did you do your TV now? Using the fixed strategy?
How does the volume affect power usage? And what about different modes (Eco, Sport, etc)?

[knudsvik]

I have measured over a few hours with different kind of input and also a range of volumes. The TV has volume as an attribute in HA, but it is not significant when it comes to power usage. The type of content is more significant for the power usage, but not an attribute in HA. Hence I have taken the average power over several hours and used that as a fixed strategy.

The modes were interesting though.. hadn't thought about it. I had dynamic mode on which consumed about 175W - 180 W. When setting it to standard mode it dropped to around 90W! Would never have guessed it. Makes it much more difficult to have this as a device in this library now, since the mode is not an attribute I have access to in HA :(

[bramstroker]

Do you know what happend before 18.00 when the power dropped to 150 and 120?
Besides that it indeed looks very consistent, not that much fluctuation.
How did you calculate the average 179?

[knudsvik]

Do you know what happend before 18.00 when the power dropped to 150 and 120?

I think it was left with a dark still picture for a while. Could do some more investigations, I know the kids were playing with the Nintendo Switch around that time.

Besides that it indeed looks very consistent, not that much fluctuation.
How did you calculate the average 179?

It should be the average of data when state of TV=on, need to double check it, seems a bit high.

But anyways, is it worth doing anything when the modes are so significant to power usage? It could be 100% off if set as a fixed value at around 175-180 W.

[bramstroker]

I am not sure if the different modes have a (big) impact on power usage, it was more of a question which you could verify.

I think it's acceptable to include in the library when the power approximation will be ~5% of reality.
But when a still image for example means the power is 25% lower than average, and the user has that on for 2 hours, it will quickly add up to be unreliable (kWh wise).

[knudsvik]

I am not sure if the different modes have a (big) impact on power usage, it was more of a question which you could verify.

As I wrote earlier it has a major impact. Changing the mode made the power use go from 180W to 90W

I think it's acceptable to include in the library when the power approximation will be ~5% of reality.

Not realistic given the major impact of the modes..