add YLXD024

[xxddff]

Device information

Official website(Japanese)
https://japan.yeelight.com/product/1624.html

Instruction manual(Japanese)
https://japan.yeelight.com/wp-content/uploads/sites/5/2021/11/%E3%82%B7%E3%83%BC%E3%83%AA%E3%83%B3%E3%82%B0%E3%83%A9%E3%82%A4%E3%83%88_%E5%8F%96%E3%82%8A%E6%89%B1%E3%81%84%E8%AA%AC%E6%98%8E%E6%9B%B8-Minas%EF%BC%86Minas-%EF%BD%90%E3%82%8D%EF%BC%89.pdf

(Because it's a Japanese model, the information page is only in Japanese.)
Translated by Google

Home Assistant Device information

ceild
Manufacturer: Yeelight
Firmware: 8

Checklist

• I have created a single PR per device. When you have multiple submissions please create separate PR's.
• For lights - I have only included the gzipped files (*.gz), not the raw CSV files.
• For lights - I have provided a CSV file per supported color mode. Look that up in Developer Tools -> States

Additional info

1. Power Measurement Accuracy

The power consumption readings are accurate at low brightness levels. However,
at higher brightness levels, the measurement error can increase to approximately
1.5W. Any suggestions for improving accuracy would be appreciated.

2. Night Light Support

This device supports a night light feature. In Home Assistant, the night light
can be exposed as a separate entity for control, resulting in two light entities
under one device:

• Main light entity
• Night light entity

How should I structure the model.json (or other configuration files)
to include power measurement data for the night light entity as well?

[github-actions]

Thank you for submitting light measurements!
Here are some visualizations for the CSV files.

[bramstroker]

the measurement error can increase to approximately 1.5W

What did make you assume that? Measurements look consistent to me.

How should I structure the model.json (or other configuration files) to include power measurement data for the night light entity as well?

You can use sub profiles for that.
https://docs.powercalc.nl/library/sub-profiles/

There are examples in the docs and multiple examples in the profile directory to be found.

[xxddff]

What did make you assume that? Measurements look consistent to me.

Sorry, that was probably my poor wording — let me rephrase.
After the full measurement run completed and I imported the generated config into HA, I tried random brightness values and compared the powercalc-created entity against the actual smart plug readings. Overall the fit is quite good. In most cases the deviation is small; it only reaches around 1.5W in a few cases, and those larger deviations seem more likely to occur at higher brightness levels.

Would reducing the color-temperature sampling step size lead to a more accurate final result?
eg. CT_MIRED_STEPS=3

You can use sub profiles for that.
https://docs.powercalc.nl/library/sub-profiles/
There are examples in the docs and multiple examples in the profile directory to be found.

Thanks — I'll go through the sub-profiles docs and the examples in the profile directory, and try to add the night light. That said, the overall power swing of the night light mode is very small, so my smart plug may not be able to resolve it reliably. I'll give it a try and see how far I can get.

[xxddff]

@bramstroker I have a couple of questions about how sub-profiles should handle standby power:

The Issue I'm Running Into

The main light and night light are mutually exclusive — when one is on, the other is off. So ideally, we should only be measuring energy for whichever mode is actually active at any given time.

Right now I have both sub-profiles set up with separate energy entities, which means we could potentially be double-counting standby power if both entities are being tracked.

Standby Power Configuration

I've temporarily set standby_power: 0 for both the main light and night light sub-profiles. Here's why:

When I tried setting just the main light's sub-profile to standby_power: 1.15 or setting both, I noticed it would still calculate the main light's standby power even when the night light was active. That doesn't seem right — if only the night light is on, we shouldn't be including the main light's standby consumption.

My question: Should the standby power be defined at the parent profile level instead? Or is there a better pattern I should use for mutually exclusive sub-profiles like this?

I want to make sure the energy calculations are accurate before we finalize this. Any advice would be helpful!