Ikea LED1624G9 "TRADFRI bulb E27 CWS opal 600lm" (RGB+TW 1st gen) standard "globe" shape A60

[nepozs]

Ikea LED1624G9 "TRADFRI bulb E27 CWS opal 600lm" (RGB+TW 1st gen)

status: measuring now by @nepozs but publication postponed to September November 2021 (when LED1624G9 E14 will be ready)
temporary abandoned by @nepozs as results are not reliable, next measure session planned to November/December 2021

[strelniece]

Hi, have a bunch of these, anyway to help out with measurments?

[nepozs]

You can just start measuring (I will start next week probably). but you must consider situation when you will not be able to use measured bulbs some days.

I haven't tested script v2 so far (if has it possibility to measure group of bulbs - I've been very busy about last month, and possibility of different methods of measurement were upgraded)- I've just started it in mode like v1
https://github.com/bramstroker/homeassistant-powercalc/tree/master/utils/measure_v2
so probably you can measure group of 2, 3 or even more identical bulbs - it will produce more reliable results than measure one bulb only.

Now there is similiar Ikea model (but E14) measuring

and when I was at home I've had so little free time, so I haven't prepared secondary circuit dedicated to parralel measure using HS110 (but it is in my plans also).

It is still possible to use old script versions ("v1" dedicated to selected hardware configurations):
https://github.com/bramstroker/homeassistant-powercalc/tree/master/utils/measure

[bramstroker]

See this wiki article for more information about contributing. https://github.com/bramstroker/homeassistant-powercalc/wiki/Contributing-new-lights

[marthubner]

Hi @nepozs, may I ask about the state of this measurement? I am looking forward to be able to implement these bulbs into my powercalc integration as I have 7 of those and they are the last piece missing for having all my lights energy monitored. I also have a Shelly plug S, so I could try to take the measurements myself, but I have no experience running scripts because I'm not a programmer.

[nepozs]

@marthubner
Current situation is more complicated than I expected:

• I've wasted lots of time (when I've had free time) to unnecessary measure models not announced here by others
• there is some model conflict with another Ikea product, so I must track down it to ensure quality of LUT database
• there is some really serious problem with Shelly and/or measuring script, so produced data needs lots of work to correct it manually

So I have to postpone publishing this measurement slightly - September is really impossible now :P (I've planned it to November lately) but I'm still working on it, however I'm really short of time now.

[nepozs]

@bramstroker @marthubner
BAD NEWS
I've abandoned measurement and publication temporary, at least until the cause of the measure problem will be found.

I work on the road (it is similar work to band tour) so usually I don't do any visual inspection when measuring, but today I found that the bulb does not change colour when measuring HS (and this is the only supported mode), and it is not first measure done because of I've found something strange in measured data and probably this is the reason.

I'm using some old v2 script version (version before ability to measure groups), just to not change environment every measure (every change in the script architecture takes too much time for me for restore working measure environment, so I thought about script update after finishing Ikea RGB + TW models which have power draw> 0.4W).

It would be nice if official script have inside version numbering, so it will be easy to say what exact version is in use.

If script could change saturation in reverse order (decreasing instead of increasing) it would be really easier to find cases with problems like this one, (and maybe all variables could be decreasing, as it will be easier to find any similar problem, but I suspect that LUT file have to have increasing values inside?).

I should prepare new or reopen old issues, as there are also other problems (Shelly and/or script and/or Hue API(?) and/or aiohue related, but I plan to use latest script version before - maybe some of them are corrected already? but next free time I will have next week).

I'll do what I can to get the measurements, but it takes a lot more time than I thought…

Problem with constant (stale) saturation is probably Ikea firmware or just Hue bridge related (so I've plans to clean-up its configuration after extensive tests of different Zigbee bulbs/hardware, but it is my only bridge driving my whole home lights, so I can't just do factory reset).

BTW This bulb has firmware 1.3.002 (I will try to update it in different environment or using native Ikea installation of my friend, but it needs some week to go)

[bramstroker]

@bramstroker @marthubner BAD NEWS I've abandoned measurement and publication temporary, at least until the cause of the measure problem will be found.

I work on the road (it is similar work to band tour) so usually I don't do any visual inspection when measuring, but today I found that the bulb does not change colour when measuring HS (and this is the only supported mode), and it is not first measure done because of I've found something strange in measured data and probably this is the reason.

Very strange, I never experienced this kind of issue before myself when measuring bulbs, and also did visual inspection from time to time.

It would be nice if official script have inside version numbering, so it will be easy to say what exact version is in use.

This is a good idea, will add this enhancement to the issue tracker. Also I will add the version used to the model.json. So we know which script version was used for measurements when refering in the future.

If script could change saturation in reverse order (decreasing instead of increasing) it would be really easier to find cases with problems like this one, (and maybe all variables could be decreasing, as it will be easier to find any similar problem, but I suspect that LUT file have to have increasing values inside?).

I need to check and test to be sure, but I think the order in LUT file does not matter. Than we could change the order in measure script.

I should prepare new or reopen old issues, as there are also other problems (Shelly and/or script and/or Hue API(?) and/or aiohue related, but I plan to use latest script version before - maybe some of them are corrected already? but next free time I will have next week).

There has been a lot of improvements in measure script in the last 2 months, so would be nice it you use the latest version, so we are sure you are not experiencing any old bugs.

I'll do what I can to get the measurements, but it takes a lot more time than I thought…

Problem with constant (stale) saturation is probably Ikea firmware or just Hue bridge related (so I've plans to clean-up its configuration after extensive tests of different Zigbee bulbs/hardware, but it is my only bridge driving my whole home lights, so I can't just do factory reset).

Yes I am just using the Hue API to change colors/brightness. This part is very simple code. When the Hue API correctly responds that the color was changed than it's a hardware/firmware/bridge issue.

[nepozs]

Yes I am just using the Hue API to change colors/brightness. This part is very simple code. When the Hue API correctly responds that the color was changed than it's a hardware/firmware/bridge issue.

Maybe this is the case - script sets values but don't test if it is really made, but this kind of problem (saturation stale) is specific only to some models of Ikea bulbs, and I think it maybe dependent on old bulbs firmware (but I can't update them now).

script is still runing

BTW It was really bad idea to remove "redundant" data from description of Ikea models - see this #132
future mappings probably will be LED1624G9-E27 and LED1624G9-E14 (so it will be different from printed models).

[bramstroker]

Because it is the same model but only the fitting (adapter/connection) is different I am pretty sure they will have the specs and same power readings

[nepozs]

Yes, it has

import aiohue

inside and it is v2 script (and really old, but I'm not able to change it every update)

[bramstroker]

Did you also see the docker version of the script? See https://github.com/bramstroker/homeassistant-powercalc/blob/master/utils/measure/README.md#run-with-docker.
That is a one liner to run the latest version of the measure script, and it won't pollute your system with python and all dependencies. You need to have docker engine installed on your machine. https://docs.docker.com/get-docker/

To update to the latest version just run docker pull bramgerritsen/powercalc-measure.
Than the docker run command again to run the script.

[bramstroker]

You also need to put a .env file first. Just copy it from .env.dist from the repository. Will improve the readme soon to also mention this step.

[nepozs]

I really don't remember why I still use this old script version, because I know about .env file (it is hidden in linux, so it can be hard to find).

[bramstroker]

I have added the version to measure script. When you start it the version will be printed first. Also the readme is improved a littlebit.