Phantom menace (RGBW)

[Kuba13 49]

Right so here is a problem:

SOMETIMES (rather sparsly but still enough to warrant me bugging Fibaro about it) RGBW makes phanthom associations for no bloody reasson.

 

It happened a while back and only replacement of specific RGBW module seemed to help that was in pre "100" era tho.

Most recent case i have are 2 RGBW modules (v 2.5) running on same HC2 4.100, reading tempriture and humidity via IN1 and IN2.

 

Confirmed at least one of them caused it (once disconnected from power the random behaviour stoped), second of those RGBW is very strong suspect. Sad thing is that the second RGBW seems to have worked fine untill the first one was disconnected.

 

Now i seek aid, advice and wisdom. if annyone had this happen to them please add your controller, its FW and mofule's FW aswell - lets see if we can help fibaro figure this one out!

[jakub.jezek 370]

1 minute ago, M.Baranowski said:

 

We have a workaround - a LUA scene, which clears all the associations once a day, I will send it to you.

 

Is it this scene (below)? Maybe slightly modified.

 

Please login or register to see this attachment.

 

[petergebruers 1,268]

17 minutes ago, AutoFrank said:

I have multiple ghosts that appear with some regularity

I'll try and check all devices and see if any associations have been created and delete them

 

failing that I'll remove the RGB and see if the issues go away

 

 

I'm pretty certain that you can't see the phantom associations. Also, by just saving the correct associations again, from the GUI, you won't fix it. The script by Fibaro, posted in this thread, forces a rewrite of the associations, so that's a better fix. But after a certain time, the phatonm reappeared (I'm talking about the previous FW 25.25) so I don't think it is a good fix.

 

I would not "remove" the module, but power it down, because for me it is easy to access the module. In the past, that's how I proved the RGBW was the cause of "phantom turn on".

 

11 minutes ago, jimicr said:

(...) I *MIGHT* have an idea:

Does the unit communicate directly to the HC2? Maybe if the signal goes thru the mesh it "piggybacks" some orphan or extra code that executes actions in the unit it bounces on? 

 

Z-Wave packets contain a destination address. So there are a few possibilities, including a variation of what you say.

 

- A device has a bug, so it thinks it needs to device "X". If 'X' exists, and is a light, the light turns on. This is, for me, the most probabl explanation. They fixed or at least reduced this, by creating new firmware. While I was typing this reply, @M.Baranowski confirmed this theory.

- A device does NOT send to "X", but somehow the message gets "distorted" due to interference (another device talking, some static from a thunderstorm, ...). By design, Z-Wave packets have a 8 bit CRC to check if that happened. Personally, I think a CRC 8 is rather weak, and I might be right. More recent devices and all Z-Wave Plus devices support CRC 16 (as an extension, so they are backwards compatible).

- The receiving device has a bug, and thinks a message for "Y" was actually for him. I can't say I have indications that this happens, but it is a theoretical possibility.

6 minutes ago, M.Baranowski said:

We have a workaround - a LUA scene, which clears all the associations once a day, I will send it to you.

 

When I was still running 25.25, once a day was not enough...

[T.Konopka 8]

 

7 minutes ago, petergebruers said:

I'm pretty certain that you can't see the phantom associations.

 

If it occurs, you can see associations in BUI after soft reconfig of RGBW. Not always, but it can be seen.

 

7 minutes ago, petergebruers said:

I would not "remove" the module, but power it down, because for me it is easy to access the module. In the past, that's how I proved the RGBW was the cause of "phantom turn on".

 

 

You are right, that's why I asked about other forgotten RGBW's, mounted somewhere in the house.

Edited June 28, 2017 by T.Konopka

[AutoFrank 372]

On 4/24/2017 at 7:43 PM, jakub.jezek said:

HI @AutoFrank,

If problem is in RGBW devices, then scene in attachment should help. It cleans up all associations. It cleans RGBW of associations through API.

 

Please login or register to see this attachment.

 

Thanks @jakub.jezek

I'll run that , much appreciated ...

 

Do you run it daily or how often ?

 

_f

 

 

[jakub.jezek 370]

@AutoFrank,

 

Well that depends on project. In customers project that scene runs once a day and that is enough. If needed, then you can run it few times per day.

 

Also that scene is not mine, but from Fibaro technician, that made this scene for customer (big problems with RGBWs). I have made a copy of that scene, for customers with same RGBW problem. But i'm glad that i did sent it just once to all my customers.

[petergebruers 1,268]

19 minutes ago, M.Baranowski said:

 

 

If it occurs, you can see associations in BUI after soft reconfig of RGBW. Not always, but it can be seen.

 

 

You are right, that's why I asked about other forgotten RGBW's, mounted somewhere in the house.

 

I've only tried soft reconfigure once or twice, I couldn't say if there were rogue associations. I'll keep that in mind. If I see a ghost, I'll try soft reconfigure to confirm rogue association(s).

I own three RGBW modules, all powered  and all running 26.25. Two set to RGB(W) mode, one set to IN/OUT.

[AutoFrank 372]

2 hours ago, M.Baranowski said:

We need to be sure, that every RGBW controller has been updated do 26.25 version and if you don't have any other RGBW, which is working, but is not connected to Home Center (therefore weren't updated).

 

Thanks @M.Baranowski

I have two RGBW modules 

1 updated successfully to the new FW

Please login or register to see this attachment.

 

The second one refused to update even after multiple resets and exclude/include attempts. I varied the distance from 2cm to 2m  and nothink works.

It will include okay but will not update the fw

 

I have removed this for now from my system and may end up throwing it in the bin.

 

[T.Konopka 8]

If it stuck during update, hard reset should resolve the problem. Maybe try to reset it one more time?

Eventually send it to us, we will take a look at it.

Do you know the procedure?

[AutoFrank 372]

14 minutes ago, M.Baranowski said:

If it stuck during update, hard reset should resolve the problem. Maybe try to reset it one more time?

Eventually send it to us, we will take a look at it.

Do you know the procedure?

 

HI @M.Baranowski

I'll give it one more go

 

Whats the procedure to send if unsuccessful ?

 

Thanks

_f

[T.Konopka 8]

We will send you an instruction in your email.

[AutoFrank 372]

Hi 

no luck getting rid of the menace

The scene didn't really help either, I ran it once per hour and still had odd associations happening.

 

I opted top remove the RGBW but it didn't exclude successfully and I tried a force delete 10 times and it still reappears.

I now have it removed from polling and disabled but it keeps giving me an initialisation error on start up and when i run  the set assoc code it still finds it and tries to remove associations, which seems odd to me.

 

I'm going to ask fibaro support to remote and remove the device and will send the two devices back to fibaro for postmortem

Until that happens I'm not considering myself out of the woods

I'll try and search for a replacement as I do like the led strips

 

_f

 

 

 

 

[petergebruers 1,268]

10 hours ago, AutoFrank said:

(...) I opted top remove the RGBW but it didn't exclude successfully and I tried a force delete 10 times and it still reappears.(...)

 

I think it is really, really odd that you cannot exclude or include these modules properly (after umpteenth attempt). I wonder... could it be a power supply problem? I only happen t know one power supply related issue: an Universal Binary Sensor with a Temp Probe can act really weird if the power supply isn't good. It is probably related to noise. And the Swipe, but that is to be (somewhat) expected because it uses capacitive sensing.

 

Anyway... what kind of supply is it? Do you have more than one to try?

 

BTW you probably force deleted the device while it was still powered... Did you try to remove power, then "force delete"?

[AutoFrank 372]

3 hours ago, petergebruers said:

what kind of supply is it? Do you have more than one to try?

 

Hi @petergebruers

I was using a laptop power supply, a lenovo 20v, 65W so should be okay

3 hours ago, petergebruers said:

BTW you probably force deleted the device while it was still powered... Did you try to remove power, then "force delete"?

 

I tried both and no luck

 

Fibaro support dialed in this morning and removed the device so hopefully should be okay now

 

will keep people posted....

 

Thanks

[petergebruers 1,268]

The PHANTOM LIVES! And it is more powerful than ever!

 

Today, the phantom turned on TWO dimmers, and would not let go! If I turned them off, a few seconds later the phantom would turn them on again. Light intensity seemed to increase slowly.

 

So, I powered down my HC2, so I am 100% it is not the cause.

 

Phantom still there!

 

Then... I powered down the Fibaro RGBW in the garden shed, connected to a 0-10 V Light Sensor.

 

When I did that, I could hear a slight whisper from the module. I'd swear it said: "AAAAAAA ---- I'll ---- be ---- back" (*)

 

After that, I could finally turn off the dimmers.

 

Conclusion: RGBW 26.25 still causes random control of other modules!

 

(*) from the movie: "The terminator". Maybe it was my imagination.

 

[jakub.jezek 370]

Hi @petergebruers,

 

In which input you have connected your Light sensor? Does phantom association appear if you reconnect sensor to IN 3 or IN4? Mine customer did that (his idea not mine) and it seems since that it work as it should. He has temperature sensor and humidity sensor connected.

[petergebruers 1,268]

On 10-7-2017 at 7:46 AM, jakub.jezek said:

Hi @petergebruers,

 

In which input you have connected your Light sensor? Does phantom association appear if you reconnect sensor to IN 3 or IN4? Mine customer did that (his idea not mine) and it seems since that it work as it should. He has temperature sensor and humidity sensor connected.

 

I'm sorry, I forgot to reply. Thank you for reminding me, I will certainly try this as a workaround. At the moment, it's IN1 and I am preparing an email for Fibaro, with all the factors I think that may be important for them to try, when they want to reproduce this issue. So far, they have not been able to do so. For now, I keep my module powered down and I hope when I power it up the "bad things" will still happen, so I can do more tests. Because of holidays, things take a lot of time here.  

[petergebruers 1,268]

On 10-7-2017 at 7:46 AM, jakub.jezek said:

Hi @petergebruers,

 

In which input you have connected your Light sensor? Does phantom association appear if you reconnect sensor to IN 3 or IN4? Mine customer did that (his idea not mine) and it seems since that it work as it should. He has temperature sensor and humidity sensor connected.

 

Because it is summer, I decided it was the right time for some nice LED garden light. So last week I implemented your suggestion, but with a twist.

 

Last week I did this:

 

R channel is now the output to the lights.

IN4 now has the filtered output of my illuminance sensor.

 

So far, so good. No phantom, light values are updating. But one week is unfortunately not enough time to say this with suffiecient certainty....

[jakub.jezek 370]

@petergebruers, since Phantom associations appeard always after few days (max), i would say week is a success.

 

I did not received any negative feedback from dealer and customer about Phantom associations for a few months, so i think it is good for now.

[petergebruers 1,268]

On 3-8-2017 at 0:24 PM, jakub.jezek said:

@petergebruers, since Phantom associations appeard always after few days (max), i would say week is a success.

 

I did not received any negative feedback from dealer and customer about Phantom associations for a few months, so i think it is good for now.

 

Hi! I'm not contradicting you in any way, but I think a week in my case is not enough, because, for instance, the ghost may turn on the light in the garden shed and I might not notice, because not all my lights warn me in case of automatic turn-off. On the other hand: I turn of the light in the cellar, by means of a script, and I mail myself when this has happened. I am certain that did not occur, so indeed that is good news. But not all my lights use this script...

 

Edit 24 aug 2107. I am not convinced the ghost is gone. Or maybe I have a different ghost. The ƒacts: i have "auto turn of scripts" that send me mails. I had a few of those, past week or so, and I am quite certain the corresponding lights "turned on by itself". I will now try a new tool (kind of Z-Wave sniffer). I cannot tell about the tool yet, I only had 15 minutes to toy with it. Mayby I have some time in the weekend.

Edited August 24, 2017 by petergebruers
Still a phantom

[petergebruers 1,268]

The RGBW module in my garden shed is now punished. It turned on the LEDs all by itself. Maybe it is not the root cause of phantom activity, but because it is at the edge of a direct connection to my HC2 I decided to try an alternate approach, that cost me 10 EUR plus some time and effort:

 

Please login or register to see this image.

/monthly_2017_09/H801_mod.jpg.649da8e7e261ecb6b1f6211adac74f8d.jpg" alt="H801_mod.jpg.649da8e7e261ecb6b1f6211adac74f8d.jpg" />

 

It is the "H801 Wifi". It is a 12-24 V RGBW1W1 PWM LED driver. It is about twice the size (volume) of an RGBW module. 100% Chinese. Based on ESP8266. It comes with and app but I did not even bother with that. I added wires to the board so I could flash my own firmware.

As you've probably guessed, this module has no 0-10V input. Correct! But right now I am soldering the ADC pin to a resistor divider so I will soon find out if I can make this work