Jump to content

Welcome to Smart Home Forum by FIBARO

Dear Guest,

 

as you can notice parts of Smart Home Forum by FIBARO is not available for you. You have to register in order to view all content and post in our community. Don't worry! Registration is a simple free process that requires minimal information for you to sign up. Become a part of of Smart Home Forum by FIBARO by creating an account.

 

As a member you can:

  •     Start new topics and reply to others
  •     Follow topics and users to get email updates
  •     Get your own profile page and make new friends
  •     Send personal messages
  •     ... and learn a lot about our system!

 

Regards,

Smart Home Forum by FIBARO Team


  • 1

Phantom menace (RGBW)


Kuba13
 Share

Question

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!

Link to comment
Share on other sites

Recommended Posts

  • 0
12 hours ago, tinman said:

sure, that's the data coming from input, the rgbw is sending then updates with values, here how it looks like on real zniffer:

@tinman o.O you have a PRO sniffer 

 

4 hours ago, petergebruers said:

If you mean "a burst of 5 messages in one second" but then with big pauses (10 seconds or more) then I think it is OK. If you mean "it is sending messages continuously at a rate of 5 per second" then that is very high. You'll get collisions with other messages and that causes delays or maybe phantom.

@petergebruers 

no, I mean, messages followed without pauses

yes, this is the strange thing, the messages are 01-201, hc2 towards RGBW

I made this little code to see the nodes

Please login or register to see this code.

message 181 to 01 must be correct, 181 is a fibaro motion sensor

 

Well, I realized that when I installed the wind direction sensor 3 days ago, I changed the sensitivity to 0.1V to make the tests that everything worked correctly.
I have reset the sensitivity to 0.5V and these messages have disappeared. None comes out, they have disappeared completely, now all the messages make sense and seem correct.

I think this may be related to the problem, and this motivates the appearance of them.
What I did not think is that hc2 was in the middle of the problem. I thought that the problem was unique to RGBW and now I have serious doubts that the problem can be generated by HC2.

 

the last test that I have done, is that I have returned sensitivity to 0.1V and the empty messages have not returned.

 

I will follow closely

 

Link to comment
Share on other sites

  • 0
3 minutes ago, rls46 said:

Well, I realized that when I installed the wind direction sensor 3 days ago, I changed the sensitivity to 0.1V to make the tests that everything worked correctly.

Yes... Confirmed. That is something to watch out for. Two possible explanations. Your analog signal might be noisy, so it sends too often. Or the RGBW has a bug at 0.1.

 

I recommend 0.5 V and poll the device.

 

6 minutes ago, rls46 said:

the last test that I have done, is that I have returned sensitivity to 0.1V and the empty messages have not returned.

I seem to remember a similar thing. I think the "spamming" will return.

 

Maybe it depends on the output (so... Wind direction)

 

Please be careful when interpreting results. Messages can be routed. And depending on where you put the sniffer it may "hear" different things. So put it closer to the device, to the controller or in the middle and compare.

Link to comment
Share on other sites

  • 0

Hello guys,

 

I want to add here some my findings from Monday.

 

I was at customers house on Monday. This customer suffers from phantom associations from his RGBWs. When i looked with CIT to active associations i saw, that 5 devices were conected to Group 1 and no other group. Both devices are on firmware 26.25.

 

So, his new dealer reconnected analog inputs to IN 3 and 4. There i did not saw any associations there except HC2. If you suffer from phantm associations from RGBW, then i suggest to find someone with CIT. You can see it nicely in module.

 

This is state of associations of those devices right now.

Please login or register to see this attachment.

 

 

Link to comment
Share on other sites

  • 0
Guest andyhud

Guys,

 

I just spotted my RGBW controllers have a firmware update to 26.25 and one of the fixes in the release notes says "Fixes False Associations"!

 

Maybe Fibaro have fixed it with this firmware update!

 

Andy

Link to comment
Share on other sites

  • 0

Hi @andyhud,

 

In most cases update fixes the issue. I have met 1 or 2 cases. Unfortunatelly we experience some cases, where it does not :(. In my case 2 customers.

 

Regards,

Jakub

Link to comment
Share on other sites

  • 0
1 hour ago, jakub.jezek said:

Unfortunatelly we experience some cases, where it does not :(. In my case 2 customers.

Add me, that makes 3 cases. Note: maybe the firmware fixes the associations. In fact, I believe it does! Why would they lie in the release notes? I do not think so!

 

It does not solve all cases of "phantom menace" but they might not be caused by associations at all. I speculate, it is either another bug or it is simply that the CRC8 of the older 300 series chip allows for some corrupt messages (due to collisions on large networks) to appear as valid packets. That explanation has been given a few times for (a) people notice random alarm frames, making devices blink (b) people report "fire alert" due to false temperature reports... EDIT: on the other hand, it might be completely unrelated or even complete nonsense. The only indication I have: on another forum a log was posted and it clearly showed "flipped bits" in the message and also in th CRC, leading to a valid but incorrect report.

 

I am only an end-user, with no tools, no SDK, no official sniffer so I cannot diagnose this issue. That is why I replaced my RGBW in the garden with an ESP8266... I wrote the code myself, so I am my own "support". Support is always available, fast and friendly. Though sometimes clueless. But now I have all sorts of diagnostic tools!

Edited by petergebruers
Link to comment
Share on other sites

  • 0
9 minutes ago, petergebruers said:

Why would they lie in the release notes?

It did not appeared when Fibaro engineers tested it.

 

9 minutes ago, petergebruers said:

It does not solve all cases of "phantom menace" but they might not be caused by associations at all. I speculate, it is either another bug or it is simply that the CRC8 of the older 300 series chip allows for some corrupt messages (due to collisions on large networks) to appear as valid packets. That explanation has been given a few times for (a) people notice random alarm frames, making devices blink (b) people report "fire alert" due to false temperature reports...

Could be. In case of associations, i found undesireable associations there. See images below. However i did not found in history what caused it.

Please login or register to see this attachment.

 

And Network Map below:

Please login or register to see this attachment.

 

13 minutes ago, petergebruers said:

I am only an end-user, with no tools, no SDK, no official sniffer so I cannot diagnose this issue. That is why I replaced my RGBW in the garden with an ESP8266... I wrote the code myself, so I am my own "support". Support is always available, fast and friendly. Though sometimes clueless. But now I have all sorts of diagnostic tools!

Well, i just have a CIT :).

 

Now OT: That ESP8266 looks great. i would probably end with this:

Please login or register to see this link.

Sorry it's in CZ. They do not have this product listed in EN portal. Those are just 2 relays controlled via HTTP mothods or another TCP/IP protocol.

 

 

Link to comment
Share on other sites

  • 0

Thank you for reply, @jakub.jezek

 

21 minutes ago, jakub.jezek said:
51 minutes ago, petergebruers said:

Why would they lie in the release notes?

It did not appeared when Fibaro engineers tested it.

 

To be on the safe side... That was a "rhetorical question". What I want to say is: I trust the engineers, the release notes and I do not call anybody a liar...

 

24 minutes ago, jakub.jezek said:

In case of associations, i found undesireable associations there. See images below. However i did not found in history what caused it.

Please login or register to see this attachment.

 

You mean... this is a device running 26.25? The latest FW? In that case... surprising result.

 

I do not really understand what I see... I see all sorts of devices, that is unlikely to be what the customer wants. I do not see the controller? It should be there?

 

27 minutes ago, jakub.jezek said:
57 minutes ago, petergebruers said:

I am only an end-user, with no tools, no SDK, no official sniffer so I cannot diagnose this issue. That is why I replaced my RGBW in the garden with an ESP8266... I wrote the code myself, so I am my own "support". Support is always available, fast and friendly. Though sometimes clueless. But now I have all sorts of diagnostic tools!

Well, i just have a CIT :).

 

I meant: I have all sorts of tools to diagnose the ESP8266 and WIFI but no tools to diagnose Z-Wave... I've got logs from the ESP, serial debug, I can use Wireshark to monitor Wifi - HTTP packets, I've got Wifi mapping tools to test coverage...

 

You have CIT but also access to Fibaro HC logs I guess?

 

31 minutes ago, jakub.jezek said:

Those are just 2 relays controlled via HTTP mothods or another TCP/IP protocol.

 

Yeah, sure, wired, why not? Looks good to me. Don't get me wrong, I am not replacing Z-Wave in my home, because of this "phantom" problem... Just a few modules.

 

I used google translate and noticed this... is this correct? "For voltage switching, for example, 230 V, Quida outputs must be amplified by an external relay or contactor."

Link to comment
Share on other sites

  • 0
5 minutes ago, petergebruers said:

You mean... this is a device running 26.25? The latest FW? In that case... surprising result.

Yes, it runs 26.25. I was there on project site (normally just dealer is presented) so i was it.

 

6 minutes ago, petergebruers said:

I do not really understand what I see... I see all sorts of devices, that is unlikely to be what the customer wants. I do not see the controller? It should be there?

Fibaro controller should be in association group number 5, which is Lifeline group. Through this group device is reporting all states. Since it went through different device, then nothing was reported to controller.

 

Hopefully with new RGBW this will be eliminated assuming, that Fibaro would like to continue with this product with Z-Wave Plus chip.

 

7 minutes ago, petergebruers said:

I meant: I have all sorts of tools to diagnose the ESP8266 and WIFI but no tools to diagnose Z-Wave... I've got logs from the ESP, serial debug, I can use Wireshark to monitor Wifi - HTTP packets, I've got Wifi mapping tools to test coverage...

Yes, there is more diagnose tools for Wi-Fi than for Z-Wave.

 

14 minutes ago, petergebruers said:

You have CIT but also access to Fibaro HC logs I guess?

Logs are possibly in HC accessable only with root access. FIbaro does not give root access that easily. I do not have one.

 

OT:

9 minutes ago, petergebruers said:

Yeah, sure, wired, why not? Looks good to me. Don't get me wrong, I am not replacing Z-Wave in my home, because of this "phantom" problem... Just a few modules.

Our customers (Control4 dealers) use Papouch I/O modules in Control4 projects, when extra Inputs or Outputs are needed. In Fibaro projects, we try to stick in Z-Wave world :D.

 

21 minutes ago, petergebruers said:

I used google translate and noticed this... is this correct? "For voltage switching, for example, 230 V, Quida outputs must be amplified by an external relay or contactor."

Well, Papouchs relays can trigger 9 - 30 V DC through those relays. So they need contactor for 230 V AC.

Link to comment
Share on other sites

  • 0

Thank you for your clarification. So 26.25 does not completely solve the association problem... Oh. Well. Indoors, my modules seem to behave well (connected to LED strips) so there has to be something specific that triggers the association bug... Keep digging...

Link to comment
Share on other sites

  • 0

Hello guys,

 

I have talked to a person (I will not mention who was it) about this "phantom associations".

 

This peeaon claimed, that problem is in Z-Wave SDK, that RGBW uses. So scene for removing association in this topic is just temporary solution.

 

If I remember it correctly, this person said this happens only to devices with more association groups. And also something regarding ID (do not know which one now) being close to Home ID.

Link to comment
Share on other sites

  • 0

@jakub.jezek Thanks.

 

10 minutes ago, jakub.jezek said:

If I remember it correctly, this person said this happens only to devices with more association groups. And also something regarding ID (do not know which one now) being close to Home ID.

 

Makes sense to me: some data is interpreted in the wrong way... That fits the observations posted by all users here. It kind of explains why not all users, and not all RGBW have this issue, it depends on some external factors (like... homeID or whatever ID being close to something else). It also kind of explains why it is very hard to get rid off... One mystery remains: it seems to happen mostly (if not exclusively!) if the "IN" mode was selected. Right? So it is not exclusively due to the SDK, it is also application code. Not necessary program logic (so maybe not a "bug" in the traditional sense), but maybe a timing issue or data size or even hardware setup (different registers, clock, ...). Or even good old C pointer stuff (invalid pointer, leading to data corruption).

 

Thanks to the release of the Z-Wave SDK I now have the official sniffer. We can all have the Zniffer! But to be honest: I have re-purposed my RGBW module and I kind of lost interest, I kind of assumed there would not be an easy fix... Just saying, it could be possible to sniff  the network and see it happen. But as the issue is a bug on the device itself, I do not think we are going to learn anything from that.

Link to comment
Share on other sites

  • 0
On 5/4/2018 at 9:27 AM, petergebruers said:

One mystery remains: it seems to happen mostly (if not exclusively!) if the "IN" mode was selected. Right?

 

From my experience it has mostly happened with devices set to RGBW mode as i don't use them as IN/OUT devices most of the time.

(12 devices perhaps over 3 different systems)

 

I also have the same experience with IN/OUT mode (4 devices) with 4 different systems (above not included).

It was set up as a way to control light scenes with a 4-way button and so scenes happened randomly (lights o when sleeping etc)

So i stopped using the RGBW device for that reason (controlling scenes) after that.

 

Have only used HC2 on all of the above.

Link to comment
Share on other sites

  • 0

Hi guys,

 

one of my customers is experiencing the same problem as described. Turning on lights itself.

 

  • Customer have 5 RGBW devices updated on 26.25. 
  • 2 of 5 devices are linked together to act as a one device
  • another 3 of 5 devices are used to control RGBW

Did you find this topic and code mentioned in it useful?

I'm just worried if this code deletes all associations,because I use association for 2 out of 5 RGBW devices.

 

Thanks and regards!

Link to comment
Share on other sites

  • 0

Hi @plutonium,

 

Code i posted rewrites associations. But it could be toyed with to set correct association. I did that few times.

 

Code i'm reffering is in post linked below.

 

 

Link to comment
Share on other sites

  • 0

@petergebruers mentioned this topic, but for me it was not the RGBW. It was actually a Fibaro motion sensor. For some reason it added node id 255 to the basic report. Basicly the whole house was going wild, since its a motion sensor things would go on/off and i only found out while i went in that room (normally the dog triggered it and i didn't find the culprit) removing the association fixed it though.

Link to comment
Share on other sites

  • 0

I too have suffered from these problems. 

A power shutdown seems to fix my issues for a few weeks, or less. 

I’ve also discovered that a ‘reconfigure’ command via Vera to the RGBW device helps for a period, however results in renumbering and renaming issues. 

 

Question: Is there a LUA command that I can send to restart the RGBW module? Suitable for Vera? I think it would be worthwhile giving a go. 

Edited by Brock
Link to comment
Share on other sites

  • 0
On 8/7/2019 at 12:26 PM, Brock said:

Question: Is there a LUA command that I can send to restart the RGBW module? Suitable for Vera?

I don't know vera, but if it has commands to change groups (aka "associations") then that might work (not to permanently fix the problem, but to avoid having to do reconfigure or power on reset to fix the phantom when you have it). The idea is you read the groups from the device, remove the bad ones (or rather keep only those you want which will very likely be the controller only) then write them to the device... This will "scrub" the phantom associations.

 

That said... If you know someone with a HCL or HC2 you can update the RGBW to 2.7 aka 27.27 which disables associations (unless you change a parameter)

 

- Added a configuration parameter that disables associations to other devices.
- The device behavior was improved while receiving next turnOn command during the color transitions.
- Version 27.27 fully supported from HC2/HCL 4.550.

 

They mention "supported from HC2/HCL 4.550." -> this is needed to show the new parameter

 

Parameter 49 Associations
This parameter allows to disable sending associations to other devices.

 

Default: "Do not send associations"

 

I've updated mine (2 out of 3 of them had to be excluded an included to get the fw update going, but after that no issues). But to be honest, ... the one that caused the most trouble (see older posts of mine) was replaced so I cannot confirm the phantom is gone. Though, of course, the idea of "disable sending associations to other devices" sounds like a permanent solution ;)

 

Link to comment
Share on other sites

  • 0

Greetings!

 

For the first time in years I experienced ghost associations related to Fibaro FGRGBW411M on my system.

 

First about my setup:

Fibaro Home Center 2 on FW version 4.600

Fibaro FGRGBW411M 5 modules with FW 27.27

Fibaro FGRGBW442CC 1 module with FW 5.0

 

This ghost menace mostly happens in the morning or evening. When RGBW module is turned on or off in bedroom it also turns on or off several other RGBW modules.

 

I have checked associations in GUI and didn't find anything suspicious.

Also parameter settings on all modules:

6 is set to 'Normal (dimmer)'

44 is set to 0

45 is set to 0

49 is set to 'Do not send associations'

 

What I did until now is:

1) Run associations scene provided by Fibaro and @jakub.jezek

2) Soft reconfigured bedroom RGBW module. After reconfiguration there was no suspicious associations found

 

Will now continue monitor these modules. Luckily for me I can see most of them at the same time so can be sure that ghost is on the run if it happen again.

 

I almost forget to mention that all my RGBW modules are controlling RGB(W) strips and no inputs are used or connected

  • Like 3
Link to comment
Share on other sites

  • 0

hello guys, I rely on you.
I had problems of associations with a rgbw that I could not update to fw 27.27, I used the scene to delete the associations, but the problems always occurred at night. I removed the RGBW from the network.
HC2 restarted, mesh network redone.

now I always have the lights that come on at home, always at night, also some wallplugs or all the lights come on.
before resetting everything and starting using recovery mode (I have 55 devices, 50VD and 20 scenes) do you have any advice?
I tried to disable all scenes, exclude all vd but it always happens at night.
I can't find associations on the devices

Please help me...

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Answer this question...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...