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


  • 0

Problems RGBW mode 0 - 10v


rls46

Question

I have a RGBW module with an ultrasonic sensor to measure a diesel fuel tank.
The fact is that the sensor is very sensitive, and although I have increased the value of the change of V to inform, the RGBW module sends constant changes of value.

I have noticed that after a while, the response to the devices becomes slow and devices are turned on and off randomly (most times other mods RGBW).

It seems to me that the RGBW mode floods the zwave network and HC2 is not able to process all the information with solvency, producing general delays and random on and off commands to modules.

Or it could be that the RGBW module when processing so many changes of values, sends corrupted information to the zwave network producing these symptoms.

Now I have only two possible solutions that make the whole system return to normal.

1.- turn off the power to the module and reconnect it (wait for it to happen again)

2.- Do not use it.

I have tried with another RGBW module with the same result.

Is there a solution to avoid this?

Link to comment
Share on other sites

12 answers to this question

Recommended Posts

  • 1

@rls46 I really appreciate your detailed reports.

 

Yeah, no good explanation exists. That is why it is called "phantom menace"

 

Yes, it looks like association. I do not disagree with that in any way. Also, it could be sending to random nodes. Anyway, the Fibaro script does not work for me.

 

This is why I keep repeating details of my setup: nobody was ever able to reproduce the problem! You can make it go away for some time but it comes back. But if you want it to go wrong? You cannot do that!

 

So I really appreciate everyone's effort to speculate what causes it. Every detail counts.

 

I think the problem will get worse if you set reports to 0.1 instead of 0.5 because it sends more reports. More reports may mean more phantom...

 

But I tested this... in situ my theory made sense. I thought it was possible to "trigger the phantom"

 

So I took the module in the house, set it to 0.1 threshold and connected it to a function generator. I swept different 0 - 10 V waveforms during the day. No phantom!

 

What I did not do... Repeat the same test in the garden shed.

 

Note.  Connection in my garden shed is sometimes direct, sometimes routed. I live in a small house, indoor means direct connection.

 

Is it possible, your working outdoor sensor has a direct connection?  And your phantom sensor is indoor, but has a routed connection? Because of walls and obstacles

 

@jakub.jezek I agree, it is not power supply. Tested that. Yes, seen that with old wall plug as well. I ran that script on a daily basis, it did not help. Maybe you need to run it a lot more. I think it cannot prevent all phantoms because it acts when it is too late. But that's speculation. Thanks for still trying to solve this phantom case.

 

Edit: @rls46 I do not want to underestimate you at all, I see you have >107 posts - but I made this explanation regarding direct & routed connection and I got some positive feedback so here it is...

 

Here is a somewhat generalised  sketch of why that could happen:

 

Please login or register to see this link.

 

  • Node C represents your sensor. It tries to send an update to your controller.
  • Z-Wave radio waves travel in straight lines.
  • The direct connection might pass through walls at a certain angle. This makes the appear much thicker, see grey box. Module A's waves go through the wall at a right angle, so the wall's apparent thickness is its normal thickness.
  • I have drawn a red arrow to represent a weak signal. The HC might not hear it. Or the module might be unable to receive the acknowledgement, sent by the HC.
  • The second red arrow is a communication attempt via another module. Unfortunately, the large metal object blocks the radio waves, so this fails too.
  • Module A and B can communicate directly with your HC, hence the green arrows.
  • Because of mesh networking, A, B and C can communicate with each other through the orange arrows.
  • Communication from C to HC should work, but with a slight delay, because module B has to buffer and forward the information to the HC. With only one hop, this is barely noticeable.

Can you apply this to your setup?

 

For context... it is part of this topic:

 

 

 

 

Edited by petergebruers
Link to comment
Share on other sites

  • 0

I agree with what you are saying. Personally, I do not think there is a solution. I tried really hard but gave up on that module for this purpose... Very long topic, but I am afraid you'd loose context if I gave a summary.

 

One last thing you could try... if you are a bit of an electronics DIY person, is add a low pass filter & buffer, so the RGBW input is cleaner and it doesn't have to report as much as it probably does now... Did not fix my problem but I have to say, my module was in a garden shed and probably did not have a direct connection (only routed connection) to my controller.

 

 

Link to comment
Share on other sites

  • 0
  • Inquirer
  • Thank you petergebruers, you confirm my suspicions. I do not have that level of knowledge of electronics.

    nor does isolating the module in another hc2 (another z-wave network) solve this problem. I have tested it.  :(

     

    Link to comment
    Share on other sites

    • 0

    Hi @rls46,

     

    Do you have RGBW controller updated to version 26.25?

     

    My customer reported me, that phantom associations disappeared, when he reconnected analog sensor from IN1 and IN2 to IN3 and IN4.

     

    Best regards,

    Jakub

    Link to comment
    Share on other sites

    • 0
  • Inquirer
  •  

    hi @jakub.jezek

    yes, I confirm that it is in v26.25 and device 0-10v in IN4

    I have another rgbw with a wind sensor and outdoor light sensor.
    This does not give me any problem, the sensors are less sensitive and configured to reports with a difference of 0.5V.
    connected to IN1 and IN2

    the ultrasonic sensor is not really 0-10v is 0-5v, although RGBW can be set to 5v, maybe the sensor is too sensitive, I'm waiting for another sensor to arrive with 0-10v to test, if it does not work I also dismiss this project

    Link to comment
    Share on other sites

    • 0

    @rls46 please try @jakub.jezek's suggestion, although it did not work for me. It is mentioned in that "phantom menace" topic... I use the same 3x RGBW  as ever, but indoors (direct connection to HC!) so no longer one in the garden shed, and I also do not use the inputs. I haven't seen a "phantom" since I took it indoors (on september 28, I wrote "The RGBW module in my garden shed is now punished.").

    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • @petergebruers 

    the RGBW that gives me problems is located inside, has direct connection to HC2, 3 meters distance.
    And the weirdest thing is that it is registered in the hc2 that I have to do tests, and it affects the devices of hc2 in production, so as if it indicates in the other thread, it seems that it is directly associated with other modules.

    the one outside does not give me any problems,

     

    a difference between a RGBW and the other, is that the outside (not problems) is powered to 12V and the inside (problems) to 24V

    Edited by rls46
    Link to comment
    Share on other sites

    • 0
    7 minutes ago, rls46 said:

    a difference between a RGBW and the other, is that the outside (not problems) is powered to 12V and the inside (problems) to 24V

    Problem in powering device? Unlikely, installers that i know did many installations using 24V DC power supply.

    I believe phantoms are related to RGBW, but have different origin. Also funny thing. I saw phantoms caused by Wall plug in one end-customer system.

     

    @rls46, can you try scene below? it is scene, that removes unwanted associations.

    Please login or register to see this attachment.

    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • @jakub.jezek @petergebruers
    Yes, when I get home, I try.


    It seems to me an association, because when it begins to happen, the actors that turn on and off are always the same, until it is resolved.

    Sometimes I have managed to solve by removing the electrical current to the module and putting it back on, but after 2 or 3 days it has happened again.

    when it happens again, it is not the same actors who turn on and off as before, there may be more, less or totally different actors.

    I am very convinced (although I can not prove it scientifically) that the trigger is the massive transfer of values. At the beginning, the outdoor wind sensor had adjusted it to the minimum 0.1V and this happened to random module ignitions. I set it to 0.5V and it never happened again (two year ago, in the same RGBW there is the wind sensor and the light sensor). This module does not have direct vision with hc2, I am sure, because it is on the outside, in a corner of the plot of the house.

    the RGBW that gives problems with the ultrasonic sensor, has a direct connection with safe hc2, it is 3 meters from hc2 without any obstacle of any kind. The most curious thing is that in hc2 there is only this module included. But the phantom threat affects modules included in another zwave network, which are in production hc2.

    hc2 I'm pretty sure it does not have to do with this problem, I've seen the zwave internal registers of hc2 and it does not seem strange, so I think it's a direct association of the RGBW commands or send on and off to nodes.

    I have a fairly large installation, 189 physical devices. Although I do not comment much in the forum, I have enough experience with hc2, many reinstallations made, many hours with fibaro :?.

     

     

     

    some things surely do not explain me well, I use google translator

    Link to comment
    Share on other sites

    • 0
    30 minutes ago, rls46 said:

    It seems to me an association, because when it begins to happen, the actors that turn on and off are always the same, until it is resolved.

    Confirmed. It is "not an association as we know it". Support told me, you cannot "see" the assocation. Though they were not sure themselves. Next time you have it, please try "read configuration", the "association" button and "soft reconfigure. The association might show up... I did one attempt but I did not see anything. Turn off helps.

     

    32 minutes ago, rls46 said:

    but after 2 or 3 days it has happened again.

    Confirmed.

     

    33 minutes ago, rls46 said:

    it is not the same actors who turn on and off as before

    Confirmed.

     

    33 minutes ago, rls46 said:

    that the trigger is the massive transfer of values.

    Confirmed, fast updates make it worse. But it also happens when I connect my 0-10 V Lux meter plus filter. It has very low noise and very slow changing voltage! The meter has a special "slow motion setting". I am an electronics engineer, I don't brag about that, but I have the tools to measured the varying signal.

     

    Yes... 0.1 or 0.5 made a difference here as well.

     

    36 minutes ago, rls46 said:

    This module does not have direct vision with hc2, I am sure, because it is on the outside, in a corner of the plot of the house.

     

    You might be surprised. I can get a direct connection about 30 meter around my house, but not at all angles. The angle is important.

     

    Please have a look at my picture in the previous post. You might not have seen it because I added it later

     

    Please try a Fibaro Key Fob or an FGMS as range detector (it is in the manual) to get a rough idea.

     

    41 minutes ago, rls46 said:

    the RGBW that gives problems with the ultrasonic sensor, has a direct connection with safe hc2, it is 3 meters from hc2 without any obstacle of any kind.

     

    That is interesting to know. It goes against my theory that connection is a key point, but that is OK. It still can be a packet collision, see next point.

     

    So you could try IN3, the script, or use it as RGBW only.

     

    42 minutes ago, rls46 said:

    The most curious thing is that in hc2 there is only this module included. But the phantom threat affects modules included in another zwave network, which are in production hc2.

     

    Less surprising than you may think. All Z-Wave devices (little white lie) use the same channel, so they all interfere with each other. It is like putting two wifi networks on the same channel. It will work, but you get reduced bandwidth.

     

    Youe need a "Software Defined Radio" to get some idea of the cleanliness of the Z-Wave channel.

     

    44 minutes ago, rls46 said:

    hc2 I'm pretty sure it does not have to do with this problem

     

    Plausible. But the radio in your HC2 is the older 300 series, just like the one in your RGBW. Z-Wave Plus devices are called "series 500" and they have a better radio. I cannot say, I cannot even speculate, if it would make a difference if the HC had a newer Z-Wave chip. But thay claim 30% extended range, so the radio has to be better.

     

    46 minutes ago, rls46 said:

    association of the RGBW commands or send on and off to nodes.

     

    It is rare, but it can also cause a broadcast alarm frame. That will make your Fibaro Relays and Dimmers blink for 10 minutes.

     

    47 minutes ago, rls46 said:

    I have a fairly large installation, 189 physical devices.

    I am impressed. > 100 physical here. Mind you, if you had only 10 modules, you probably did not have the phantom!

     

    48 minutes ago, rls46 said:

    Although I do not comment much in the forum, I have enough experience with hc2, many reinstallations made, many hours with fibaro

     

    Thanks. Glad to meet you here. So we can talk on a technical level. Good!

     

    Good luck!

     

    I am happy... someone tries to chase the phantom... again...

     

    50 minutes ago, rls46 said:

    some things surely do not explain me well, I use google translator

     

    It is better than my Spanish. But you can always try, then I'l combine it with google translate, some french and popular song lyrics to decipher it. Dos servezas por favor, I am Flemish, I like a beer once in a while.

    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • @petergebruers 

    yes, we have the same symptoms, it's good to know that I'm not alone
    I reconnected after two days off, now 5 minutes running and the ghost has not appeared, everything is working correctly.

    we have to wait

    no more modules added to hc2, only RGBW, must have direct connection to hc2. but I show you the distance in the picture.

    when I reappear I will do more tests and inform

    ok, I understand your drawing, it is possible that the outer RGBW has a direct connection.

     

    wine is also good for me  ;D

    Please login or register to see this attachment.

    Edited by rls46
    Link to comment
    Share on other sites

    • 0
  • Inquirer
  • @T.Konopka can you close this thread?

    we continue with this, thanks

    Please login or register to see this link.

    Link to comment
    Share on other sites

    Guest
    This topic is now closed to further replies.
    ×
    ×
    • Create New...