dead nodes

[Mib 15]

I have a problem with dead nodes. They follow me through every software version. Now I run on 4.058 Beta.

I tried the mesh network reconfiguration. I get tihs result:

 

[09:00:50] 476: Requesting neighbor list from the node is in progress
[09:01:10] 476: Getting new neighbor list failed

 

So what does it tell me? What can I do to solve this prob?

Thanks

MIB

[Mib 15]

@ power-trip-forum

full ack!

How can I have my HC2 checked?

[Mib 15]

Rafał Ciesielski

03 Dec 15

 

No, I’m serious

Please make a reconfiguration a mesh network for all devices (configuration tab → Z-Wave network).

If it doesn’t work please change the location of Your central.

Regards

 

 This is the support answer but still got dead nodes and "Getting new neighbor list failed" message ...

[Guest Kuuno]

ok i think this is the same case i have/had

 

try this: power off all fuses for 5 sec in the house and power on again (HC2 can stay powered) and if it all starts working then it is a rouge module.

 

i bought a radio frequency scanner and looked what happens in the air and HC2 is not transmitting any data but ALL of the modules are transmitting like crazy. It seems like they end up in an infinite loop

 

the next steps are to take out fuses by groups and find where the module is...

 

i tell you honestly...if it was a broken Hc2 it would be simplest fix ever!

[Mib 15]

It was a busy weekend. The result of all testing which was suggested here: If I use no Motion Sensor I get no dead nodes

Please login or register to see this image.

/emoticons/default_sad.png" alt=":(" srcset="https://forum.fibaro.com/uploads/emoticons/[email protected] 2x" width="20" height="20" />

I can even test range successfully with a wall plug. Reconfiguration of Mesh network works partly if I do it one by one.

First I get a failed - then I get a received.

So all of my 7 Motion Sensors are out of order

Please login or register to see this image.

/emoticons/default_sad.png" alt=":(" srcset="https://forum.fibaro.com/uploads/emoticons/[email protected] 2x" width="20" height="20" />

MIB

[petergebruers 1,265]

I fully added all my mains powered devices last Thursday. Thats's 35 nodes. Works like a charm! Very responsive, Wall Plug probes green everywhere. I've also put back about 15 battery powered devices and thats also 100% operational, including scenes. That includes 5 Aeon MultiSensor and one Aeon MultiSensor Gen5. I do own a FGMS Motion Sensor, but I have not included it yet.

Are your Motion Sensors the FGMS? Did you set p 4 "Illumination Threshold" to something < 50? That could explain why they cause lots of traffic, but I can't say if it would explain everything. But if 7 of these sensors start to report small lux changes at the same time? Sometimes, but not always, Fibaro support can see that in your logfiles. In fact, in my case, there was probably so much traffic, that paradoxically, nothing appeared in the log. Actually, a very quiet log could be a sign of trouble.

Anayway, the frequent & erratic reporting of an FGMS was reported as bug 2762. Workaround published by Fibaro:

"For now you can disable reports on lux change (parameter 40 = 0) or set it to value greater than 100 lux and turn on frequently automatic reports (parameter 42) -> for example once for 2 minutes. If you had parameter 40 on low value (<50) it won't affect your battery life (or impact will be minimal)."

The workaround improves the accuracy of the reporting and reduces the "chatter" of the sensor.

[kandetvara 3]

I can confirm that the fibaro Motion sensors are a source of Problems. I have 3 of them and they literally do whatever they want and they do not respond to changing the values or sticking to the changes. In some instances they go to infinite loops.

 

Some examples:

 

1. I set parameter 40 to 10lux. Nevertheless all of the motion sensors report changes of 1 lux (!!). Up and down – from 63>64 and from 64>63 – for HOURS. Including excluding doesn’t make a change (right beside the HC2). Reconfiguration doesn’t either.
2. Parameter 60 is set to 0,6. Doesn’t matter either – temperature changes are reported randomly to the HC2.
3. Parameter 66 – to offset the temperature has no effect whatsoever

 

And the list doesn’t end – they must jam the whole network. I can not proof this but if I look at my log there must be massive issues with the motion sensors…

 

All the best

 

kandetvara

[petergebruers 1,265]

(...)they must jam the whole network. I can not proof this but if I look at my log there must be massive issues with the motion sensors(...)

I can't prove this, I only have one to experiment with, but it sounds "plausible". Filling the log and triggering scenes every second doesn't serve a purpose either. Did you try the workaround I mentioned in my previous post (i.e. set reporting to default for temp & lux or at least make sure p40 > = 100 and use wake up interval 120 sec or higher)?

[kandetvara 3]

hi,

 

no I did not set it back to default or para 40 to > 50 or 100, since the device would lose its purpose - which is to measure the light intensity to control the roller shutters and the lights... with a lux value of > 50 lux the tool is worthless...

 

But thanks for your tips - always welcome !!

[petergebruers 1,265]

It would report lux 100% accurately at the wake up interval. Parameter 40 doesn't count in that case. It's a bit like the MultiSensors. I have mine set to 5 minute wake up. Would that work for you?

[boerremk 83]

I also use parameter 42 for lux (and 62 for temp), works perfect. I have mine set on 480sec.

[Momos 102]

Hi All,

 

Talking about dead devices... maybe this is related or not.

 

In the HC2 interface there are some rooms where is see in the corner that I have dead devices.

 

Please login or register to see this image.

 

However nothing happens when I click it and more important there is no device in that room which I can not control... All seem to work ok.  Unless some AI form installed some new devices and didn't tell me about it.

 

Any ideea what that could mean?

 

Regards

[boerremk 83]

Hi All,

 

Talking about dead devices... maybe this is related or not.

 

In the HC2 interface there are some rooms where is see in the corner that I have dead devices.

 

Please login or register to see this image.

 

However nothing happens when I click it and more important there is no device in that room which I can not control... All seem to work ok.  Unless some AI form installed some new devices and didn't tell me about it.

 

Any ideea what that could mean?

 

Regards

 This is not room based, but section based, every section has this feature. In you case on "Parter" (section) level, not "Office"(room)

[Momos 102]

True, but still I don't have any device which I can't control... All devices are responsive.  I would see the X mark on a dead device right ?

[boerremk 83]

True, but still I don't have any device which I can't control... All devices are responsive.  I would see the X mark on a dead device right ?

It is part of the interface, even if there is no need for it to be. If a node is really dead, you will indeed see the X mark.

[Momos 102]

Hmm might be true but I don't believe it. Some sections have that "wake up dead nodes in section" , some don't.  A better explanation must be, as yoda says

[HomeSystem.sk 69]

Hey Guys, I think I have solution for this, not sure if someone figured that out already.

 

One of my clients has the same issue, many dead nodes, but his mesh network is quite good with many dimmers and relays. The dead nodes could not comunicate at all, even dimmer located just 2 meters from HC2 did not wake up. The switches connected to S1 worked.

 

We removed and added all the dimmers with this issue and it worked for few minutes, than again dead nodes and while I was trying to wake them up, even more devices got dead (only switches and dimmers, no battery powered devices).

 

Long story short - the issue was parameter 40 (can be 60 as well) on motion sensor. It was set up on 30 seconds or something like that. It made the network crazy. After changing to 300seconds and waking up the sensor, everything just started to work perfectly. So please, if you have this issue, check theis parameters, they can not be that low.

[fi-bar-oh 47]

Hey Guys, I think I have solution for this, not sure if someone figured that out already.

 

One of my clients has the same issue, many dead nodes, but his mesh network is quite good with many dimmers and relays. The dead nodes could not comunicate at all, even dimmer located just 2 meters from HC2 did not wake up. The switches connected to S1 worked.

 

We removed and added all the dimmers with this issue and it worked for few minutes, than again dead nodes and while I was trying to wake them up, even more devices got dead (only switches and dimmers, no battery powered devices).

 

Long story short - the issue was parameter 40 (can be 60 as well) on motion sensor. It was set up on 30 seconds or something like that. It made the network crazy. After changing to 300seconds and waking up the sensor, everything just started to work perfectly. So please, if you have this issue, check theis parameters, they can not be that low.

 

what is that parameter for? and exactly what device? 40 / 60 have nothing to do with timing on my sensors?

[petergebruers 1,265]

This was reported on bugzilla, but not directly as a cause of dead nodes. But that sounds very plausible to me. If p40 < 50 or so, the sensor reports very small changes in light level, probably very fast too.

Please login or register to see this link.

The workaround has been mentioned before: "For now you can disable reports on lux change (parameter 40 = 0) or set it to value greater than 100 lux and turn on frequently automatic reports (parameter 42) -> for example once for 2 minutes. If you had parameter 40 on low value (<50) it won't affect your battery life (or impact will be minimal)."

[chaicka 195]

Probably I can share a little more on this as originator of the bug report.

 

I have since changed parameter 40 to 50 lux (previously 20 lux) and parameter 42 to 90 seconds (previously 900s) on all FGMS.

 

These changes have stopped the random 1-3 lux change reporting and may have contributed to reduction in the z-wave network choking/busy where I experienced delays in motion-based scenes quite frequently in the past.

 

Please login or register to see this attachment.

 

[kooy 5]

So any update on fixing this?

Its been reported since 2015-01-29 in bugtracker.

Its been a year now since.

 

Or can fibaro confirm this product as not work correctly and put a statement so i can get a refund