FIBARO DEAD MODULES

[sergiobaiao 5]

I'm having this very same issue for a long time with one of our customers. We installed a Fibaro system on an apartment, with 12 lighting modules (between dimmers and relays) along side with 6 roller blind modules. The module which is most distant from the HC2 is about 9~10m, all of them are installed on the living room of the apartment. Almost everyday, no matter what i change in the configuration of the HC2 (have tried pooling time, not marking modules dead, etc, etc etc) i get modules marked as dead. Sometimes they show as awake but all commands sent to the modules don't work.

Sometimes i get error messages on the web interface like:

[01] ID 24: Sending error

[01] ID 24: Sending error

My client is asking me to remove all hardware from his apartment, and i really don't know what to do. I've tried to solve this problem with Fibaro Support from my distributor here in Brazil (Munddo Distribuidora) but no matter what they try nothing is getting solved.

I really want some word from someone from Fibaro to see what i can do to solve this with my client. This client is a very influent person, which not only knows a lot of people, but also owns a magazine in my city, and if he decides to talk about this with people he knows, this may ruin my company.

There should be something that can be done to solve this, as well as to prevent it from happening again. Modules that are always powered on (not on battery) should never get disconnected. Fibaro must do something about this.

My company sell and install automation systems for 3 years. We work with other brands than Fibaro, and Fibaro is the only one giving us trouble. Automation is something that is supposed to ALWAYS work, seamlessly, flawlessly, so the client doesn't even know it's there. It's very hard to sell a product that gives problems almost every week.

[robmac 173]

see if you can borrow something like this?

Please login or register to see this link.

[elmips 4]

sergiobaiao, sending error mean you got problem with connection with that device. Try have a lower traffic in network, lower pooling, bigger wakeups

A.Socha, i don't understand why lower pooling time make lower zwave traffic ? please can you explain this point ?

I also had this kind of dead nodes issue (8 dead nodes per day, everyday different nodes) but i saw that if i set a bigger pooling time to my Horstmann ASR-ZW (pooling of 1200) i had now just 1 dead nodes per day.

but this setup mean that i don't know exactly when my boiler is "ON" or "OFF" in the event panel. It's not logging precisly...

if i set to 0 i have no logging of the boiler.

if i set to 30, i have a precise looging but the result is a lot of dead nodes in my system.

[robmac 173]

Simple congestion as each node can only handle so many packets a second. When many devices talk at the same time some nodes can get overloaded.

+ very low tolerance of HC2 to accept slow response. Possibly being able to set how long HC2 waits for response would be a good feature?

Polling regularly increases the chance.

Many associations set so devices receiving reports increases chance.

Devices that report to associated devices to often. e.g. I have a PID controlled coffee machine that I have to reduce the sensitivity of power recording on. Same would be true of rapidly changing temperature.

[A.Socha 133]

elmips, you was not understood me well

by lower pooling i mean bigger time .

1600 = low pooling

20 big pooling he spam every 20s

i hope now it is clear

robmac, in zwave they are strictly defined and we can't use own.

example

send command basic set to switch

sending 1 frame basic set

no response in time sending second frame

no response in time sending third frame

still no response sending frame by using routing

if you got 100 device that can try connect with that device trying route by all 99 nodes

if you in that time send other command it's waiting in queue (you see lags then) if that frame wait too many you can saw error sending error

[elmips 4]

A.Socha, yes thanks for your explaination, it's crystal clear now !

Please login or register to see this image.

/emoticons/default_icon_wink.gif" alt=";-)" />

yes Robmac, thanks. i understand, it's like a kind of broadcast storm compare to the ethernet world. what surprise me is that i dont have a huge installation and i start to have those node saturation... it feel like a bit weak.

Please login or register to see this image.

/emoticons/default_icon_curve.gif" alt=":-/" />

[Lauri 6]

If you replaced the HC2 did you added all the modules to the new HC2 or you used backup?

What happens if you set polling time to 0?

[sergiobaiao 5]

If you replaced the HC2 did you added all the modules to the new HC2 or you used backup?

What happens if you set polling time to 0?

I've added all the modules again, one by one, mainly because i wanted to make sure all of them were working. I haven't tried the polling set to 0.

[ Added: 2014-02-10, 18:49 ]

there is one very strange thing that i just noticed

all of a sudden, without me doing nothing, all the modules are saying:

Configuration: Device not upgraded

and some also give this:

Device kind: Dimmer

Device type: FGD211 Dimmer ANZ/BR

Producer: Unknown

Configuration: Device not upgraded

i didn't make any changes for the past week. the only time i saw this was when i hit the "reconfigure z-wave network" option and it failed for all devices.

what ta h**l is going on?

[anonymous66 9]

If you replaced the HC2 did you added all the modules to the new HC2 or you used backup?

What happens if you set polling time to 0?

I've added all the modules again, one by one, mainly because i wanted to make sure all of them were working. I haven't tried the polling set to 0.

[ Added: 2014-02-10, 18:49 ]

there is one very strange thing that i just noticed

all of a sudden, without me doing nothing, all the modules are saying:

Configuration: Device not upgraded

and some also give this:

Device kind: Dimmer

Device type: FGD211 Dimmer ANZ/BR

Producer: Unknown

Configuration: Device not upgraded

i didn't make any changes for the past week. the only time i saw this was when i hit the "reconfigure z-wave network" option and it failed for all devices.

what ta h**l is going on?

Did you try DSD37-ZWEU/US?

This repeater solved many of my disconnection problems.

[sergiobaiao 5]

this is the topology on the client

As you can see, the modules are very close to each other, i don't think there is a need for a repeater.

Also, here in Brazil we use a different frequency than US

Please login or register to see this attachment.

[Guest Kuuno]

which modules give errors in the picture?

are all frequencys the same? hc2 and every dimmer?

[sergiobaiao 5]

Kuuno,

The modules giving problems are really random, and yes, all of them are using the same freq.

[A.Socha 133]

sergiobaiao, You was trying rebuild mesh network?

[hvb83 1]

I basically have 3 modules that give me problems every now and then, 2 of which are Fibaro's. From the physical location I would say it shouldn't be a coverage problem, but just for the sake of ruling out as many possibilities as possible, I'll try installing one more module as a hop in between.

Still appears to me that rebuilding the mesh doesn't do anything. At least I've never really noticed that dead nodes occur less often after rebuilding the mesh a couple of times or using the modules extensively. Right now, using the poll dead nodes script it's bearable, but I understand that it's a totally different problem if the problem occurs on all your modules every now and then.

[sergiobaiao 5]

sergiobaiao, You was trying rebuild mesh network?

I didn't, but Fibaro support crew from Munddo (our local distributor) did. Today i went one more time to the client apartment and had to remove and add two modules again, because they lost connection, like, forever.

What's most interesting, and this was done under a Teamviewer session with Munddo's support, that when i removed the modules, they didn't get removed from the HC2 pages. I added the modules again, they got a new ID and i was able to control them, but i couldn't remove the old ones, I had to restart HC2 two times so i could remove the old modules configuration, because the HC2 kept trying to communicate with them

[A.Socha 133]

when you was removing you got log device id xx was deleted?

where xx was a d of that device? when you got 0 mean that device was not added to that hc2

[sergiobaiao 5]

when you was removing you got log device id xx was deleted?

where xx was a d of that device? when you got 0 mean that device was not added to that hc2

I got "0", but the modules were added to the HC2 previously. they just got lost in the wild

[A.Socha 133]

sergiobaiao, 0 mean that device was removed from hc2

[sergiobaiao 5]

sergiobaiao, 0 mean that device was removed from hc2

I know that. The module ID's that i removed were 22 and 25. After removing them, they were still present on the HC2 web interface, even tough i sucessfully removed them using removal procedure

[A.Socha 133]

sergiobaiao, You was not understood me

0 in that log was showing that modules was removed from hc2 before you removed it again

[sergiobaiao 5]

sergiobaiao, You was not understood me

0 in that log was showing that modules was removed from hc2 before you removed it again

only if HC2 did it by itself. I didn't. That's the problem i'm having constantly here. Modules get disconnected, never come back, even when not marked as dead they don't respond. When i remove the module using the remove procedure, i give the module id 0 removed message. When i add it again, it works ok for some time.