Why do my Fibaro Wall Plugs keep going dead?

[Ole 14]

I manage three houses (home, work, parents house) using Fibaro Home Center (both HCL and HC2) and for controlling the panelheaters I use a lot of wall plugs. Why on earth do they keep going dead? Every month between 1-3 of them go dead. I then have to force remove from Home Center, reset the wall plug and then ad it again. This very rearly happens with any other device, but with the wall plugs, very often. Anybody have an idea why that is? It is very frustrating.

[Stian1 0]

Hi, I'm experiencing the same issue. 

 

I have four FGWPE/F Wall Plug Gen5 , and 1-2 times a month they just disconnect and appear dead on the Z-Wave controller.

 

Most of the time I can just unplug them and plug them back in and they will reconnect and start working. 

On a few occasions I've had to reset them and reinclude them in the network.

 

Very annoying when it's connected to a heater and it's stuck in the ON position with no way to turn it off remotely. Just watching the temperature rise until I get home...

 

I have several other devices in my Z-Wave network, and I only have problems with the Fibaro Wall Plugs. 

 

I have the latest firmware according to Fibaro Home Center Lite. 

 

Was hoping this problem was fixed by a more recent firmware, but appears it's not.

 

I will also send a case to [email protected], and really hope this can be fixed by a firmware update.

 

 

[r0b 0]

Apparently there are different firmware versions available for different regions. It would be good to compare those amongst the people having issues.

 

@Stian1What are the serial numbers/region/firmware versions of your plugs?

 

Mine are:

00084606 EU, firmware 3.2
00123738 EU, firmware 3.2

[akatar 208]

Same thing here (well almost)

 

i have one wallplug, when i use this with a load and i turn the plug off, it goes on after a second. i can switch it off again but it will turn on again. this is only with a load.

then after a while the plug is "dead" on my system and i have to remove it and add it again.

 

any thoughts why?

[akatar 208]

I tested it with a zigbee plug, and the same thing happens. it must be something in the psu i connect. 

any ideas why?

[petergebruers 1,268]

20 hours ago, akatar said:

then after a while the plug is "dead" on my system and i have to remove it and add it again.

Thank you for joining this topic. That indeed sounds like the issue described here.

 

Next time this happens, can you please try this: unplug, plug back in and observe the LED ring. If it is RED then the Wall Plug has been factory reset and is no longer part of your Z-Wave network. 

 

@r0b already confirmed on OpenHAB forum, his plugs glow RED when they are dead.

 

20 hours ago, akatar said:

i have one wallplug, when i use this with a load and i turn the plug off, it goes on after a second. i can switch it off again but it will turn on again. this is only with a load.

Interesting, I think you are the first to mention that turning a load OFF turns it back ON. If it were the other way around, turning off after on, then I would say "overload detection". If there is no code doing that, I speculate that somehow the load causes the plug to see a button press... But your next post makes me think something else is going on...

 

16 hours ago, akatar said:

I tested it with a zigbee plug, and the same thing happens. it must be something in the psu i connect.

So your load is a switching mode power supply? Those can be rather "nasty" loads when they start up, causing high inrush current. I've checked the specs of a few open frame supplies used for LED strips and 60 A peak is not uncommon. But then I would expect the plug to turn off after on, not "on after off".

 

Do you have a spare FGD212? It would be interesting to see if your Dimmer 2 can turn on/off this load without issues.

 

Or maybe a FGS-2x3 with recent firmware, it also does switching at 0 crossing so should have less issues with that load.

 

We're not talking about an inductive load, like an iron core transformer? I did some tests with that and was not able to cause resets but it might be worth checking again, I was using the older Fibaro Wall Plug, not the ZW5 version (I don't have one here to test).

 

On 3/3/2020 at 3:50 PM, Stian1 said:

I have four FGWPE/F Wall Plug Gen5 , and 1-2 times a month they just disconnect and appear dead on the Z-Wave controller.

 

Most of the time I can just unplug them and plug them back in and they will reconnect and start working. 

On a few occasions I've had to reset them and reinclude them in the network.

 

Next time this happens, can you please try this: unplug, plug back in and observe the LED ring. If it is RED then the Wall Plug has been factory reset and is no longer part of your Z-Wave network.

 

On 3/3/2020 at 3:50 PM, Stian1 said:

Very annoying when it's connected to a heater and it's stuck in the ON position with no way to turn it off remotely. Just watching the temperature rise until I get home...

Mmmm yes Fibaro does not make "thermostat modules" but there are other brands so... But then those are not "plugs" sou you would have a nice DIY project to fit them in an enclosure...

 

Purely resistive heater or does it have a fan?

[akatar 208]

1 hour ago, petergebruers said:

Thank you for joining this topic. That indeed sounds like the issue described here.

 

Next time this happens, can you please try this: unplug, plug back in and observe the LED ring. If it is RED then the Wall Plug has been factory reset and is no longer part of your Z-Wave network. 

 

@r0b already confirmed on OpenHAB forum, his plugs glow RED when they are dead.

 

Interesting, I think you are the first to mention that turning a load OFF turns it back ON. If it were the other way around, turning off after on, then I would say "overload detection". If there is no code doing that, I speculate that somehow the load causes the plug to see a button press... But your next post makes me think something else is going on...

 

So your load is a switching mode power supply? Those can be rather "nasty" loads when they start up, causing high inrush current. I've checked the specs of a few open frame supplies used for LED strips and 60 A peak is not uncommon. But then I would expect the plug to turn off after on, not "on after off".

 

Do you have a spare FGD212? It would be interesting to see if your Dimmer 2 can turn on/off this load without issues.

 

Or maybe a FGS-2x3 with recent firmware, it also does switching at 0 crossing so should have less issues with that load.

 

 

 

I already tried that, the plug is glowing green so the plug has a gateway connected (no more then that, because if the controller is down it also is glowing green)

it is only red if the gateway is removed from the plug or has been erased for some kind of reason.

 

the psu is a 230 vac to 24vac convertor and the load on that is 25watt

 

i do not have a spare dimmer, but i do not think a dimmer would do anything good for a on/off psu. dimmers have often a slow start.

 

I orderd a new wall plug, let's see what brings that. i guess i have it on tuesday, give or take a day.

 

[petergebruers 1,268]

3 hours ago, akatar said:

I already tried that, the plug is glowing green so the plug has a gateway connected (no more then that, because if the controller is down it also is glowing green)

it is only red if the gateway is removed from the plug or has been erased for some kind of reason.

Okay, so next time it turns "dead" on your homeseer and does not come back to life when you unplug and plug back in, I expect it will glow "red". If it is dead and "green" then we have an interesting variant of this "Fibaro Wall Plugs keep going dead" issue...

 

3 hours ago, akatar said:

the psu is a 230 vac to 24vac convertor and the load on that is 25watt

They can be pretty nasty... Remember my topic about switch contacts welded by such a PSU?

 

3 hours ago, akatar said:

i do not have a spare dimmer, but i do not think a dimmer would do anything good for a on/off psu. dimmers have often a slow start.

You might think that but it is actually a good thing and they even sell such "slow start" devices to avoid inrush current problems. For example, if you had 10 of those PSU with high inrush current connected to a single switch (or relay) and turn them all at once (when the sine wave of your mains is at its peak voltage), that might trip your circuit breaker. I looked up such slow start devices, they are pretty expensive. So most solutions use an NTC like the "ESB1"

 

Please login or register to see this link.

 

Now as far as I know, sticking of the contacts aka "welding" is not an issue on the Fibaro Wall Plug (any version), but users have reported that the ZW5 version has very fast response to overcurrent so even though the maximum steady current may be below limit (13 A) it is sometimes not possible to turn on loads, because the load momentarily exceeds that limit at turn-on and then the ZW5 turns off the load again.

 

 

 

Edited March 8, 2020 by petergebruers

[akatar 208]

5 minutes ago, petergebruers said:

Okay, so next time it turns "dead" on your homeseer and does not come back to life when you unplug and plug back in, I expect it will glow "red". If it is dead and "green" then we have an interesting variant of this "Fibaro Wall Plugs keep going dead" issue...

 

 

 

 

No, i already did that, it's green then. so it has a "controller" stored in his memory

9 minutes ago, petergebruers said:

Okay, so next time it turns "dead" on your homeseer and does not come back to life when you unplug and plug back in, I expect it will glow "red". If it is dead and "green" then we have an interesting variant of this "Fibaro Wall Plugs keep going dead" issue...

 

They can be pretty nasty... Remember my topic about switch contacts welded by such a PSU?

 

Yes, but it doesn't explain why the plugs (zigbee and fibarpo wallplug) goes back on when you turn the switch off.

[petergebruers 1,268]

6 minutes ago, akatar said:

No, i already did that, it's green then. so it has a "controller" stored in his memory

Very interesting... And unfortunate because that means there is something else going on that makes these plugs disappear from the network.

 

Did you notice it reports very implausible values for Power or Energy? I mean eg extremely high values? @robmac and @r0b and myself have seen those and we're wondering if it is related to this issue (eg NVRAM corruption).

 

8 minutes ago, akatar said:

Yes, but it doesn't explain why the plugs (zigbee and fibarpo wallplug) goes back on when you turn the switch off.

No indeed and I have only one idea but it seems unlikely... But I'll say it anyway...

 

Do  you have an unusually high resistance (impedance) on that socket? Or does it happen anywhere in your house? I am speculating that the load causes the voltage to drop and maybe this causes the power supply of the Wall Plug to drop and cause a reset. Only problem with that theory is that I think the internal supply can hold enough energy and the inrush current is only very brief. On the other hand, I have seen devices reset when there is a very high "spike" near them (a PC that would crash when someone fired a professional flash nearby).

 

What happens if you have a high load but purely resistive, like a kettle or a heater?

[akatar 208]

27 minutes ago, petergebruers said:

Very interesting... And unfortunate because that means there is something else going on that makes these plugs disappear from the network.

 

Did you notice it reports very implausible values for Power or Energy? I mean eg extremely high values? @robmac and @r0b and myself have seen those and we're wondering if it is related to this issue (eg NVRAM corruption).

 

No indeed and I have only one idea but it seems unlikely... But I'll say it anyway...

 

Do  you have an unusually high resistance (impedance) on that socket? Or does it happen anywhere in your house? I am speculating that the load causes the voltage to drop and maybe this causes the power supply of the Wall Plug to drop and cause a reset. Only problem with that theory is that I think the internal supply can hold enough energy and the inrush current is only very brief. On the other hand, I have seen devices reset when there is a very high "spike" near them (a PC that would crash when someone fired a professional flash nearby).

 

What happens if you have a high load but purely resistive, like a kettle or a heater?

 

No. the reported value is 25 watt, ik tested it with a watercooker but then all is fine.

 

but i must say, the problem solved itself, the "thing" is a piece of artwork with lights and the wife decided to put it in the addict for storage.

so as long as it's stays there the problem is solved

 

now i have to look what i am going to do with the wallplugs i orderd.

 

something else, where can i buy the: ACC-UZB3-H-STA

i only find it in usa webshops

 

 

[petergebruers 1,268]

Thanks, @akatar - so it clearly has to do with the load. But problem "solved" indeed. Erm ACC-UZB3-H-STA - you want a "H" version of the UZB, that is with Asian filter... I don't know the answer to that... "mouser.be" lists seven available but 8 weeks shippings so sounds as if they are not in the EU.

Edited March 8, 2020 by petergebruers

[akatar 208]

@petergebruers i am going to buy a aeotec gen 5 stick to convert that one in a sniffer

 

[petergebruers 1,268]

1 minute ago, akatar said:

@petergebruers i am going to buy a aeotec gen 5 stick to convert that one in a sniffer

I wouldn't do that, the Aeotec has a antenna amplifier that is not supported by the Zniffer firmware and it won't work particularly well unless you solder a bit

 

I use this one instead of the UZB-3, so I know this one is OK...

 

Please login or register to see this link.

[Stian1 0]

 

@petergebruers

 

"Next time this happens, can you please try this: unplug, plug back in and observe the LED ring. If it is RED then the Wall Plug has been factory reset and is no longer part of your Z-Wave network."

It does not show red LED ring. In about 90% of the cases, which is the primary case I'm troubleshooting right now, the plug comes back on the network without reincluding after replugging. 

 

I have attached a log from OpenZwave from the last occurence of this problem:

 

Node015 is the wall plug. The start of the log show working communication, and the problem starts here:

2020-03-02 07:29:43.247 Info, Node015, WARNING: ZW_SEND_DATA failed. No ACK received - device may be asleep.

2020-03-02 07:29:43.247 Warning, Node015, WARNING: Device is not a sleeping node.

2020-03-02 07:29:45.470 Error, Node015, ERROR: Dropping command, expected response not received after 1 attempt(s)

 

 

 

"Mmmm yes Fibaro does not make "thermostat modules" but there are other brands so... But then those are not "plugs" sou you would have a nice DIY project to fit them in an enclosure...

Purely resistive heater or does it have a fan?"

It's this: 

Please login or register to see this link.

 

 

Please login or register to see this attachment.

[Stian1 0]

On 3/4/2020 at 1:06 PM, r0b said:

Apparently there are different firmware versions available for different regions. It would be good to compare those amongst the people having issues.

 

@Stian1What are the serial numbers/region/firmware versions of your plugs?

 

Mine are:

00084606 EU, firmware 3.2
00123738 EU, firmware 3.2

 

My firmware version is also 3.2, EU region.

[robmac 173]

On 3/8/2020 at 6:15 PM, petergebruers said:

ACC-UZB3-H-STA - you want a "H" version of the UZB, that is with Asian filter... I don't know the answer to that... "mouser.be" lists seven available but 8 weeks shippings so sounds as if they are not in the EU.

The ones shown in stock are available in US and fedex to europe. The 8 weeks is a restock time.

 

Not sure about transits to Asia.

 

This is my last from Mouser to UK.

 

UZB3 as my zniffer case is held together with tape again. I do treat them badly.

 

 

• Friday , 3/06/2020
  8:54 am

  XXXXXX GB

  Delivered
• 7:13 am

  CARDIFF GB

  On FedEx vehicle for delivery
• 4:43 am

  CARDIFF GB

  At local FedEx facility

• Thursday , 3/05/2020
  9:12 pm

  ROISSY CHARLES DE GAULLE CEDEX FR

  International shipment release - Import
• 8:04 pm

  RIDGMONT GB

  Departed FedEx location
• 5:26 am

  STANSTED GB

  In transit

  Package available for clearance
• 3:50 am

  ROISSY CHARLES DE GAULLE CEDEX FR

  Departed FedEx location
• 3:33 am

  ROISSY CHARLES DE GAULLE CEDEX FR

  In transit
• 3:32 am

  STANSTED GB

  At destination sort facility
• 3:23 am

  ROISSY CHARLES DE GAULLE CEDEX FR

  In transit
• 12:06 am

  ROISSY CHARLES DE GAULLE CEDEX FR

  In transit

  Package available for clearance

• Wednesday , 3/04/2020
  11:00 pm

  ROISSY CHARLES DE GAULLE CEDEX FR

  At local FedEx facility
• 6:54 pm

  ROISSY CHARLES DE GAULLE CEDEX FR

  Arrived at FedEx location
• 4:33 am

  MEMPHIS, TN

  In transit
• 4:08 am

  MEMPHIS, TN

  Departed FedEx location

• Tuesday , 3/03/2020
  11:10 pm

  MEMPHIS, TN

  Arrived at FedEx location
• 9:27 pm

   

  Shipment information sent to FedEx
• 8:30 pm

  FORT WORTH, TX

  Departed FedEx location
• 5:40 pm

  MANSFIELD, TX

  Left FedEx origin facility
• 1:35 pm

  GRAND PRAIRIE, TX

  Picked up

[robmac 173]

On 3/8/2020 at 11:48 AM, petergebruers said:

So your load is a switching mode power supply? Those can be rather "nasty" loads when they start up, causing high inrush current. I've checked the specs of a few open frame supplies used for LED strips and 60 A peak is not uncommon. But then I would expect the plug to turn off after on, not "on after off".

Possibly Fibaro could tell us the spec.

 

It is given for the G version that I have but that may be different hardware. This is great information. note inrush and electronic ballast specs. 3A is not a lot. Some computer power supplies are more than that.   

 

Well done but could you post the values for older devices please.

 

 

Please login or register to see this attachment.

[petergebruers 1,268]

Cross referencing this post because it might be related somehow (inadvertent change of parameter 1, "always on").

 

 

[Giovanni_Genoa 11]

@petergebruers

I have 6 FGWP.102, SW ver.3.2.

 

One managing power supply to home water pump – 255 W – never Failed

One is managing power to Light 10.5 W – LED Lamp powered via 220 V – never Failed

One is managing power to modem 6.5 W – via AC/DC adapter – Never Failed

One is Used only for Christmas Three.

Two are manging very low power pieces for my small aquarium, about 5W each.

 

They are plugged into two sockets of the UPS, and manage power:

-          the first one to led light through power adapter, also this is managed with HC2 Scene

-          The second one to a small water pump 220 V to simulate waves, time on/off are managed with HC2 scene.

 

 

Randomly this last had Communication problem: I inverted the use with the Led… no chance!  After a while the problem returns. Therefore I returned everything as it was in origin.

 

Further the led ring glows green as soon as plugged in but it does not show any light when power is ON.

I need to pull out from socket and plug in again! But still the led ring is not showing any colour when power is on.

 

And more….

Recently I was on holidays and from App on my phone I noted again the same problem, But I had no possibility to pull/push the “nice” piece in the socket). But it was “funny” that the problem was extend also to other 14 devices in range…. All of them never had a problem since installation! (Sincerely I did never checked before for this last extension of the problem!)

 

Once at home I downloaded (

Please login or register to see this link.

) the neighborslist of all devices having communication problem. All of them had a common device connected: the FGWP102 ID 176 and this was showing also ID1 (HC2).

 

I tried to refresh the Mesh, but unsuccessfully, they were always communicating throw the failed FGWP  and there was no response from them to HC2. Therefore, no chance to renew the Mesh.

It seems that the various devices were communicating with the FGWP but the FGWP was not communicating with the HC2. OK, I removed the FGWP and suddenly all other devices were OK, they used other links of the mesh to communicate with HC2.

 

This is confirmation that devices were communicating with the Failed FGWP (for them communication was ok!) but the FGWP was not communicating with HC2 (ID1), even though ID1 was in the neighborlist fo the FGWP.

Having FGWP Failed, I renewed the mesh successfully, having always the Failed FGWP un-powered.

I plugged in again the FGWP and it was working again.

 

24 hours later again, but no problem with other devices

 

In conclusion….

-          It’s not a problem of interference with magnetic field of UPS (I have another FWGP connected to the same UPS and a third one with another UPS – No problem).

-          It’s not a problem of power load the second FGWP is working normally with 3.5 W

-          It’s not a problem of power adapters (the connected water pump, even if small is using 220V - 5W)

-          It is not a problem of type of use, (I have another E-Motor,  one with Led lamp (220W/10,5W), one LEd with poeer adapter, and a Power adapter for Modem).

Only one doubt: I purchase them in various step, this is that last one that I Purchased (anyhow already out of guarantee), considering what other users claimed (there are also other various topic with FGWP), probably it could be a problem affecting a batch(s) of products. (Fibaro should Investigate, their QM would be Happy).

 

Next step:

Reset to factory ….and ….In God I Hope.