Fibaro FGS213 Switch issues

Hardware
1

I’ve been trying to troubleshoot why this relay has been causing me issues. Just general issues like not turning on with the motion sensor, or not updating the interface when it turns on or off. Looking int he logs, I see this when I turn it off using HA.

2019-10-19 13:50:23.293 Info, Node058, Value::Set - COMMAND_CLASS_SWITCH_BINARY - Switch - 0 - 1 - False
2019-10-19 13:50:23.293 Info, Node058, SwitchBinary::Set - Setting node 58 to Off
2019-10-19 13:50:23.294 Detail, Node058, Queuing (Send) SwitchBinaryCmd_Set (Node=58): 0x01, 0x0a, 0x00, 0x13, 0x3a, 0x03, 0x25, 0x01, 0x00, 0x25, 0xd5, 0x0b
2019-10-19 13:50:23.294 Detail, Node058, Queuing (Send) SwitchBinaryCmd_Get (Node=58): 0x01, 0x09, 0x00, 0x13, 0x3a, 0x02, 0x25, 0x02, 0x25, 0xd6, 0x09
2019-10-19 13:50:23.294 Info, Node058, Sending (Send) message (Callback ID=0xd5, Expected Reply=0x13) - SwitchBinaryCmd_Set (Node=58): 0x01, 0x0a, 0x00, 0x13, 0x3a, 0x03, 0x25, 0x01, 0x00, 0x25, 0xd5, 0x0b
2019-10-19 13:50:23.899 Detail, Node058,   Received: 0x01, 0x04, 0x01, 0x13, 0x01, 0xe8
2019-10-19 13:50:23.900 Detail, Node058,   ZW_SEND_DATA delivered to Z-Wave stack
2019-10-19 13:50:27.911 Detail, Node058,   Received: 0x01, 0x07, 0x00, 0x13, 0xd5, 0x00, 0x01, 0xce, 0xf1
2019-10-19 13:50:27.911 Detail, Node058,   ZW_SEND_DATA Request with callback ID 0xd5 received (expected 0xd5)
2019-10-19 13:50:27.911 Info, Node058, Request RTT 4616 Average Request RTT 4441
2019-10-19 13:50:27.912 Detail, Node058, Removing current message
2019-10-19 13:50:27.912 Info, Node058, Sending (Send) message (Callback ID=0xd6, Expected Reply=0x04) - SwitchBinaryCmd_Get (Node=58): 0x01, 0x09, 0x00, 0x13, 0x3a, 0x02, 0x25, 0x02, 0x25, 0xd6, 0x09
2019-10-19 13:50:28.518 Detail, Node058,   Received: 0x01, 0x04, 0x01, 0x13, 0x01, 0xe8
2019-10-19 13:50:28.519 Detail, Node058,   ZW_SEND_DATA delivered to Z-Wave stack
2019-10-19 13:50:32.312 Detail, Node058,   Received: 0x01, 0x0d, 0x00, 0x04, 0x10, 0x3a, 0x07, 0x60, 0x0d, 0x01, 0x01, 0x25, 0x03, 0x00, 0x90
2019-10-19 13:50:32.312 Info, Node058, Response RTT 4400 Average Response RTT 5014
2019-10-19 13:50:32.312 Error, Node058, Cannot find endpoint map to instance for Command Class COMMAND_CLASS_SWITCH_BINARY endpoint 1
2019-10-19 13:50:32.376 Detail, Node058,   Received: 0x01, 0x0d, 0x00, 0x04, 0x10, 0x3a, 0x07, 0x60, 0x0d, 0x01, 0x01, 0x25, 0x03, 0x00, 0x90
2019-10-19 13:50:32.376 Info, Node058, Response RTT 4464 Average Response RTT 4739
2019-10-19 13:50:32.376 Error, Node058, Cannot find endpoint map to instance for Command Class COMMAND_CLASS_SWITCH_BINARY endpoint 1
2019-10-19 13:50:32.773 Detail, Node058,   Received: 0x01, 0x07, 0x00, 0x13, 0xd6, 0x00, 0x01, 0xe6, 0xda
2019-10-19 13:50:32.774 Detail, Node058,   ZW_SEND_DATA Request with callback ID 0xd6 received (expected 0xd6)
2019-10-19 13:50:32.774 Info, Node058, Request RTT 4862 Average Request RTT 4651
2019-10-19 13:50:32.866 Detail, Node058,   Received: 0x01, 0x12, 0x00, 0x04, 0x00, 0x3a, 0x0c, 0x60, 0x0d, 0x01, 0x01, 0x32, 0x02, 0x21, 0x32, 0x00, 0x00, 0x00, 0x00, 0x91
2019-10-19 13:50:32.867 Info, Node058, Response RTT 4955 Average Response RTT 4847
2019-10-19 13:50:32.867 Error, Node058, Cannot find endpoint map to instance for Command Class COMMAND_CLASS_METER endpoint 1
2019-10-19 13:50:37.912 Error, Node058, ERROR: Dropping command, expected response not received after 1 attempt(s)
2019-10-19 13:50:37.912 Detail, Node058, Removing current message
2019-10-19 13:50:37.913 Detail, Node058, Notification: Notification - TimeOut

So it seems there’s a problem with COMMAND_CLASS_METER in OpenZwave, and it’s causing the packet to be dropped as invalid. So the interface says it’s still on when the light has turned off.

This device also has 4800ms response time. I’m not sure if that error is a cause or a symptom. My mesh network is robust, and an extender is installed about 5 feet away. I have excluded it and repaired it in insecure mode to help latency, and it didn’t help.

2

Hi @cexshun,

I have an HA instance running with various Fibaro running fine.

I’m wondering if you have tried to either move the switch closer to your zwave antenna or tried to completely removed and re-add the device?

3

Yes to both

4

Hmm I have tested a few of my Fibaro Dimmers - no issues in the logs and they are working as expected. I’m on HA 0.99.3. Though I’m planning to upgrade later this weekend, to latest release.

5

Why is your extender 5 feet away ?
That doesn’t help anything.
How far from your z-stick to your 213 switch ?
Try removing the extender as your 213 switch is an extender in its own right.
Are there any other z-wave devices (not battery) close (between hub and 213 sw) even if at 90° ?
Not sure you can add a 213 as a secure node (though the manual implies you can) (unlike the dimmer 2 where you have to, if you want the scene functionality)

Edit: Please remember to ‘heal’ your network after each change you want to test. AND then leave it whilst it does the healing. Depends on the size of your network but I allow 10 to 15 mins

6

Why is your extender 5 feet away ?

Because it’s the farthest away from my hub, on a different floor and through several walls. I put the extender in the middle of the floor in the wide open area to obtain full area coverage.

How far from your z-stick to your 213 switch ?

As the crow flies, maybe 40 feet. But it is through a wood floor and possible has to propagate through a few walls. Zwave map shows the device as 1 hop away before reaching the z stick.

Try removing the extender as your 213 switch is an extender in its own right.

The extender was recently (day I posted this) added to try and improve reliability in that area of the house. It was problematic before then.

Are there any other z-wave devices (not battery) close (between hub and 213 sw) even if at 90° ?

Several. With 30 zwave nodes, at least half not on battery, there are several in the path. The node lists 15 neighbors, 9 being AC powered.

Please remember to ‘heal’ your network after each change you want to test.

Yep, heal after every new device, and allow the zwave map to propagate.

At this moment, the response time is down to 298ms, which is completely acceptable. But I’m still getting the endpoint errors which is causing dropped packets and unreliable performance.

7

Everything is pointing to an OpenZwave issue on this switch. And since we’re still locked on the ancient OZW using locked configuration files, I’m not seeing much hope in it working. I replaced it with an Aeotec Nano I had lying around. Not only does it work perfectly, but it also has a hugely superior radio. Instead of being a hop away, it’s connected directly to the stick. Response time of 40ms.