O.k. thanks for the explanation. I’ve been digging for some more information about these plugs: But all documentation/tutorials I’ve found are referring to touchlink pairing - even when connecting to a HUE bridge.
More than this, there seems to be no way to do a real factory reset of it - only option is an unpair from a remote using touchlink as well
Anywway - seems I have to buy something new for it
And OH will crash hard when the binding is stopped (let me know if you’re interested in the core dump):
get_java_var: invalid file descriptor
#
# A fatal error has been detected by the Java Runtime Environment:
#
# SIGSEGV (0xb) at pc=0x00007ff7eb8f47d5, pid=2667, tid=0x00007ff7e530a700
#
# JRE version: OpenJDK Runtime Environment (8.0_131-b12) (build 1.8.0_131-b12)
# Java VM: OpenJDK 64-Bit Server VM (25.131-b12 mixed mode linux-amd64 compressed oops)
# Problematic frame:
# C [libNRJavaSerial.so+0x77d5] read_byte_array+0x55
#
# Core dump written. Default location: /opt/openhab2/userdata/core or core.2667
#
# An error report file with more information is saved as:
# /opt/openhab2/userdata/hs_err_pid2667.log
Compiled method (c2) 1696268 14685 s! 4 gnu.io.RXTXPort$SerialInputStream::read (98 bytes)
total in heap [0x00007ff83b1f2c10,0x00007ff83b1f3b38] = 3880
relocation [0x00007ff83b1f2d38,0x00007ff83b1f2dd8] = 160
main code [0x00007ff83b1f2de0,0x00007ff83b1f33c0] = 1504
stub code [0x00007ff83b1f33c0,0x00007ff83b1f3420] = 96
oops [0x00007ff83b1f3420,0x00007ff83b1f3428] = 8
metadata [0x00007ff83b1f3428,0x00007ff83b1f3498] = 112
scopes data [0x00007ff83b1f3498,0x00007ff83b1f38b0] = 1048
scopes pcs [0x00007ff83b1f38b0,0x00007ff83b1f3a60] = 432
dependencies [0x00007ff83b1f3a60,0x00007ff83b1f3a70] = 16
handler table [0x00007ff83b1f3a70,0x00007ff83b1f3ae8] = 120
nul chk table [0x00007ff83b1f3ae8,0x00007ff83b1f3b38] = 80
On a positive note, I transferred primary controller from a zstick to the zwave side (using OZW and Zensys), then transferred SIS/SUC, and everything seems good after changing the associations to the new controller!
Yes! There’s also a model (HUSBZ-1) with just the zigbee for nearly the same price. IIRC, the zwave is zw500 with v4.05 firmware. So very similar to zstick but newer firmware. And cost me less than a backup zstick! I will likely post a separate thread about it… maybe after zigbee is working on it. IMO, it would be a perfect off the shelf unit. I have a lot of Linear devices and they’ve been very solid for me.
I’ll play with the logging to see what I’m missing.
Hi @chris,
would you have a look at the color channel again? The “brightness” control do not act as expected:
When moving the brightness slider in Paper UI, the saturation of the bulb is set to the value selected by the brightness slider - BUT: the brightness value has been set internally as well and will become effective through some subsequent command (e.g. color, saturation)… hard to explain
Sorry for not reporting earlier. I haven’t noticed before since I was more keen on the white spectrum control and so did just basic testing on the color channel which is working fine when leaving the brightness slider untouched.
I have tried everything I can think of and can’t figure out why I’m not getting all of the logging. But I’m also not seeing any more details in other peoples logs, so I’m not sure exactly what is missing from them. I’ve only seen people reporting use of the TI controllers. Does the code have a difference in the logging between the Ember and cc2531? Has anyone else gotten an Ember device working?
Well, the logging is obviously different, but they are both under the same com.zsmartsystems.zigbee namespace. Enabling this works for me. I do see some basic stuff from the ember driver, but I’d have expected more as I don’t see any low level frames being sent.
Yes - me. It’s being used commercially by a customer, so I think it works ;). Maybe the dongle you have isn’t quite compatible with the NCP code (I did say I wasn’t sure it would be)…
I’am using an CC2531EMK Coordinator which is already programmed to my Zigbee network. (PAN ID, channel and network key is already set) I’ve added the coordinator as a thing and changed all the values exactly as they are programmed on the stick. Except the “extended PAN ID”, which I don’t now. (I tried to find it out using Ubiqua packet analyzer, but my network doesn’t seam to have such an ID)
But the coordinator cannot be initialized with those values. Here are the logs regarding the problem.
2017-07-03 23:03:18.611 [INFO ] [ndler.ZigBeeCoordinatorCC2531Handler] - Serial port [/dev/ttyACM1] is initialized.
2017-07-03 23:03:23.616 [WARN ] [etwork.impl.ZigBeeNetworkManagerImpl] - Dongle reset failed. Assuming bootloader is running and sending magic byte 0x01.
2017-07-03 23:03:28.618 [WARN ] [etwork.impl.ZigBeeNetworkManagerImpl] - Attempt to get out from bootloader failed.
2017-07-03 23:03:28.629 [INFO ] [ndler.ZigBeeCoordinatorCC2531Handler] - Serial port [/dev/ttyACM1] is closed.
Is there a way to configure the binding to accept the already programmed values?
My guess is the issue isn’t related to the EPAN ID - if you don’t reinitialise the network, then it should use the existing settings. I’d suggest to enable more debugging by enabling the com.zsmartsystems.zigbee namespace.
Hi Chris, I’ve enabled the debug log for org.openhab.binding.zigbee and com.zsmartsystems.zigbee. Here is the output. I’ve removed the PAN ID’s an the network key, but everything is set. The EPANID was generated automatically.
What could cause this problem? I also tried the console app of zsmartsystems on my Mac with the same adapter an there I was able to initialize the coordinator and I was able to see the network packets. Could this maybe be a problem that the device does not work properly on my Rasberry Pi? Maybe a driver problem?
I see the binding is sending the magic byte - I’m not sure that this is correct though so this might be a problem. I would suggest to make sure that you plug in the device, leave it for 1 minute, then start the binding and see if that helps. In theory the stick should be out of the bootloader by then so an incorrect magic byte shouldn’t matter…
If it’s not that, then it might be issues with the serial driver. I’ve seen some problems with the communications to the TI sticks when running in OH versus running with the debug console. The errors I’ve seen are low level COM port issues so I have a theory (rightly or wrongly!) that it might be related to the drivers which are different in the console.