I checked a little bit more with the new stick. Original flashed with firmware by the shop.
Still the same issue.
Some things i figured out:
No interface is set for the devices of Zigbee. @Micha@Sim or is this normal for zigbee?
I tried to active it wit the “setInterface” funktion. But no change
Yes there are some packets we lose at the begining while we boot. But then we have no more lost packeges in the logs. Is this stopped after a couple of tries.
Is there any possibilities, that it is issued somehow because of my enocean usb-stick.
You could perhaps send me on private your zigbee configuration file. You can change of course the sensitive information, like the password and perhaps PAN id. This way I could check it out.
Please ensure you don’t have some other module accessing the stick. Odd things could happen if you have it accessed like that.
Please ensure you actually have the proper serial device set in the config file. You could have some other device pointed at instead and in such case you get a behavior similar with what you are experiencing:
I just obtained the above by setting the serial interface for the zigbee module to a zwave stick. Of course it doesn’t reply to the TI zigbee ‘ping’.
For raspberry, it probably should be the one you have in the config file.
Maybe your should change it to /dev/ttyACM0
as this is shorter and less prone to errors.
You may also want to set a PAN id, don’t let it be random. But this doesn’t solve your current situation.
I’m out of ideas right now, I’ll check again tomorrow the zigbee2mqtt sources to see if they have something special for starting the stick, but I don’t expect to find anything new
Veto
The serial-by-id approach should make sure, that no other USB device was mistakenly taken the defined /dev/...
I gave that hint cause I did not know how many devices were installed…
I just committed new sources. Binaries should be available soon.
I just increased the timeout for the first packet sent at stick initialization and I added some attempts to retry, including a soft reset of the stick if ‘ping’ fails.
There is a remote chance of fixing your issue, but it’s worth a try.
I also looked over zigbee.mqtt sources and some python ports that claim to work with that firmware, I couldn’t spot anything obvious there…
Probably it would be much easier if I would have such a stick to test it
I wonder why ZIGBEE 3.0.x is mandatory with latest firmware (correct me if I’m wrong). All guys using cc2530/31-chips should run into problems because the chips are not powerful enough. Please follow this issue here: https://github.com/Koenkk/zigbee2mqtt/issues/1445
Is there anybody using cc2530/31 and ZIGBEE 3.0.x successfully?