Supporting the ACT Solutions ZCU201 USB Z-wave Interface

[1audio]

The US ZWave and the EU ZWave run on different frequencies. They will not communicate. I will have access to one of the Homepro USB modules later this week. I will need to assemble a seperate machine to test it. Changing interfaces will screw up my system and with 20+ devices its a pain to reconstruct everything.
The ZWave stuff is not expensive but it does cost between $5-$10 us per device still. The rest of the cost is in dimming circuitry and reliability. The Leviton and Cooper stuff is much better made than the Homepro, and it shows. But the Homepro still works.

If you find a ZWave device that doesn't meet the minimum compatibility and has a ZWave logo complain loudly to the Zen-sys guys. They are supposed to be pretty aggressive on this. I know the current Harmony devices with ZWave are not compliant and they don't have logos any more.

There is software from both Homeseer and Controlthink that will do some tricks for backing up ZWave systems and restoring them. They should work with EU ZWave stuff. I like the Homeseer solution. You don't need a portable controller, just the software and your USB device. You set up your system with the USB and a laptop. Then transfer the stick to LMCE and it imports the devices from the stick. If you named the devices with the homeseer app it should be possible with a revised ZWave stack for LMCE to import those names and "autoconfigure" most of the setup.

[caiman]

On the same note, since it is possible to send some custom commands with LMCE, is there any way to get info - even manually or the hard way - on the state of the network, routes in place etc ?

Homeseer has a great software that allows you to poll that information, but because it will work only with their own USB controller which uses the US frequencies, no luck to get it here in Europe.

If that's not possible at all with LMCE, is there any chance to have some functionality added to the driver of the ZCU010 ?

If not, are there any other troubleshooting tools or network mapping tools available in europe ?

many questions here, but my z-wave network doesn't work well, and I'd like to dive into it to fix it

Caiman

[hkmod25]

Merten Configuration software is very good and give a nice possibilities to configure and troublesh. your network.
It is also possible to transfer your upset when work is done to HomeSeer.
But:
1. You need Merten Conf. Tool HARDWARE to use this software.
2. New wall (dimmer/appliance) modules from ACT recognises only with basic functionality.
3. New wall (transmitter) modules from ACT not recognises at all.
4. Merten still not make available Configuration Files Creation Software.

[1audio]

The Homeseer setup software Z-Tool on their site ftp://homeseer.com/pub/setupztool1_0_0_0.exe works with a number of ZWave modules and is independent of frequency.

The issues about features comes from two aspects of ZWave. first there are standard features, something that increases with each new rev. Those unfortunately can be retrofitted into modules. All ZWave approved devices need to support those features (as of manufacture approval). Second are proprietary features that are unique to a vendor. Some of the things mentioned about Merten are proprietary.

The latest generation of controllers have a system rebuild function that tests the linking between modules and recreates the routing tables. Its not a true mesh like the Internet. It has fixed routing options. And sometimes they aren't ideal(!) The update tests all of the links and recreates the routing. the routing is only stored in controllers. And the user or manufacturer doesn't have access to muck with the algorithms, those are locked down by Zen-Sys in the chip libraries.

Updating the LMCE Zwave module could add a lot of functions and make it better. Better (newer) controllers also help.

[totallymaxed]

1Audio,

So the ZWave-computer interface (e.g. the ZCS201) has the configuration stored? That would explain why lmce knew I have three devices after a re-install.

All SUC's like the ZCS201/ZCU201 store the config you send from the Master Controller... you can move the SUC to another LinuxMCE system and the new Core will interrogate it and sense the existing z-wave config and ask if you want to use it or create a new one.

[totallymaxed]

Hari... I am really sorry! I missed your post above... I have been really busy for last several days. So big apology to you.

no problem! Hope your business is going fine

I am away from the office until late Friday... but I will try to test your code then.

great!

thank you very much and best regards,
Hari

Hari i got back very late from my trip so i did not have any time to test the drive... I will look at it on Monday.

[totallymaxed]

Andrew,
Do you have any experience with Merten/LMCE implementations, in particular how their software, touch panels and their central units?
Their stuff is real nice for retrofitting older houses with home automation, which is my situation. Prices are a bit steep, but if I include wiring etc with the purchase of cheaper stuff, it doesn't look just as bad...
Anyway. I'm not even considering it if it doesn't work properly with LMCE, so I'm real curious what you think.
Mark

Yes we do have some experience with Merten's Connect system... and we like it very much. However its early days in terms of the technical side and there also maybe some restrictions on what we can make openly available here. Its too early to say yet on this however.

[cyr]

I did some munching to the spcp8x5 code and managed to compile and load the module on 0704. If somebody feels like sending me a ZCU201...

Thanks for the patch, I got the module to compile for my kernel (2.6.21.5) and it loaded.

However, when I loaded the module with my ZCU201 plugged in, this is what I got in my kernel log:

Code: [Select]

Jan 24 18:17:55 kira kernel: usbcore: registered new interface driver usbserial
Jan 24 18:17:55 kira kernel: drivers/usb/serial/usb-serial.c: USB Serial support registered for generic
Jan 24 18:17:55 kira kernel: usbcore: registered new interface driver usbserial_generic
Jan 24 18:17:55 kira kernel: drivers/usb/serial/usb-serial.c: USB Serial Driver core
Jan 24 18:17:55 kira kernel: drivers/usb/serial/usb-serial.c: USB Serial support registered for <NULL>
Jan 24 18:17:55 kira kernel: spcp8x5 2-2:1.0: <NULL> converter detected
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB0
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB1
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB2
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB3
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB4
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB5
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB6
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB7
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB8
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB9
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB10
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB11
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB12
Jan 24 18:17:55 kira kernel: usb 2-2: <NULL> converter now attached to ttyUSB13
Jan 24 18:17:55 kira kernel: usbcore: registered new interface driver spcp8x5
Jan 24 18:17:55 kira kernel: /usr/src/spcp8x5/spcp8x5.c: SPCP8x5 USB to serial adaptor driver v0.01

I didn't expect the driver to register 14 devices

I'm not much of a kernel hacker, so I have real idea how to debug this...

Worse things happened when I tried actually opening the device (ttyUSB0) in my test program...

Code: [Select]

Jan 24 20:15:10 kira kernel: BUG: unable to handle kernel NULL pointer dereference at virtual address 00000000
Jan 24 20:15:10 kira kernel:  printing eip:
Jan 24 20:15:10 kira kernel: cf988866
Jan 24 20:15:10 kira kernel: *pde = 00000000
Jan 24 20:15:10 kira kernel: Oops: 0000 [#1]
Jan 24 20:15:10 kira kernel: Modules linked in: spcp8x5(F) usbserial(F) rfcomm(F) l2cap(F) wlan_tkip wlan_xauth ipt_MAS$
Jan 24 20:15:10 kira kernel: CPU:    0
Jan 24 20:15:10 kira kernel: EIP:    0060:[<cf988866>]    Tainted: PF      VLI
Jan 24 20:15:10 kira kernel: EFLAGS: 00010086   (2.6.21.5 #4)
Jan 24 20:15:10 kira kernel: EIP is at spcp8x5_close+0x56/0x190 [spcp8x5]
Jan 24 20:15:10 kira kernel: eax: c1ce6ad0   ebx: 00000246   ecx: c9dede80   edx: c9dede74
Jan 24 20:15:10 kira kernel: esi: 00000000   edi: 00000000   ebp: ca9b6a00   esp: c9dede60
Jan 24 20:15:10 kira kernel: ds: 007b   es: 007b   fs: 00d8  gs: 0033  ss: 0068
Jan 24 20:15:10 kira kernel: Process zw_serial (pid: 7459, ti=c9dec000 task=c1ce6ad0 task.ti=c9dec000)
Jan 24 20:15:10 kira kernel: Stack: c0102ab3 ca9b6a00 cf98a5c0 ca9b6a08 c2280ca0 00000000 c1ce6ad0 c010e870
Jan 24 20:15:10 kira kernel:        c653911c c653911c ca9b6a00 c2280ca0 c6539000 ca9b6a08 cf9aa343 cd8027f8
Jan 24 20:15:10 kira kernel:        c01d880f cf9aff40 c6f954e0 fffffe00 c6539000 0bc00000 c6f954e0 00000100
Jan 24 20:15:10 kira kernel: Call Trace:
Jan 24 20:15:10 kira kernel:  [<c0102ab3>] common_interrupt+0x23/0x30
Jan 24 20:15:10 kira kernel:  [<c010e870>] default_wake_function+0x0/0x10
Jan 24 20:15:10 kira kernel:  [<cf9aa343>] serial_open+0x103/0x160 [usbserial]
Jan 24 20:15:10 kira kernel:  [<c01d880f>] kobject_get+0xf/0x20
Jan 24 20:15:10 kira kernel:  [<c01f58fd>] tty_open+0x14d/0x2c0
Jan 24 20:15:10 kira kernel:  [<c0152c60>] chrdev_open+0x70/0x130
Jan 24 20:15:10 kira kernel:  [<c0152bf0>] chrdev_open+0x0/0x130
Jan 24 20:15:10 kira kernel:  [<c014ed7f>] __dentry_open+0xaf/0x1c0
Jan 24 20:15:10 kira kernel:  [<c014ef45>] nameidata_to_filp+0x35/0x40
Jan 24 20:15:10 kira kernel:  [<c014ef9b>] do_filp_open+0x4b/0x60
Jan 24 20:15:10 kira kernel:  [<c014effa>] do_sys_open+0x4a/0xe0
Jan 24 20:15:10 kira kernel:  [<c014f0cc>] sys_open+0x1c/0x20
Jan 24 20:15:10 kira kernel:  [<c01027c6>] sysenter_past_esp+0x5f/0x89
Jan 24 20:15:10 kira kernel:  =======================
Jan 24 20:15:10 kira kernel: Code: 18 b8 70 e8 10 c0 89 54 24 14 8d 54 24 14 89 44 24 1c 8b 45 04 05 1c 01 00 00 e8 36 $
Jan 24 20:15:10 kira kernel: EIP: [<cf988866>] spcp8x5_close+0x56/0x190 [spcp8x5] SS:ESP 0068:c9dede60
Jan 24 20:16:14 kira kernel: usb 2-2: USB disconnect, address 2
Jan 24 20:16:14 kira kernel: BUG: unable to handle kernel paging request at virtual address 0100012c
Jan 24 20:16:14 kira kernel:  printing eip:
Jan 24 20:16:14 kira kernel: c011f1d3
Jan 24 20:16:14 kira kernel: *pde = 00000000
Jan 24 20:16:14 kira kernel: Oops: 0002 [#2]
Jan 24 20:16:14 kira kernel: Modules linked in: spcp8x5(F) usbserial(F) rfcomm(F) l2cap(F) wlan_tkip wlan_xauth ipt_MAS$
Jan 24 20:16:14 kira kernel: CPU:    0
Jan 24 20:16:14 kira kernel: EIP:    0060:[<c011f1d3>]    Tainted: PF      VLI
Jan 24 20:16:14 kira kernel: EFLAGS: 00010202   (2.6.21.5 #4)
Jan 24 20:16:14 kira kernel: EIP is at queue_work+0x3/0x30
Jan 24 20:16:14 kira kernel: eax: cefee8c0   ebx: cefee8c0   ecx: 00000000   edx: 0100012c
Jan 24 20:16:14 kira kernel: esi: 00000008   edi: c2280ca0   ebp: c475ac18   esp: ce049ea4
Jan 24 20:16:14 kira kernel: ds: 007b   es: 007b   fs: 00d8  gs: 0000  ss: 0068
Jan 24 20:16:14 kira kernel: Process khubd (pid: 512, ti=ce048000 task=c1225af0 task.ti=ce048000)
Jan 24 20:16:14 kira kernel: Stack: c3646c01 cf9a99d4 c3647b68 ca8a69a0 cf852543 c475ac00 c475ac18 cf98a4e0

etc. etc.

I had to reboot the machine.

[hari]

I had to reboot the machine.

ok, i'll look into that.. i hoped to come by with those little mods, but i have to go through the whole code.

I'll provide a new module soon.

best regards,
Hari

[cyr]

Great, I'll be happy to test it (although not on my router/asterisk machine this time I think  ).

[hari]

Great, I'll be happy to test it (although not on my router/asterisk machine this time I think  ).

sorry, it was meant to be a test i don't have the device myself, so i can't try..

i've made some changes and it should now register all usb stuff correctly (you should not get plenty of devices any more). But be aware, i still have 3 warnings left (initialization from incompatible pointer type).

I'm looking at that at the moment but if you are not faint hearted you may try that in between (v0.02):
[link removed - crashes]

best regards,
Hari

[hari]

Great, I'll be happy to test it (although not on my router/asterisk machine this time I think  ).

sorry, it was meant to be a test i don't have the device myself, so i can't try..

i've made some changes and it should now register all usb stuff correctly (you should not get plenty of devices any more). But be aware, i still have 3 warnings left (initialization from incompatible pointer type).

I'm looking at that at the moment but if you are not faint hearted you may try that in between (v0.02):
[link removed - crashes]



best regards,
Hari

ok, wait another 15 minutes, i found the problem and am fixing it.

[hari]

ok, lets try that:

http://vt100.at/spcp8x5.ko-003

no guarantees

i'm still verifying some low level routines.

best regards,
Hari

[hari]

ok, lets try that:

http://vt100.at/spcp8x5.ko-003

no guarantees

i'm still verifying some low level routines.

best regards,
Hari

i had not seen that you are not using 2.6.20 so I have committed the actual version into the trunk:
http://svn.charonmedia.org/trac.cgi/browser/home/hari/spcp8x5

[hari]

nobody?