switchlinc state not recognized in ISY

[bbrally]

Long ago I set up my switchlincs through the ISY and set scenes for the switchlincs to control. Everything has been working fine, but now I am trying to write some programs for the ISY and am now running into problems. The problem is that when I test for an on/off control of the switchlincs, three switches in the house get no response.

 

When I look in the console at the state of these switches, no changes in states are displayed when I activate these three switchlincs. All the other switches in the house change their states in the console properly, just not these three.

 

These three switchlincs activate their scenes properly, and I can activate the scenes properly from the console. I just can't seem to use them in programming a condition.

 

Any thoughts as to why this is? I've tried restoring these devices with no luck.

[LeeG]

Either a problem with the link records in those three Switchlincs or the link records in the PLM for those three Switchlincs. This is assuming good communication between the PLM location and the SwitchLincs.

 

Run a Show Device Links Table for one of the SwitchLincs that is not reporting state changes. When the display is complete click Compare. A right click on the Switchlinc node with a Restore Device will rebuild the device link database. If that does not resolve the problem it is likely a PLM link record problem. A Restore Modem (PLM) will rebuild the PLM link database.

[bbrally]

I ran the compare device links and got the results below

 

I'm assuming this is OK?

 

I then restored the PLM, but still have no luck. Any other thoughts?

 

[LeeG]

Depends on what type of device 13.1C.09 is. If this is a SwitchLinc the link database is not correct for a SwitchLinc. There must be a link record in the device link database that starts with E2 for the ISY to be aware of state changes. It gives the appearance the ISY thinks this is a Responder only device.

 

What does the Admin Console show on line two on the right side, below the node name, for the device type and the firmware level?

 

Also what ISY firmware is being used?

[bbrally]

Here's a quick screen shot of the info you requested. But if I understand correctly, the switchlinc needs to be a controller. The switches I'm having problems with have been setup as responders only. I'm going to try creating a scene for that switch in which its a controller and then try again.

 

[bbrally]

That solved the problem. I made a scene for each switch making it the controller. Thanks for your quick help.

[LeeG]

Something is very wrong. When a device is added to the ISY the Controller link from the device to the ISY PLM is established when the device has Controller capability. This has nothing to do with how the device may be added to Scenes, such as adding them only as Responders. I have never seen the ISY add a SwitchLinc without the Controller link. The only means to fix this is to Delete the device from the ISY and add it back. Before assigning to any Scenes do a Show Device Links Table and insure there is a link record that is

 

E2 01 PLM Insteon Address xx xx xx

 

Without this link the Device will not communicate state changes.

 

EDIT: creating a Scene with the SwitchLinc as a Controller should not be necessary. If that Scene is ever accidently removed from the ISY state change will no longer be seen. The device should be Deleted and added back as described above

[bbrally]

OK, So I deleted all the scenes I created, then deleted all those devices from the ISY and then added them back. I then did a compare device links and found record mismatches on those devices, so I restored those devices. I still have record mismatches. I've also since restored the PLM, but still the record mismatch. Everything seems to work fine though.

 

[Xathros]

I believe that mismatch is not an issue. That record starts with a 00 which I believe indicates the end of the link database on the device. The ISY does not maintain a matching end of list record (it simulates one like 00 00 00 00 00 00 00 00 instead) and thus the mismatch. Only the first 00 is required on the device for end of list and the remaining old data is useless/ignored. Since the ISY's simulated version does not contain this old data you get a mismatch.

 

Are state changes now showing in the admin console before creating any scenes with the affected devices? It looks like you now have the necessary controller record in the device.

 

-Xathros

[LeeG]

bbrally

 

As Xathros has already indicated that mismatch can be ignored. The first byte of that link record is 00 which indicates End Of List. None of the 7 bytes after the 00 mean anything. You have the required E2 01 ........ link record and should now be seeing status updates without generating dummy scenes with these SwitchLincs as Controllers

 

You are running an older level of ISY. That End of List link record carries an "ignore" annotation on later levels of ISY. I suggest moving to 3.3.10 Official release at least. There were some issues with the ISY and link records in much earlier ISY images. That may explain how the Controller E2 01 link record was lost. Had to do with file management on the SD card.

[bbrally]

All status changes are being reflected on the control panel. I've been running 3.3.10 for a while now, but as long as everything works I'm happy. Thanks again for your help.

[LeeG]

Apologies. The "ignore" annotation has been removed. Checked on my 3.3.10 and 4.0.2, it is not there anymore.

[Michel Kohanim]

Hi LeeG,

 

"ignore" has not been removed. I just checked and we do have:

[identical/Ignore]

 

With kind regards,

Michel

[LeeG]

Michel

 

We may be describing different situations. The OP End Of List has the remains of a previous link record as is the case below. Maybe I am remembering the 00.....00 equals 00.......00 case and the situation below never had an ignore.

[Michel Kohanim]

Hi LeeG,

 

Records with logical deletes (starting with 22), should be Ignored. This said, the same record should exist in both ISY as well as the device.

 

With kind regards,

Michel