Scenes not working with newer 2476S revision

[brkent]

I recently purchased and installed a couple of additional 2476S SwitchLinc switches, but neither will respond to Scene commands through my ISY 99i/IR & 2413S setup. They respond reliably to individual direct commands through the ISY, and I was able to manually link them reliably with other Insteon responders, but they don't work at all with Scenes (although the ISY 99i/IR appears to successfully program them for Scene use).

 

They appear to be a newer revision of the 2476S: Rev6.0 on the faceplate, and I believe it showed v.41 in the Universal Devices Administrative Console - although they are both now uninstalled.

 

I swapped them both out for older versions of the 2476S that I have: Rev5.7 on the faceplate, and showing v.3A in the console. I have a number of these throughout my house, and they all have worked reliably with Scenes. Once I replaced the Rev6.0/v.41 switches with Rev5.7/v.3A, both locations now work reliably with Scenes.

 

Before I return these newer switches, I thought I'd check here if there might be some known compatibility issues with the ISY or perhaps mixing the different switch revisions (I'm doubtful that both of these are somehow bad units)?

[LeeG]

Is the ISY running 3.2.2?

 

The new SwitchLincs are I2CS devices which require 3.2.2. Scenes are one of the things that will not work without I2CS support. The SwitchLinc will have to be Deleted and added under 3.2.2

[brkent]

Thanks for the quick response.

 

I'm presently using 3.1.17, and I see 3.2.2 is in beta.

 

That certainly makes sense given what I've seen, so I'll upgrade to the new version when I install these switches again.

[JackdawFool]

Thank you for posting this question - I was just coming here to ask the exact same thing.

 

In my case I have three ToggleLinc Relays (2466S) that show up as v.41 in the ISY admin console. Direct commands to these switches work perfectly but the switches will not respond to scenes. They also link perfectly when linked directly to other devices without the use of the ISY.

 

When viewing the device link table, I get very strange results. It produces a matrix showing every possible value counting downward for 32 rows:

 

0FF8 : F8 F9 FA.FB.FC FD FE FF

0FF0 : F0 F1 F2.F3.F4 F5 F6 F7

0FE8 : E8 E9 EA.EB.EC ED EE EF

...

0F08 : 08 09 0A.0B.0C 0D 0E 0F

0F00 : 00 01 02.03.04 05 06 07

 

I'm also running 3.1.17 on my ISY. I'll upgrade to 3.2.2 and report back. (Just need to figure out how to upgrade to a beta version...)

[LeeG]

JackdawFool

 

The Show Device Links Table looks like that because it is an I2CS device.

 

In the 3.2.2 Available post is a link to the download of 3.2.2 Zip file. DO NOT unzip file. Go to Help | Manually Upgrade My Lighting to install the downloaded Zip file. There is a prompt to take a backup which I strongly recommend. After the install clear the Java cache and use a URL for 3.2.2 Admin Console. When 3.2.2 Admin Console is running go to Help | About and confirm the Firmware line and UI line are there and both indicate 3.2.2.

[Brian H]

I second the I2CS opinion.

I have also seen that device linc table reverse count down.

[LeeG]

The technical reason the Show Device Links Table looks odd is this. The link table is read one byte at a time with a Peek command. The Peek command tells the device which byte of memory to read by address. The address of the first byte is F8. The Peek command should return the value at location F8. Instead the Peek echoes back the address it was told to read rather than actually reading memory. It does this because it is an I2CS device which likely no longer supports the Peek command. Most I2CS devices are dropping Peek/Poke which have been around since 2005. There are faster methods for reading memory and the I2CS devices are forcing applications like the ISY to use the faster method.

 

The next address read is F9, then FA, then FB and so on for a total of 8 bytes. Then the next link record is read which starts at F0, F1, F2, F3, and so on. That is why the data looks so odd. It is actually the addresses being read rather than the data.

 

I purposely left out some of the details. They are not necessary to understand the concept of what is happening.

[JackdawFool]

I've updated my ISY to the 3.2.2 beta and all my problems with the v.41 switches are solved.

 

I had to delete the switches from the ISY device list just as you said and re-add them using the standard linking procedure.

 

The communication time between the ISY and the switches both when adding the devices and when placing them in scenes is noticeably faster than the older versions... I'd say twice as fast.

 

And thank you very much as well for the more detailed explanation for the results I was seeing in the link tables. I greatly appreciate the help!

[webminster]

I'm assuming this is the problem, that topic is not the same, but assume the problem is the same.

Now all I have to decide is which is more important to me... my client programs or the status updates

[LeeG]

The symptom appears externally to be different. In reality it is the same as you are experiencing. The ISY is notified of a status change because of a Scene established when the ToggleLinc is added to the ISY. Whether it be an explicit Scene that is defined later or the Scene the ISY automatically establishes at device add time to support status change notification, Scenes cannot work because functional link records cannot be written below 3.2.6.

 

You are right on about the choices although some folks have not had issues with Conductor. Not sure that is really a show stopper. It only gets worse over time as all new devices will eventually be I2CS devices. Would be impossible to add new or replacement devices if the ISY firmware is not maintained at the latest level.

[webminster]

Thanks for the info.

Main issue I've seen with Conductor, relative to the discussion, is that it tends to break across ISY firmware revisions. While I haven't tried it yet, I doubt it will work with 3.2.6 (like it broke between 3.1.9 and 3.1.16 at least). That's the root of the choice.