Can control all devices but status doesnt work

[bsobel]

Ive noticed with many of my devices (but not all) the Status program item doesnt not work. Also, when I turn a switch on manual the state in the admin console is not updated. However, if I query the switch, I can reliably read its status all the time. However, some switches are sending status (so I know that the PLM is working).

 

I have 478 links in the PLM and 7 access points in the house. I've installed filterlincs on the ups's and tv's, so kinda stumped as what to try next.

 

Anyone else seen this?

Bill

[upstatemike]

I am seeing this as well. I have one switch I wanted to use as a trigger but cannot get it to register local control. I can query it fine. I can control it from the ISY GUI fine. I can link it to another switch and the link works 100%.

 

This tells me the switch transmits fine and my communication with the PLM is fine. I finally decided to give up and wait till 6.8 comes out to see if that changes anything.

 

Something to try....

 

Put the switch into a scene.

Verify you can control the switch by selecting the device in the GUI.

Then verify you can control the switch by activating the scene.

Finally, go into the scene and select the switch there and try to control it.

 

Do all three methods control the switch reliably?

[G W]

Could it be that the switch is not transmitting the data when it is turned on? Maye the switch is bad.

[Michel Kohanim]

Hi Bill,

 

Please go to Tools->Diagnostics->Event Viewer and set the level to 2. Then, go to the problematic device and click on it. If you see any traffic, then we might have a bug. If you do NOT see any traffic the please:

1. Try restoring your device (right mouse click ->Restore Device). If this does NOT work then

2. Try factory resetting your switch, redo step 1, and retry. If this does NOT work either then we have a defective (or OLD) switch

 

With kind regards,

Michel

 

Ive noticed with many of my devices (but not all) the Status program item doesnt not work. Also, when I turn a switch on manual the state in the admin console is not updated. However, if I query the switch, I can reliably read its status all the time. However, some switches are sending status (so I know that the PLM is working).

 

I have 478 links in the PLM and 7 access points in the house. I've installed filterlincs on the ups's and tv's, so kinda stumped as what to try next.

 

Anyone else seen this?

Bill

[upstatemike]

I tried these steps with my switch and get no activity in the event viewer.

 

The switch is a 2476S relay switch that reports as v.2C I also have an ICON ON/OFF v.60 doing the same thing.

 

Query works fine

 

Rebuild works fine

 

Adding to a scene as a controller works fine (so it does transmit to other switches it is linked to)

 

I will swap it in the next day or two and see if a new switch solves the problem.

[Michel Kohanim]

Hello upstatemike,

 

Thanks so very much; I am smelling i2 related issues here! Do you have more than one of those ICON v.60s? If so, do you have the same problem with all of them?

 

With kind regards,

Michel

 

I tried these steps with my switch and get no activity in the event viewer.

 

The switch is a 2476S relay switch that reports as v.2C I also have an ICON ON/OFF v.60 doing the same thing.

 

Query works fine

 

Rebuild works fine

 

Adding to a scene as a controller works fine (so it does transmit to other switches it is linked to)

 

I will swap it in the next day or two and see if a new switch solves the problem.

[upstatemike]

I have 5 of this version ICON in the basement and all reported OK

 

I have 3 in the garage and all reported OK

 

I have 1 other one on the 2nd floor near the problem switch and it fails to report.

 

I took a spare switch and attached it to a cord and plugged it into an outlet near my PC and observed it reports OK.

 

I then took the test switch up and plugged it into an outlet on the same circuit as the problem switch and it reported OK.

 

CONCLUSION 1: I have 3 bad switches that all just happen to be on the 2nd floor.

 

I tried the original bad switchlinc again while the test switch was plugged in and it reported its status!

 

CONCLUSION 2: The signal is somehow marginal on the upstairs circuits such that acks work when the PLM initiatates a communication session but somehow messages initiated by the switches do not get through as valid. Some switches, such as the test switch, must transmit with just enough extra strength to get through. Having the test switch on the circuit allowed the original switch to work by acting as a repeater.

 

I left the test switch plugged in and tried the other 2 upstairs problem sitches (ICONS) again but they still did not work. I then tried the original problem switch again and it did not work despite the fact that the test switch was still plugged in and acting as a repeater. I operated the test switch again and it still reports reliably.

 

CONCLUSION 3: The upstairs switches have a demon or other malevolent spirit residing in them and no amount of testing will give conclusive results. Even though they work perfectly in evey other respect, programming, linking, etc, it looks like I will still have to replace them if I want them to reflect accurate status.

 

I should also note that none of the problem switches are ever reported as having communications issues during the daily 3 AM status query.

[G W]

(plays the theme from 'The Twilight Zone.')

[IndyMike]

Upstate,

 

I was all set to say this was a case of a "missing link" in your Icons. Glad I waited for your last post - saved myself some embarrassment.

 

I don't have any good explanations, but possibly some additional tests (I feel your pain):

 

1) Try programming one of your "problem" switches with an X10 address. If the ISY can see the switch transmitting the X10 On command (when the switch is locally activated) you do not have a signal strength problem. If you can't see the X10 transmission we haven't learned anything.

 

2) Try linking one of your "problem" switches to the "test switchlinc". Right now you can't tell if the test switch is hearing the problem unit. With the units linked, you'll be able to tell if the test switch can hear the problem unit. If the PLM still can't see the problem unit (even with the test switch repeating) we have a conundrum. Possible destructive interference (not supposed to happen)?

 

3) Try swapping your test switch with one of your Icon's (put the Icon on the extension cord). If the Icon still can't communicate with the PLM try restoring it with the switch plugged in next to your computer. If the test switch (installed in the old Icon position) can't communicate with the PLM you may have a noise source that is activating unattended.

 

4) Move an Accesspoint to the circuit - not a solution, just another data point.

 

I understand that you've already put a lot of effort into this. I just hate when the troubleshooting trail ends with malevolent spirits (I had my share). If we can put enough data points together, maybe Michel can solve this one for us.

 

IM

[upstatemike]

I tried the Accesspoint first since that was the easiest option... no difference. Maybe I could increase the range of the RF link by using more than one accesspoint in a sort of RF array... maybe I can arrange a dozen or so into a geometric configuration where the RF adds together into a sort of super Accesspoint?

[upstatemike]

Once again I deleted the problem switch from ISY, did a factory reset on the switch, and added it back with the option to delete all existing links. Still no good.

 

The test switch plugged in at the same location is still fine. I'm too annoyed to mess with it anymore... clearly I have at least 3 switches to replace and perhaps many more... Since there are no other communication symptoms to alert me I might have others that are bad and not know it until I stumble on the fact that the ISY GUI isn't keeping up with local changes.

 

To really know for sure, I will need to use the diagnostic window and check all 105 switches one by one to see if local control is tracking properly in the ISY GUI. I guess I know what I'm going to be doing all next weekend! And even when I find them, how do I convince Smarthome they are bad? They work correctly according to every test the helpdesk tech is going to use to verify a problem.

[upstatemike]

I tried the suggestion of programming the problem switch with an X-10 address. When I operate it locally I do not get any X-10 ON commands but always get X-10 OFF when the switch is turned OFF. It is clearly the switch and not a signal issue.

 

Once I test all the switches and determine the true scope of the problem I guess I can do a temporary fix by having an event to query the problem switches once a minute to keep the status correct.

[Michel Kohanim]

Hello upstatemike,

 

First of all, thanks so very much for the diagnostics. Would you be kind enough to go to Tools->Diagnostics->PLM Links Table and then count the number of links in the PLM?

 

We have been hearing of intermittent issues with PLMs that have more than 2000 links. I have to think that PLM could be a culprit since you already mentioned that these devices can control others (the responders listen for state changes) and thus the only missing link is the PLM itself.

 

With kind regards,

Michel

 

 

 

Once again I deleted the problem switch from ISY, did a factory reset on the switch, and added it back with the option to delete all existing links. Still no good.

 

The test switch plugged in at the same location is still fine. I'm too annoyed to mess with it anymore... clearly I have at least 3 switches to replace and perhaps many more... Since there are no other communication symptoms to alert me I might have others that are bad and not know it until I stumble on the fact that the ISY GUI isn't keeping up with local changes.

 

To really know for sure, I will need to use the diagnostic window and check all 105 switches one by one to see if local control is tracking properly in the ISY GUI. I guess I know what I'm going to be doing all next weekend! And even when I find them, how do I convince Smarthome they are bad? They work correctly according to every test the helpdesk tech is going to use to verify a problem.

[upstatemike]

The PLM link count comes back with 182.

[Michel Kohanim]

Hello upstatemike,

 

For an installation as large as yours, 182 seems to be very low. You do not have any scenes?

 

With kind regards,

Michel

 

The PLM link count comes back with 182.

[upstatemike]

Yes I have quite a few scenes.

 

Also I just swapped the test switch for the problem switch. I went into the ISY and did a replace (I didn't need to since nothing was linked to either switch but I just did it as a quick way to transfer the name to the new switch).

 

The ISY has been busy for over 5 minutes now transferring LNK-DATA UNKNOWN... how do I abort this?

[upstatemike]

OK the process ended on its own after it shut down the admin console. I restarted the admin console, deleted the new switch and re-added it cleanly. It works fine and reports local actions correctly.

 

I will have to re-add the problem switch to do any more testing with it.

 

As for the link count... maybe I am not doing it correctly. The whole PLM table diagnostic screen is kind of scary because there is no description of what the buttons actually do. If I hit START, what exactly am I starting? If I hit LOAD then I will be loading what from where to where? (I tried the Help wiki link but it went nowhere).

 

So, should I hit LOAD then COUNT? Just hitting count comes back with 0.

[IndyMike]

Hi Upstate,

 

Very curious that you reliably see the X10 Off commands but no On commands. Since this is a relay switch, there aren't any option in between (no dimming).

 

Could you possibly verify whether the switch will communicate the Off state reliably in Insteon? You can do this through the "tools/event viewer" set to level 2, or by turning the switch on from the Admin console and then off at the unit.

 

When you posted that you had 182 PLM links, that was 182 Hex (386 decimal) correct? Still sounds a bit low, but I believe I remember that you're running many programs.

 

 

I tried the suggestion of programming the problem switch with an X-10 address. When I operate it locally I do not get any X-10 ON commands but always get X-10 OFF when the switch is turned OFF. It is clearly the switch and not a signal issue.

 

Once I test all the switches and determine the true scope of the problem I guess I can do a temporary fix by having an event to query the problem switches once a minute to keep the status correct.

[bsobel]

Ok, I swapped out the Switchlinc for a new Keypadlinc and I'm still seeing the same issue. The load is dimmable flourescent bulbs, so I removed al of them thinking they were generating noise. Same issue, I cant get the ISY/PLM to see commands from the unit. I dont even see the unit transmitting commands, however if I link it to another switch it starts transmitting (based on the access point's led flashing).

 

Its like its supposed to be linked to the PLM but that link isn't actually getting put into the device. Is that possible?

[Michel Kohanim]

Hello bsobel,

 

I assume that you can control the switch from within ISY, correct? Not having statuses back from a device usually mean the following:

1. Noise -- which you have ruled out

2. Reception -- is it possible that the PLM is NOT hearing

3. No slave link in the PLM -- you can rule this out by doing a Restore Device on your device

4. Defective device

 

The next course of action is to move the problematic switch closer to the PLM. If this works, then the problem is either 1 or 2. If not, then the problem is either 3 or 4.

 

With kind regards,

Michel

 

Ok, I swapped out the Switchlinc for a new Keypadlinc and I'm still seeing the same issue. The load is dimmable flourescent bulbs, so I removed al of them thinking they were generating noise. Same issue, I cant get the ISY/PLM to see commands from the unit. I dont even see the unit transmitting commands, however if I link it to another switch it starts transmitting (based on the access point's led flashing).

 

Its like its supposed to be linked to the PLM but that link isn't actually getting put into the device. Is that possible?

[upstatemike]

Hi Upstate,

 

Very curious that you reliably see the X10 Off commands but no On commands. Since this is a relay switch, there aren't any option in between (no dimming).

 

Could you possibly verify whether the switch will communicate the Off state reliably in Insteon? You can do this through the "tools/event viewer" set to level 2, or by turning the switch on from the Admin console and then off at the unit.

 

When you posted that you had 182 PLM links, that was 182 Hex (386 decimal) correct? Still sounds a bit low, but I believe I remember that you're running many programs.

 

 

Ok here is the final test of this evil evil switch. I connected it to my test cord and added it back to the ISY. I created a test scene and made it a controller for the scene and added anoher device as a responder. Here is what it does:

 

The switch always switches the local load reliably.

 

The switch always responds reliably hen controlled from the ISY

 

The responder in the scene always responds when the switch is operated on or off (so the switch always transmits something) i can also see the LEDs on Insteon devices flash whenever the switch is pressed on or off.

 

Turning the switch on never registers in the ISY even though the linked responder goes on.

(I am watching both the device status and the event viewer.

 

Turning the switch off sometimes but not always registers when the switch is turned off. The responder device always shuts off reliably.

 

Banging rapidly on the switch paddle (either on or off) will eventually cause the ISY to register the change.

 

CONCLUSION: This switch does transmit something when it is operated and that transmission will reliably operate any devices linked to the switch. Whatever it is transmitting does not include a validly constructed message that the ISY is looking for. Because the switch can be forced to eventually register in the ISY by banging on the paddle, it seems likely that the problem involves the transmission of corrupted messages rather than the failure to transmit the message at all.

[IndyMike]

Upstate,

 

You've convinced me that it is indeed a switch problem. Actually, you had me convinced when when you found that X10 signals made it to the ISY.

 

It's almost as if this (these) units were programmed with a "HOP" count of 0. That would prevent any of your other switches from repeating the signal. Said differently, nearby linked devices can respond but not repeat the signal. Distant devices can't reliably hear the device. This would also explain why the accesspoint did not improve things.

 

If the above is true, the PLM should be able to "hear" your problem unit if you get it close enough.

 

 

Turning the switch on never registers in the ISY even though the linked responder goes on.

(I am watching both the device status and the event viewer.

 

Turning the switch off sometimes but not always registers when the switch is turned off. The responder device always shuts off reliably.

 

Banging rapidly on the switch paddle (either on or off) will eventually cause the ISY to register the change.

 

CONCLUSION: This switch does transmit something when it is operated and that transmission will reliably operate any devices linked to the switch. Whatever it is transmitting does not include a validly constructed message that the ISY is looking for. Because the switch can be forced to eventually register in the ISY by banging on the paddle, it seems likely that the problem involves the transmission of corrupted messages rather than the failure to transmit the message at all.

[upstatemike]

But remember the PLM hears the switch fine for query responses and acks... just not the paddle presses.

[upstatemike]

Hello upstatemike,

 

For an installation as large as yours, 182 seems to be very low. You do not have any scenes?

 

With kind regards,

Michel

 

 

OK I think this is more accurate. I now have a count of 812.

[bsobel]

It really seems these switches aren't being programmed to send on/off presses. Im going to go play with homelinc and see if adding the swtich there starts getting it to send...