Having problems - all programs not running

[fisix]

Weird problem that seems to have just started today.  None of my programs are executing - it's as if the ISY isn't receiving anything back from the various sensors/switches in the system.  Including motion sensors, light switches (I have some programs tied to fast on fast off), and others.

 

Any hints on what to try/look at?

[fisix]

I should add - it's just the programs.  My tie-ins to Alexa are still working.  Could a power outage have done something weird with the module that links the ISY to the power lines?

[stusviews]

What is your firmware? your UI? Expect the unexpected if they don't match. Check that first.

 

And, yes. A power outage can cause that, too. Try a reboot (Configuration,  bottom of right panel). If that doesn't work, then unplug both the PLM ans USY Power up the PLM wait a few seconds and power up the ISY

[fisix]

Firmware and UI are 4.6.2 and match.  I'm going to go through both the reboot and the hard power boot and I'll report back.  I've tried both but had no luck.

 

I tried to get an 8 scene remotelinc to connect while rebooting and such before, but no joy.  The ISY can send control signals to devices, can read device statuses properly (on/off), and if I run the "then" part of a program, the device does what it should, but the ISY just can't trigger any of the programs (none of the "if" conditions tied to device status/controls work).

 

I just tried the reboot.  Didn't work.  Error log is below, and will try the hard boot next.

 

Error Log:

2017/09/25 22:32:11 System -5 Start

2017/09/25 22:32:13 System -170001 [Network] Established

[fisix]

Hard boot didn't work either.  I tried to replace cables and place the modem in a different socket, to see if that helped, but nothing.

 

Clearly the motion sensor is wireless, but is perhaps the fast on/off feedback through the wireless channel rather than through the wired?  I can control everything through the UI, as far as I can tell, but the remotelinc and the motion sensors aren't working.

 

Could the dual band modem have lost its wireless band?  It's the typical one from smart home, and it may be about 7 years old...

 

Is there a way to diagnose the modem?  Unfortunately, I don't have an extra one around...

[fisix]

Well... actually, I was able to reload the remotelinc, and now the programs tied to those switches run properly.

 

I'm not sure what I'm seeing.  I'm guessing there's some kind of link mapping that's now wrong for the fast on/off (multiple switches), and for the motion sensor.  Maybe a data munging, or maybe the PLM is funky and needs to be remapped.

 

Should I try "restore devices"?

 

I've just now backed up the ISY, and the last backup I have is from version 4.5.4 of the firmware.  Not sure how compatible the backups are...

[fisix]

Ok, I figured it out.  I restored a single switch, and now it works fine.  That means I probably need to restore all devices to be safe.  Sigh.

[fisix]

All is good on the western front.  Restore devices worked fine.  Used this:

 

http://wiki.universal-devices.com/index.php?title=ISY-99i/ISY-26_INSTEON:File_Menu#Restore_Modem_.28PLM.29

 

Thanks!

[oberkc]

I wonder, given the modem restoration having solved your problem, if this is a sign of impending PLM failure.

[Brian H]

How old is the 2413S?

As pointed out they are known to have a failed power supply. After two years and a few months.

One clue would be Link Database in the PLM empty or have few entries. Restore fixes for awhile.

Not responding but a power cycle fixes it for awhile is another.

As is a dim or out LED on the side.

 

There is a thread here on the PLM issues.

http://forum.universal-devices.com/topic/13866-repair-of-2413s-plm-when-the-power-supply-fails/

[fisix]

It's either 6 years old or 4 years old.  I purchased one with my 99i, then got another when I upgraded to the 994i.  I can't find the 99i or the other modem... so I can't tell what the date is exactly, I guess.

 

That'll suck if it's going bad.

[fisix]

How do I check if the link database is the PLM is empty?  I was trying to diagnose last night but didn't really find a way to drill down into the functioning (or not) of the PLM.

[stusviews]

Tools, Diagnostics, Show PLM Links Table. Even a low count (not empty) is an indication of a failing PLM.

 

BTW, what's the four digit code on the white label on the back of the PLM?

[apostolakisl]

Tools, Diagnostics, Show PLM Links Table. Even a low count (not empty) is an indication of a failing PLM.

 

BTW, what's the four digit code on the white label on the back of the PLM?

 

Yes, there is a 4 digit code on a white label.  example: 1015 means tenth week of 2015

 

And you should consider getting a new PLM or look at the instructions on how to repair your PLM (on this forum).  I bet it crashes again within  a week or two.

[Brian H]

You will probably also see a hardware revision number on a white sticker.

Like my latest is V2.3. The one Smarthome says fixed a problem they found.

 

Thread on how to repair.

http://forum.universal-devices.com/topic/13866-repair-of-2413s-plm-when-the-power-supply-fails/

[fisix]

Tools, Diagnostics, Show PLM Links Table. Even a low count (not empty) is an indication of a failing PLM.

 

BTW, what's the four digit code on the white label on the back of the PLM?

 

Mine is V1.5 1130.  Which makes no sense if I interpret 1015 as the tenth week of 2015.  Unless there's a time warp.  And V1.5 I don't think goes back to 2011... let alone the 30th month of that year. 

 

I'm working on finding a backup.  Thanks for the links! 

[Brian H]

I have seen Date Code format changing between early and later modules.

I have a failed V1.5  1122. Now repaired as a spare.

Yours is 30th week of 2011. Mine is 22nd week of 2011.

 

I have also seen a production code sticker on some. That confuses even more.

[apostolakisl]

Yes, there is a 4 digit code on a white label.  example: 1015 means tenth week of 2015

 

And you should consider getting a new PLM or look at the instructions on how to repair your PLM (on this forum).  I bet it crashes again within  a week or two.

Sorry, my bad, had it swapped.  1015 would be 15th week of 2010.  I answered by memory and should have actually looked at it.  It is typically pretty obvious since you generally know roughly how old the unit is.  You get into trouble when something was made in the 12-15 week of the year where year and week are no longer obvious.  But if you look at multiple devices, certainly one of them will make it obvious.

[stusviews]

Over the years, the position of which is the week (not month) and which is the year is reversed. It's usually, but not always, obvious which is which.

[apostolakisl]

Over the years, the position of which is the week (not month) and which is the year is reversed. It's usually, but not always, obvious which is which.

OK then.  Yes, something like 1213 would not be obvious.

[G W]

OK then. Yes, something like 1213 would not be obvious.

Test

 

Sent from my SM-G955U1 using Tapatalk

[apostolakisl]

Test

 

Sent from my SM-G955U1 using Tapatalk

?

[G W]

?

I'm demonstrating a bug in Tapatalk so they can see it and hopefully fix it.

 

 

Everytime Tapatalk quotes a message from you it changes your name to

 

quote name=""a" post="220594"

 

 

Sent from my Nexus 6P using Tapatalk

[fisix]

Alright fellas, just an update.

 

The system went back into the mode where none of the programs that required transmissions from switches/sensor BACK to the ISY were working, AND, on top of that, I couldn't run any of the programs using the various Alexa devices around the house.

 

Just to remind, last time, the device-->PLM programs weren't working, but I could still verbally control things through Alexa... so, it at least seemed like signals from the PLM were getting transmitted out to the devices, but signals from the devices weren't being received by the PLM.

 

I didn't do a lot of testing this time, unfortunately, but given the info you all provided before regarding the PLM failure rates, and how old mine was, I went ahead and ordered the caps to perform a repair.  I did look at the LED light, and it didn't look dim to me... but whatever.

 

In any event, tonight I pulled the old caps, installed the new ones, and then hooked everything back up.

 

Initially, I noted that the device-->PLM programs still weren't working, but PLM-->device controls WERE working, and now I could control devices using Alexa (meaning, I could run the "then" parts of programs and control devices).  So, I thought maybe I had made some headway, but things still weren't back to normal.

 

Finally, I again went through the Restore Devices process.  The process completed with no errors, and now all the programs are working as normal... meaning that both the PLM--> devices and devices-->PLM communications appear to be working.

 

That's great, but I'm not feeling very confident about the reliability of the system at the moment.  Which is a bummer after years of it being bullet proof.

 

So, right now I'm in testing mode.  If you guys have any thoughts on what might be happening, other than the caps going bad in the PLM, I'd love to hear it.  Some possibilities:

1) some other component in the PLM is going bad - it seems like this would be tied to the Rx channel, but it would be pretty far upstream because it appears to be broken for both wired and wireless Rx signals.

 

2) some electrical component in the ISY is going bad - this could be tied to the Rx channel from the PLM to the ISY (Maybe the daughter card in the PLM is going bad?).  This could be the issue - the PLM is plugged into the wall, and the ISY is plugged into a UPS - their grounds aren't the same, so I guess there's a chance that a power surge/outage could have caused a differential voltage to build up between them, enough to damage the communication components, even if they are only linked by transformers or optoisolators.

 

3) some logic component in the ISY is going bad - this could be a memory issue, but it'd have to be pretty mild to basically make the state machine crap out but still allow control signals to be transmitted to devices using the GUI. 

 

4) some device in my group of devices is somehow corrupting the communication channel.  I don't know how this might happen, but from other threads it sounds like some of the Insteon devices can go bad just the PLM does and perhaps send out noise on the communication channel... which could basically fuzz the system and generate erratic behavior.

 

Just something to mull on while I wait to see if the system falls back into non-working mode.  This last time, it took 2-3 weeks for the system to die, from when I last Restored Devices.

 

Thanks everyone.

[fisix]

Another update:

 

After replacing the caps, the PLM has been working flawlessly.  Yay!

 

Thanks again for everyone's help and commiseration.