On/Off Switches (2477S) blocking communication?

[dpace]

Edit: A few corrections

Installing some new on/off switches has created problems in my existing network. I have been trying to troubleshoot this one for some time, and fortunately I have been able to narrow down the issue somewhat.

Original environment:

• 40 SwitchLinc Dimmers (power line only, model 2476D, puchased back in 2011)
• 3 On/Off switches  (2477S, dual-band, purchased 1/2019)
• ISY 994i (v.5.0.14)
• a few motion detectors

Everything was working fine until I introduced some new On/Off Switches (2477S, dual-band, purchased 2/2019). 

Here's a play-by-play:

1. Added an On/Off switch (to respond to motion sensor), two nearby dimmer switches on the same circuit stopped reliably sending or receiving requests to/from the ISY. (Timeouts about 70% of the time.)
2. Tried a factory reset on the new switch - problem still exists.
3. Removed the On/Off switch. Problem disappeared, and communication was restored to the two affected dimmers.
4. Added a different On/Off switch at the same location. Problem re-appeared. 
5. Removed the On/Off switch and replaced with a dimmer switch. Communication problems disappeared.
6. Linked the motion sensor to the dimmer switch and it has been working flawlessly for a week.
7. Hoping this was an odd issue only impacting that location, I installed an On/Off switch in another location (to control a fan). 
8. Now I have communication issues with the dimmers that are located near THAT on/off switch.

After turning on logging in the Event Viewer to show the most info (3. Device communication events), and comparing before/after, I see no difference between the communication before/after the install of the On/Off switch (other than time-out errors). 

The three existing On/Off switches (those bought a month before) are currently operating bathroom fans; all of them function fine (including working within programs), with adjacent dimmers in the same box, and do not exhibit any of these issues.

So it appears I can add/remove my old power-line dimmer switches throughout my home without issue, but anywhere I introduce one of these new On/Off switches, I have problems communicating with nearby dimmers on the same circuit.

I suppose my next step is to take one of the existing on/off switches that I bought before, and place it into the same switch location where I am having trouble with the new switches - that would at least confirm whether or not this is unique to a particular batch of On/Off switches (warranty issue). 

Anyone have any thoughts on what might be happening?

[dpace]

Quick update:

I've now taken that next step and used one of my existing 2477S on/off switches which works fine and placed it into operation in one of the locations where the new 2477S on/off switch seems to disrupt communication. The first time I supplied power, everything was fine and communication between all units worked as expected. Great, this seems to be an issue isolated to the new batch of switches I bought. So then I tried two more of the new switches (I bought a five-pack), and each of them failed. Further evidence this batch is not working as designed, I thought. Then I disconnected and reconnected power to the switch which was working fine, and now it "fails" (prevents communication to other switches on the same circuit) just as the other switches did. If I disconnect the on/off switches, all communication is restored with no errors/delays whatsoever.

This issue is driving me nuts. I am in the process of using a test wire (short romex line run to 3-prong on one end with exposed ends on the other side to connect to each test switch) to help me try to identify the pattern here.

Until I do, I am not quite sure what direction to go. I suppose I could just reach out for technical support on this one? I felt like I've done enough of this stuff in the past that certainly I could figure this out on my own. I will probably go that route and cross my fingers if the next bit of troubleshooting doesn't bear any fruit.

I will keep the thread updated in hopes this could save someone else this pain. 

Doug

[lilyoyo1]

I've always been dubious when troubleshooting issues and multiple devices have the exact same issue. This is especially true when you take something existing and it behaves the same way. When you add the amount of switches insteon sells in any given day, I'm sure we would've heard about more people having issues. 

What are the devices that you're trying to control with the new switches? Are the new switches replacing existing insteon switches? 

[dpace]

Got another update, and I think this is it.

This appears to be an issue with the PLM. It finally hit me to check on communication between devices (e.g. scene controller/responders). That works fine even with the new on/off switches. 

So I plugged my PLM into a different outlet (on a different circuit) and now I am getting communication events through my ISY even to devices with the on/off switch on the same circuit.

All good with the obvious exception of the physical challenges of using a different PLM outlet. I had already wired a new outlet into my server closet which solely serves this PLM (after first replacing it only to find the issue persists). Now, with the communication issues introduced by these on/off switches, the PLM is unable to communicate to a select group of devices unless I abandon both outlets (on two different circuits) that I already have wired to my server closet?

Chalk up yet another frustration with the ever-troublesome, often failing, and extremely finicky PLM.  

Perhaps I will need additional access points, and maybe even filters, to get my network back to what it was before the on/off switches?

For now, I will leave these on/off switches out of my network, and hope for an epiphany while I let my frustration cool from the time I have spent on this issue to date.

 

[dpace]

12 minutes ago, lilyoyo1 said:

I've always been dubious when troubleshooting issues and multiple devices have the exact same issue. This is especially true when you take something existing and it behaves the same way. When you add the amount of switches insteon sells in any given day, I'm sure we would've heard about more people having issues. 

What are the devices that you're trying to control with the new switches? Are the new switches replacing existing insteon switches? 

Completely agree with your logic, and as you see above you were exactly right. It was indeed something else, and that was the PLM. To answer your questions, the switches weren't replacing existing switches - they had basic paddles before. All my on/off switches are controlling fans (either bathroom exhaust or ceiling fans). 

[Techman]

Is you PLM plugged into a power strip or UPS?

What is the load on the switches you're having a problem with.

Are both sides of you power line bridged via ISY dual band modules?

Were / are any other electronic devices, i.e. UPS, power supplies, computer, etc,  plugged into the same circuit as the PLM?

 

[dpace]

3 hours ago, Techman said:

Is you PLM plugged into a power strip or UPS?

The PLM is plugged into its own dedicated outlet.

Quote

What is the load on the switches you're having a problem with.

The load on each of the switches I'm having trouble with is varied (recessed LED lights, step lights, single overhead light). Then again, these are the switches that I am unable to communicate with after putting in the on/off switches. These switches work fantastic without the on/off switch in the mix. The loads on the on/off switches are fans, but they never even need to be on at all for these problems to occur. As soon as the switch is online, the PLM is unable to reach the other switches - and I get an error in the ISY confirming the failure to communicate. 

Quote

Are both sides of you power line bridged via ISY dual band modules?

 

I would imagine so (I have enough dual-band devices that I would strongly suspect), but I will confirm this.

Quote

Were / are any other electronic devices, i.e. UPS, power supplies, computer, etc,  plugged into the same circuit as the PLM?

I believe the current outlet is on a circuit removed from other electronics but I will do a thorough check on that. This outlet was actually added after the PLM refused to communicate on an existing dedicated outlet that did have electronics. Getting an outlet on a circuit with no other electronics can be a tall order these days!

Still does not add up - communication to roughly 50 devices is excellent without the new switches. With them, the PLM fails to reach the same powerline switches on the same circuit it otherwise reaches with no issue. But perhaps my system is weak (due to interference, etc.) and somehow these new switches are just the proverbial straw that broke my little Insteon camel's back?

I will diligently work on confirming the items mentioned above and follow-up once complete. 

Thanks for the help.

[Techman]

Take a look at these trouble shooting tips:  https://wiki.universal-devices.com/index.php?title=INSTEON:_Troubleshooting_Communications_Errors

You can ignore the paragraph that suggests plugging a dual band module into the PLM

Try doing a factory reset / restore device on one of the problem devices to see if that makes a difference.

Try unplugging some of your electronics such as TV's UPS, Power supplies to see if that corrects the communication problem with the switch.

What is the firmware version of the problem devices, it can be found in the ISY when you clock on the device. 

Do a 4 tap test with your dual band modules to verify that both sides of the power line are coupled

You can also try setting your event viewer to level 3 then try to communicate with a problem device. Copy and paste the results in a post.

Power line noise from electronics can be eliminated by plugging the electronic device into an Insteon Filterlinc. 

 

 

 

 

 

 

[mwester]

Pray tell, good people who've posted on this thread (and on others with similar suggestions), why the fixation on "circuits" with Insteon?   So let me understand --  somehow noise on one circuit cannot pass through the circuit breaker panel to the one that has the PLM?  Yet, at the same time, when everything is working, the signal from the  PLM freely flows to all the circuits on that half of the power line, passing unhindered through all of those same circuit breakers that somehow at the same time manage to block noise?  And then, as it relates to a signal-sucker on one circuit, do these magical circuit breakers also have some marvelous ability to isolate the signal from the PLM so that its amplitude is only affected on the circuit that has the signal sucking device?  Really?  Can this be true??  Such a marvel, these magical circuit breakers!  Amazing!    <sarcasm mode="off"/>  Ok, clearly none of that is true... so while there are MANY excellent reasons to have isolated circuits (and I especially recommend ones with a dedicated ground), the truth is that Insteon is NOT one of those reasons!  I keep seeing well-meaning posts that suggest that there's something magical about a circuit breaker... and it's just not the case (I wish it were, maybe I wouldn't need to have my walls littered with Insteon filterlincs...)

 

So, the first rule of troubleshooting: EVERYTHING in your house, and i mean EVERYTHING that's plugged into any outlet, wired into any circuit, even if the power switch on it is turned off, is suspect.  By all means, start with things closer to the PLM and the switches that arent' working -- heck, you gotta start somewhere, and that's as good a place as any.  But do not think for a moment that the problem can't be on some other room, on some other floor, or even in an out-building -- heck, it doesn't even have to be on your OWN house -- while less likely, it is absolutely possible that your neighbor sharing a transformer with you is putting the noise on the wire (consider for a moment why the X10 world has that switch to select which "house" to use -- and consider that the X10 signal is a small percentage different than the Insteon signal -- so of course Insteon signals can and will travel to other houses!).

Also consider the stacking effect -- the Insteon signal is a certain voltage put on the electrical line when the line is normally quiet (which happens 120 per second, for a very short period of time).  And as  you add more and more "things" that suck down that signal to the system, the signal-to-noise ratio will drop, eventually getting to the point that your devices can't communicate.  Pulling out a device, or disconnecting a load somewhere (anywhere) can be just enough to make it work again -- "ahah!" you'll shout, found the problem! - but in fact, you could have pulled out any number of other devices or loads that could have just gotten things to point where they'll work.  For a while.  It's pretty tricky to get this stuff working, and you're working blind because none of these devices actually report useful information like signal-to-noise ratios...

Now, all that said - it could be your PLM.  Mine certainly went through a period of reduced sensitivity before it failed entirely.  So maybe replacing it isn't a bad idea.  But don't go being surprised if you're right back where you're at now sooner or later.

I see a truckload of filterlincs in your future, I do.

 

[oberkc]

5 hours ago, mwester said:

why the fixation on "circuits" with Insteon?

My experience, based upon suggestions from others around here, was that putting a PLM on a dedicated circuit was very helpful.  While I have never measured it, I assume that the capacitance and resistance of wire is such that a noise-maker far away has less impact on a given device than a noise-maker plugged into the same circuit and even less impact that one plugged into the same outlet.  If that were not true, why would moving a device "close" to the PLM matter?  (Some suggest temporarily plugging a device into the same outlet as the PLM as a trouble-shooting method.)  If signals (good or bad) were not degraded over distances of wire, why is there a communication range for insteon?  Additionally, I suspect that the devices most often plugged into the outlets near the PLM are some of the worst offenders in the insteon world, including UPS and power supplies.

As for me, I consider the PLM to be of sufficient criticality (EVERY device is linked to it) that putting it on a dedicated circuit (or at least filtering the other devices) to be well-worth considering and failure to do so as a potential source of trouble.  

[lilyoyo1]

6 hours ago, mwester said:

Pray tell, good people who've posted on this thread (and on others with similar suggestions), why the fixation on "circuits" with Insteon?   So let me understand --  somehow noise on one circuit cannot pass through the circuit breaker panel to the one that has the PLM?  Yet, at the same time, when everything is working, the signal from the  PLM freely flows to all the circuits on that half of the power line, passing unhindered through all of those same circuit breakers that somehow at the same time manage to block noise?  And then, as it relates to a signal-sucker on one circuit, do these magical circuit breakers also have some marvelous ability to isolate the signal from the PLM so that its amplitude is only affected on the circuit that has the signal sucking device?  Really?  Can this be true??  Such a marvel, these magical circuit breakers!  Amazing!    <sarcasm mode="off"/>  Ok, clearly none of that is true... so while there are MANY excellent reasons to have isolated circuits (and I especially recommend ones with a dedicated ground), the truth is that Insteon is NOT one of those reasons!  I keep seeing well-meaning posts that suggest that there's something magical about a circuit breaker... and it's just not the case (I wish it were, maybe I wouldn't need to have my walls littered with Insteon filterlincs...)

 

So, the first rule of troubleshooting: EVERYTHING in your house, and i mean EVERYTHING that's plugged into any outlet, wired into any circuit, even if the power switch on it is turned off, is suspect.  By all means, start with things closer to the PLM and the switches that arent' working -- heck, you gotta start somewhere, and that's as good a place as any.  But do not think for a moment that the problem can't be on some other room, on some other floor, or even in an out-building -- heck, it doesn't even have to be on your OWN house -- while less likely, it is absolutely possible that your neighbor sharing a transformer with you is putting the noise on the wire (consider for a moment why the X10 world has that switch to select which "house" to use -- and consider that the X10 signal is a small percentage different than the Insteon signal -- so of course Insteon signals can and will travel to other houses!).

Also consider the stacking effect -- the Insteon signal is a certain voltage put on the electrical line when the line is normally quiet (which happens 120 per second, for a very short period of time).  And as  you add more and more "things" that suck down that signal to the system, the signal-to-noise ratio will drop, eventually getting to the point that your devices can't communicate.  Pulling out a device, or disconnecting a load somewhere (anywhere) can be just enough to make it work again -- "ahah!" you'll shout, found the problem! - but in fact, you could have pulled out any number of other devices or loads that could have just gotten things to point where they'll work.  For a while.  It's pretty tricky to get this stuff working, and you're working blind because none of these devices actually report useful information like signal-to-noise ratios...

Now, all that said - it could be your PLM.  Mine certainly went through a period of reduced sensitivity before it failed entirely.  So maybe replacing it isn't a bad idea.  But don't go being surprised if you're right back where you're at now sooner or later.

I see a truckload of filterlincs in your future, I do.

 

Come on @MWester, we know you hate insteon but that book you wrote is overboard. Zwave is no where near perfect the way you try to portray. 

Yes setting up the system a certain way allows you to optimize it's efficiency and capabilities. However the same applies to every protocol and system. The fact that he has older Powerline only devices means the circuitry is even more important. Let's be real, if this was about 7+ year old zwave technology, the conversation would be similar (albeit for different reasons). The fact his system has worked for so many years without problems is a testimony to how well it can work. 

Right now he's having issues and through troubleshooting he'll get it back to normal. Simply coming on here to bash a technology that you don't like doesn't help him. It's been a while since we've bashed differing technologies. We can create a post to do all that but at least try to help the man first.

[dpace]

Thanks all for the help. While I'm sure we could all do with less bickering, a little venting here and there (usually) keeps everything in check to a degree. I know at least I can get pretty darn frustrated with this. My system has certainly not been flawless - I have had issues here and there, and through careful testing I can typically get things operational again.

While I do appreciate the troubleshooting steps, I admit they seem a bit of a cut & paste job. I liken it to AT&T asking me to reboot my computer when I'm having trouble with my internet. But fine, I also appreciate how taking all these steps in an orderly fashion makes the best of everyone's (unpaid) time before taking on my case as a special on-off. But thank you lilyoyo1 for understanding my particular situation.

I should also clarify: My earlier posts imply I have had all these devices operational for six years here. I actually moved about two years ago, and through a combination of not wanting to continue supporting that Insteon system after I leave and my own frustration with lowball offers on my last home (will. not. digress. grrr), I actually pulled out the entire network and relocated to my current address. (Bitter? lol) That experience gave me a bit more exposure to restoring devices, invalid links, and some comedy (e.g. "Sorry Olivia, I went downstairs and forgot your overhead is still linked as a responder to the hall switch - I'll fix that tomorrow honey. Please go to bed.") 

While the prior house was built in 2012, the current home was built in 1996. Oddly, I generally have less trouble in this home. Used to have more drop-outs, some unexpected "all-on" events, and one troublesome arc-fault breaker that kept tripping on Insteon signals. 

My PLM is new (within the past year, and my third one at that) - I think/hope that is not the issue. But this idea of putting the PLM close to the breaker box is intriguing and something I would love to do. My server closet is in the center of my house. Just about as excited to run a new ethernet wire out to the garage as I am to find a "clean circuit" and bring that electrical line into that server closet (making THREE circuits that deliver power to that closet - haha). 

 Will keep everyone updated with the eventual "solution" (band-aid or permanent fix - feel free to draw your own conclusions).

[mwester]

5 hours ago, lilyoyo1 said:

Come on @MWester, we know you hate insteon but that book you wrote is overboard. Zwave is no where near perfect the way you try to portray. 

Yes setting up the system a certain way allows you to optimize it's efficiency and capabilities. However the same applies to every protocol and system. The fact that he has older Powerline only devices means the circuitry is even more important. Let's be real, if this was about 7+ year old zwave technology, the conversation would be similar (albeit for different reasons). The fact his system has worked for so many years without problems is a testimony to how well it can work. 

Right now he's having issues and through troubleshooting he'll get it back to normal. Simply coming on here to bash a technology that you don't like doesn't help him. It's been a while since we've bashed differing technologies. We can create a post to do all that but at least try to help the man first.

I said NOTHING wrt to alternate technologies -- it's all in your head. I grow weary of your crap - put me on your ignore list and do us ALL a favor.

 

My post provides valuable accurate technical detail on power-line signals in the real world.  Feel free to dispute that -- but YOUR response is clearly a knee-jerk reaction that was posted without even reading any of it.  Since you won't read my posts anyway, please just ignore them rather than poison them with your attacks.

[apostolakisl]

I've seen similar issues.  Where everything is working great, I add a new switch and stuff that worked 99.99% of the time suddenly starts failing 10% or 20% of the time.  I can't explain that with signal suckers or noise makers.  I suspect it has something to do with overlapping or timing issues with the Insteon comm introduced with one new device.  Without an oscilliscope you'll never know.  I've also seen where then adding another new device made things right again.  

Having stuff on separate circuit breakers is really only helpful when trouble shooting.  You can shut down circuit breakers when looking for troublesome devices, easily taking out whole sections of the house at once when narrowing things down.  I see no reason to have a PLM on its own breaker.  While you are increasing the wire run to the closest noise/sucker, you are also increasing distance to your next Insteon device.  So you'll have noise attenuation, but you'll also have signal attenuation.

If you have some lamplincs, consider playing around with those.  They are super easy to move around and test.  I have found that sometimes plugging in a lamplinc somewhere makes the whole system work great.  But plug it in somewhere else and things get flaky.  Fortunately, since replacing my entire Insteon setup with dual band devices a couple years ago, I have very little of these issues.

[lilyoyo1]

9 hours ago, mwester said:

I said NOTHING wrt to alternate technologies -- it's all in your head. I grow weary of your crap - put me on your ignore list and do us ALL a favor.

 

My post provides valuable accurate technical detail on power-line signals in the real world.  Feel free to dispute that -- but YOUR response is clearly a knee-jerk reaction that was posted without even reading any of it.  Since you won't read my posts anyway, please just ignore them rather than poison them with your attacks.

In this particular post you're right you didnt. However most times that's not the case. In most posts you dog insteon and tout zwave. 

If you actually tried to assist without talking crap you wouldn't get crap back. If you can't stand for someone saying something back don't say it. 

[lilyoyo1]

9 hours ago, dpace said:

Thanks all for the help. While I'm sure we could all do with less bickering, a little venting here and there (usually) keeps everything in check to a degree. I know at least I can get pretty darn frustrated with this. My system has certainly not been flawless - I have had issues here and there, and through careful testing I can typically get things operational again.

While I do appreciate the troubleshooting steps, I admit they seem a bit of a cut & paste job. I liken it to AT&T asking me to reboot my computer when I'm having trouble with my internet. But fine, I also appreciate how taking all these steps in an orderly fashion makes the best of everyone's (unpaid) time before taking on my case as a special on-off. But thank you lilyoyo1 for understanding my particular situation.

I should also clarify: My earlier posts imply I have had all these devices operational for six years here. I actually moved about two years ago, and through a combination of not wanting to continue supporting that Insteon system after I leave and my own frustration with lowball offers on my last home (will. not. digress. grrr), I actually pulled out the entire network and relocated to my current address. (Bitter? lol) That experience gave me a bit more exposure to restoring devices, invalid links, and some comedy (e.g. "Sorry Olivia, I went downstairs and forgot your overhead is still linked as a responder to the hall switch - I'll fix that tomorrow honey. Please go to bed.") 

While the prior house was built in 2012, the current home was built in 1996. Oddly, I generally have less trouble in this home. Used to have more drop-outs, some unexpected "all-on" events, and one troublesome arc-fault breaker that kept tripping on Insteon signals. 

My PLM is new (within the past year, and my third one at that) - I think/hope that is not the issue. But this idea of putting the PLM close to the breaker box is intriguing and something I would love to do. My server closet is in the center of my house. Just about as excited to run a new ethernet wire out to the garage as I am to find a "clean circuit" and bring that electrical line into that server closet (making THREE circuits that deliver power to that closet - haha). 

 Will keep everyone updated with the eventual "solution" (band-aid or permanent fix - feel free to draw your own conclusions).

You're welcome. I don't think it's your plm as you would have more trouble communicating than you do now.

Being in your server closet could be an issue due to the amount of electronic equipment that you have in there. (Assuming alot since you specifically called it a server closet). I would definitely recommend getting it away from as much other equipment as possible. Unfortunately being tied to Powerline does have its drawbacks. 

[apostolakisl]

22 hours ago, dpace said:

Thanks all for the help. While I'm sure we could all do with less bickering, a little venting here and there (usually) keeps everything in check to a degree. I know at least I can get pretty darn frustrated with this. My system has certainly not been flawless - I have had issues here and there, and through careful testing I can typically get things operational again.

While I do appreciate the troubleshooting steps, I admit they seem a bit of a cut & paste job. I liken it to AT&T asking me to reboot my computer when I'm having trouble with my internet. But fine, I also appreciate how taking all these steps in an orderly fashion makes the best of everyone's (unpaid) time before taking on my case as a special on-off. But thank you lilyoyo1 for understanding my particular situation.

I should also clarify: My earlier posts imply I have had all these devices operational for six years here. I actually moved about two years ago, and through a combination of not wanting to continue supporting that Insteon system after I leave and my own frustration with lowball offers on my last home (will. not. digress. grrr), I actually pulled out the entire network and relocated to my current address. (Bitter? lol) That experience gave me a bit more exposure to restoring devices, invalid links, and some comedy (e.g. "Sorry Olivia, I went downstairs and forgot your overhead is still linked as a responder to the hall switch - I'll fix that tomorrow honey. Please go to bed.") 

While the prior house was built in 2012, the current home was built in 1996. Oddly, I generally have less trouble in this home. Used to have more drop-outs, some unexpected "all-on" events, and one troublesome arc-fault breaker that kept tripping on Insteon signals. 

My PLM is new (within the past year, and my third one at that) - I think/hope that is not the issue. But this idea of putting the PLM close to the breaker box is intriguing and something I would love to do. My server closet is in the center of my house. Just about as excited to run a new ethernet wire out to the garage as I am to find a "clean circuit" and bring that electrical line into that server closet (making THREE circuits that deliver power to that closet - haha). 

 Will keep everyone updated with the eventual "solution" (band-aid or permanent fix - feel free to draw your own conclusions).

I really wouldn't start running new wires at least not in the walls and such.  I would do some extension cords and free runs of cat6 to test before  running new wires in the walls/attic/etc.  But before messing wtih that, I would get some dual band lamp lincs or on/off modules.  Plug them in at various locations and see if you can't get the signal to/from the PLM.  The fact that adding new Insteon devices caused your problem tells me this is not an issue with noise or signal suckers, but rather the exact nature of the signal propagation through your home.  Putting some additional dual band devices (in what might seem like random locations) can change the "web" of signal propagation. The Insteon range extenders have the same Insteon signal processing/repeating as the lamp lincs and on/off modules, so don't waste your money on them. Lamplincs are cheaper and you can always use them to control something as well.

[dpace]

2 hours ago, apostolakisl said:

I really wouldn't start running new wires at least not in the walls and such.  I would do some extension cords and free runs of cat6 to test before  running new wires in the walls/attic/etc.  But before messing wtih that, I would get some dual band lamp lincs or on/off modules.  Plug them in at various locations and see if you can't get the signal to/from the PLM.  The fact that adding new Insteon devices caused your problem tells me this is not an issue with noise or signal suckers, but rather the exact nature of the signal propagation through your home.  Putting some additional dual band devices (in what might seem like random locations) can change the "web" of signal propagation. The Insteon range extenders have the same Insteon signal processing/repeating as the lamp lincs and on/off modules, so don't waste your money on them. Lamplincs are cheaper and you can always use them to control something as well.

Appreciate the recommendation, and yes, that is exactly what I am doing. I have done enough wire runs (electrical as well as low voltage) to have learned to test and retest before making things final. (I'm also a software developer by day, so it's just part of my nature... at least when my head is on straight.)

So yes, I have since been playing around with plugging my PLM into various sockets using an extension cord. In addition to that, I mentioned earlier I have a little "tester" that has proven valuable in all this - it's just a few feet of electrical cable (standard 14/2 romex) run to a 3-prong plug on one end with exposed leads on the other which I use to quickly wire in various devices to various locations without having to open up existing switch locations. Connecting that to one of my new on/off switches makes quick work of proving out that, with it plugged into a particular plug, my PLM ceases to communicate with other powerline dimmers on the same circuit, even though communication to the on/off switch on this tester works just fine. Unplug the tester, and everything returns. (Fortunately, the lights I want to test are also Alexa-enabled via ISY Portal, so I am just the odd guy yelling commands while plugging and unplugging . "Kids, ignore Dad - he's trying to fix the lights.") 

One consistent recommendation has been to add some additional dual-band items (lamplinc, on/off modules, access point/range extenders) to improve the reliability of my network. What I find peculiar: my new on/off switches are all dual-band. So wouldn't the addition of 5 more (to the existing 3 already in place) do exactly that, while instead I am seeing a negative impact?

As I read more about what the Insteon Noise Filter (FilterLinc) actually does, certainly several devices in my server closet could be the worst offenders for noise. In there, I have: A/V receiver, gaming systems, NAS, multiple video streaming devices, gigabit fiber modem, router, multiple ethernet switches (including a PoE switch), UPS, power conditioner, infrared repeaters, and last but not least, the ISY.  Several of those give me pause, but one in particular: my power conditioner. That pretty massive device (Monster HTS 3500 MKII), has all kinds of circuitry in it to reduce several types of noise. And those noise filters, according to the FilterLinc product info, can cause issues: 
Another common, yet easily correctable, problem affecting INSTEON signals is signal attenuation, or absorption. Because some devices are built with power supplies that kill noise on the AC line—and they often mistake INSTEON signals for noise—these devices may absorb INSTEON signals before their arrival at the receiving module. But the 10-amp plug-In noise filter eliminates those obstacles by blocking INSTEON signals from getting to them in the first place. The most common sources of signal attenuation include computers and audio/video equipment, particularly big-screen TVs.

In addition to the potential of adding a FilterLinc in front of that power conditioner (which itself is in front of many of the devices in that closet), my search for a clean nearby circuit is going well. I have a dedicated circuit that solely serves my dining room outlets (all of which are unused, so basically a free dedicated circuit), which coincidentally is between my server closet and my garage, making it a great potential. Every test using my PLM connected to an outlet hidden behind a buffet table (yay!) in that room has been successful. I am seriously considering running that new ethernet wire into that wall, as it just seems to make sense to get that PLM away from all that gear. 

Yes, while I recognize mwester's points regarding how all circuits are fundamentally connected (how else can I have connection), if my problems go away by using this plug, I will probably go that route, and probably add a FilterLinc or two into that closet for good measure.

By the way: No one has suggested a good troubleshooting tool they use other than trial and error. Is there such a thing? (There was a reference to an oscilloscope which seems like a fun - yet costly - thing to own, but still not sure that would help me much.)

[Techman]

Quote

 

Just about everything you mentioned in your server closet has the potential to wreak havoc on the powerline and interfere with the Insteon mesh network..  The Filterlincs are rated at 10 amps so you'll probably need a few of them.

Unfortunately there's no troubleshooting tools other than trial and error and a good understanding of the Insteon mesh network. 

I think adding dual band modules and filterlincs are your best option.

Run a diagnostics on your on/off switches. Right click on the device, then click on diagnostics |  Show device link tables. After the table fully populates click on compare. That will show if you have any mismatched device links. If so then do a restore device.

If the diagnostics runs unusually slow then you most likely have powerline signal / noise issues.

 

 

 

[apostolakisl]

@dpace

No, there is no good tool besides an oscilliscope.  With one, you can plug things in, turn things on/off, etc and directly see on your screen what it does to your power line noise and insteon signals.  There is no budget tool that does this, only trial and error.

Your system worked fine without your new dual band devices.  So the issue is not as simple as noise from your other gizmos.  It wouldn't have worked before adding the new devices if that were it.  I am not saying that noise from that stuff is playing no role, just that it isn't the sole reason.  Personally, I would keep your PLM where it is.  I have a big trip-lite power strip in my server room.  I have an outlet with a filterlinc plugged in.  The filterlinc has a passthrough plug that I plug my PLM into.  The filtered plug on the filterlinc has my UPS plugged into it and then the big triplite power bar is plugged into that.  So everything, except the PLM, in that room is behind the filter.  If you need more than 10amps, then you'll need more than one filterlinc.

Counter-intuitively, adding even more devices can undo the bad caused by adding the first few.  It is a complicated web of receive and repeat that makes Insteon work.  I suspect what you have done with the new devices is alter the pathway back to your plm.  Add a few more devices and you may flip the path around to something else that works.

In summary, I would filter everything like I described in that computer room, and play around with some dual band plug in modules of one sort or another.

[lilyoyo1]

4 hours ago, dpace said:

By the way: No one has suggested a good troubleshooting tool they use other than trial and error. Is there such a thing? (There was a reference to an oscilloscope which seems like a fun - yet costly - thing to own, but still not sure that would help me much.)

As already mentioned an oscilloscope is the best way. The funny thing about troubleshooting is that most of us has been doing this for so long, it's just as easy to troubleshoot with or without equipment. When it comes to installation, I generally look at the usual suspects and plan around them

I know insteon does make a diagnostic keypad which may help you. 

One thing I would try is to swap an off in place where one of your dimmers are to see if you experience the same issue. You can also place an existing switch into one of the locations as well to see if the same thing happens.

[dpace]

3 hours ago, lilyoyo1 said:

One thing I would try is to swap an off in place where one of your dimmers are to see if you experience the same issue. You can also place an existing switch into one of the locations as well to see if the same thing happens.

Actually yes, I have definitely confirmed this to be the case. After installing the on/off, I introduce the problem. Swap out that on/off switch with my old powerline dimmer switch, and no issue at all. That is how I have it running now actually. (See #5 in my original post.)

This has happened in two places, on two different circuits. I can seemingly add as many of my dimmers as I want (I still have four more unused) without issue.

[dpace]

5 hours ago, apostolakisl said:

Your system worked fine without your new dual band devices.  So the issue is not as simple as noise from your other gizmos.  It wouldn't have worked before adding the new devices if that were it.  I am not saying that noise from that stuff is playing no role, just that it isn't the sole reason.  Personally, I would keep your PLM where it is.  I have a big trip-lite power strip in my server room.  I have an outlet with a filterlinc plugged in.  The filterlinc has a passthrough plug that I plug my PLM into.  The filtered plug on the filterlinc has my UPS plugged into it and then the big triplite power bar is plugged into that.  So everything, except the PLM, in that room is behind the filter.  If you need more than 10amps, then you'll need more than one filterlinc.

Counter-intuitively, adding even more devices can undo the bad caused by adding the first few.  It is a complicated web of receive and repeat that makes Insteon work.  I suspect what you have done with the new devices is alter the pathway back to your plm.  Add a few more devices and you may flip the path around to something else that works.

In summary, I would filter everything like I described in that computer room, and play around with some dual band plug in modules of one sort or another.

Brilliant. Thank you. Excellent job of explaining this.

I do have four more of these on/off switches which I planned to install. Perhaps I will boldly move forward with installing those. Then, if the issue persists (which means now I have tons of my old powerline switches unable to communicate to my PLM), I will see if a FilterLinc helps. That is, after I find a nice comfortable chair to enjoy a period of self-pity. 

You mentioned you bit the bullet a while back and switched everything over to dual band, and that has greatly improved your situation. With the two cases of seeing my old powerline switch work fine where the new dual-band on/off did not, I admit I have a fear they just do not play that well together. But logically, I realize it just does not make sense.

[arw01]

There us a discontinued x10 power meter that shows signal strength on a led graph.  I have used it to help find signal suckers in the past, but filters are king. My microwave is a sucker, but 10 amps is not large enough.

I have some dual band in the same poly box running the same lights that are flakey for communications. I had a massive improvement when I put my plm on a filterlinc and forced it to use wireless when I was replacing it with the new one.  Timeouts went away and reprogrammjng flew through prior trouble spots!

[Brian H]

JV Digital Engineering still has the XTB-F15. 15 Amp plug in filter available. If 15 amps is heavy enough for you and it has an acceptable look for where it maybe installed.

http://jvde.us/xtb-f10.htm