SwitchLinc 2476S - ISY Turns light on but loses connection

[thelovins]

First let me say that I am pretty new to the ISY but I do have other switches that are working and I appreciate any advise that I can get on this.

 

Here is my setup, I have an ISY 99i that is on the same phase as a 2 way circuit that I am installing a 2476S switch on to control one light. I have made sure that I am on the same phase by locating a outlet that loses power with I trip the breaker for the switch I am working on.

The PLM is plugged directly into the outlet. (I have also tried the PLM on other outlets and it is not on a surge protector.)

 

At first I could not get my ISY to find the switch by entering in the address or by doing a auto linc. I overcame this by taking advise from a different forum and updating my ISY to v. 3.2.4

 

Now the problem is if I manually turn the switch off, it shows communication in the ISY and shows as OFF.

 

Then if I turn it ON through the ISY, the light will come on but I get a message in the ISY that it can not communicate and the ISY still show the switch as OFF.

 

If I manually turn the light off again, the switch instantly refreshes in the ISY and shows as linked.

 

 

This is the second switch that I have tried in the same location with the same result. I have verified that there is only one switch that runs this light, and the 2476s works fine as a switch to turn it on and off without any problems.

 

I have also used a volt meter to verify which wire is the line and which one is the load.

 

Any thoughts on what could be going on?

 

I really appreciate the help.

[LeeG]

Herre is what I posted on the Smarthome forum after looking at the trace.

 

Thanks for the trace. The SwitchLinc did not answer the On request. Since the SwitchLinc turns On in response the issue is communication from the SwitchLinc to the PLM. That is odd because there were several commands from the PLM during the device add to the ISY that must have been answered for the device to have been added. This adds up to either a wiring problem at the SwitchLinc, something the SwitchLinc is powering when turning it is turned On is interfering with powerline communication or the phases are not coupled. First thing to do is disconnect the Red load wire. Cap it off and see if the Admin Console can turn the SwitchLinc On and Off with the load attached without generating errors. I know it is believed the SwitchLinc is on the same phase but perhaps it is not. An easy test is to remove the SwitchLinc from its current location. Connect it to an Appliance cord (plug on one end, bare wires on the other) and plug the cord into the same outlet as the ISY PLM.

[oatflake]

Did you ever get this situation resolved? I am attempting to troubleshoot a very similar situation.

 

I have a several new 2476S switches - one location, however, gives me a lot of trouble. I can have the ISY-99i link the device without problems, as long as the switch is in the "off" position. I can turn it on, but then I can't communicate with the device as long as it is on. If I walk over to the switch and manually click the off button, the ISY-99i shows the device as "off" and I can talk to it again.

 

Originally this is a light fixture with two compact fluorescent bulbs in it. I thought I would remove them and put two 40 watt incandescent bulbs in, but that only helped a slight amount - I can now click the "off" button on the ISY-99i admin page and maybe about 1 in 10 times it will respond to the off request.

 

I've tried moving my two access points all over the house, but it appears that it may be that this one light fixtures is just too "noisy" to work with. Is there any other troubleshooting ideas that might work?

[Brian H]

It does sound like a noise or power line problem on that devices circuit. Could be the CFLs made just enough noise to kill the communications and the incandescents brought the problem level down just enough for it to be intermittent.

 

It may not be the fixture itself.

Can you do a little testing by turning off the breaker for the fixtures SwitchLinc and see what other things turn off. Then you maybe able to determine the problem device by leaving an individual device off and seeing if the SwitchLinc then works better.

[oatflake]

Turns out this circuit runs all over my house. It touches over half of the conventional light switches, and most of those have CFLs attached. It's daylight right now, so they're all switched off, but I'm still wondering if the CFLs can cause problems when off.

[ELA]

CFLs downstream of a conventional switch that is off should not be an issue. Are you having the issue when they are all off?

 

You most often hear about CFLs degrading and causing noise. However, all the CFLs I have tested were signal suckers! There may be some that are not but many are.

I have no CFLs in my home for that and other reasons.

[oatflake]

Yes, unforuntately it appears I am having the issue when everything in my house is off, except for another 2476S which is on another switch. That one seems to have no trouble working at all, regardless of how many other CFLs are on at a time!

 

It seems unlikely, but I'm thinking maybe the light fixture itself might be generating some kind of interference?

 

As far as solutions go - one thought I had was getting a 2477D (Dual-band) dimmer. I'm assuming that the dimmer should at least be able to *receive* wireless signals even if the line noise is super bad? Or is the wireless part of this device only used to communicate with other wireless-only devices?

[oatflake]

Just a follow up. I managed to solve the problem, but I'm not sure why it works.

 

I had a LampLinc Dimmer Dual-Band #2457D2 plugged into an outlet which was on the same circuit as the light switch (2476S) that was giving me problems. I also had another light switch (2476s) that was working fine; only one of the switches was not responding to OFF commands. However, as soon as I unplugged the LampLinc module, the light switch worked.

 

I've heard about "bad" dual-band modules or access points, but I'm not sure this module is defective - it turns on lamps just fine and works as my wireless access point for my other wireless devices like motion sensors.

 

Is it possible this module was intercepting the communications?

 

Anyway, it turns out I was able to relocate the module to another circuit and I can live without the lamp control this was originally plugged into for now.

[Raynentye]

Hi! I'm having a very similar problem with a 2476S as well. However, in my situation, I have multiple Insteon devices on the same breaker circuit and its the only one in the entire house I'm having issues with. The problems originally started occurring before I had my ISY994/IR PRO and was manually programming the 2476S to a couple of 2486DWH8 keypads. While programming, the 2476S didn't seem like it was sending its signalling properly. I was constantly getting an error message from the furthest Insteon Keypad; I seemed to have hit a window of opportunity when it finally seemed to link up all of a sudden. My second indication that something was wrong was when I was able to send a command from the keypads to the 2476S and it would respond, but if everything was off and I turned on the 2476S from it, the 2486DWH8s would not respond. This is without anything else Insteon or other otherwise on throughout the house. I also unplugged the lamp that was on the 2476S's switching circuit. At this particular point, I decided to hold off until my ISY came in and then I'd try programming the 2476S with it.

 

This brings me up to 2 days ago when I finally received my ISY and installed it. It sees everything I installed and also sees all the Scenes I've programmed manually. Its able to control everything and flawlessly sense the status of every device; with one exception: the 2476S I've been having issues with.

 

With all this long and boring explanation of my weeks long troubleshooting (which I have not gone into major detail, but long story short, I once removed all the other Insteon devices with the exception of a Keypad on the same phase and the 2476S and still had problems with the 2476S not sending its signal.) aside, my conclusion is the 2476S is just plain defective. But my curiosity is this: What version 2476S do you have and where did you get it from? I'm wondering if there's either a bad batch or a weak link somewhere. Mine is v6.0, Date Code 1209, Address: 1C.D1.0B and I bought it from Amazon.com early this month.

 

Any info would help! Thank you in advance!

[oatflake]

I bought all 4 of my SwitchLinc Relays about about 2 months ago from smarthome.com. They are V6.0 I think, 1C.CX.XX range, so I think they're pretty close to yours. Sorry, don't have a moment to take the face plate off to look at the datecode (I think that's where they are).

 

However, all 4 of mine are working fine now, and I only ever had a problem with 1 of them. The other 3 worked great.

 

I suppose it could be a mfr problem, or maybe some new software behavior, but I don't know enough about these things to make that determination!

[ELA]

Hello oatflake & Raynentye,

 

While it is possible for a device to be noisy or defective in a way that corrupts other devices communications it is also possible that you have a signal strength issue. While each Insteon device retransmits the signal ( aiding in hops) each device also is a load, and enough loads together bring the Original message send signal level lower.

 

You often hear that more devices will enhance your network communications. In my opinion that is often not the case. I have one room in my house with 11 devices that now experiences high reliability. (This is after a lot of work filtering out all signal suckers). All it takes to bring that reliability down is to add one more device in that room, or plug in a laptop ( signal sucker).

 

I have also purposely created a test network that simulates that room's situation with marginal signal strengths. In that test setup adding certain devices keeps the reliability near the same, adding certain other devices almost totally kills the communication reliability.

 

It can be difficult sorting these types of issues out.

One method I would suggest is pulling the power tab (power down) some devices on the circuit with issues. Then see if the keypad that is farthest away gets more reliable. Once the farthest device away is more reliable then you can add the lamplinc back in to see if it alone corrupts everything. Or if reliability is still good maybe it was previously just one too many loads on that circuit ( low overall signal strength).

 

If you find that it appears that the total number of devices on that circuit is the issue then make sure there are no signal suckers at all on that circuit. You can read posts about what people have identified as signal suckers and filter them all if need be.

 

While in theory adding more Insteon devices adds to simulcasting power I have found that they do not always appear to sync up with each other as well as one might hope.

 

Another thing that can help is to diagram your home circuits and Insteon devices with approx. wire lengths and paths. You can then better visualize how long the signal path might be from your system PLM to each device. Relocating your system PLM to a more central location can be very helpful, if you find that the path lengths seem lopsided. It is all about getting the maximum signal strength to each device that you can.

 

One thing I have yet to experiment with is adding multiple dual band devices into a room with lots of devices in hopes that the first power line hop then becomes strong enough to improve reliability.

As you found, one dual band device alone does not appear to help and may just hurt. I had the same situation in my worst case room. I did try two dual band devices in that room but it was still not enough.

 

What made my worst case room improve the most was filtering all signal suckers ( that took three filters) and moving my system PLM right near the service cabinet. Moving the PLM allowed more of the original send signal to reach that room.

 

Best of luck to you both

[Raynentye]

Hello oatflake & Raynentye,

 

While it is possible for a device to be noisy or defective in a way that corrupts other devices communications it is also possible that you have a signal strength issue. While each Insteon device retransmits the signal ( aiding in hops) each device also is a load, and enough loads together bring the Original message send signal level lower.

 

You often hear that more devices will enhance your network communications. In my opinion that is often not the case. I have one room in my house with 11 devices that now experiences high reliability. (This is after a lot of work filtering out all signal suckers). All it takes to bring that reliability down is to add one more device in that room, or plug in a laptop ( signal sucker).

 

I have also purposely created a test network that simulates that room's situation with marginal signal strengths. In that test setup adding certain devices keeps the reliability near the same, adding certain other devices almost totally kills the communication reliability.

 

 

I have to say, this intrigues me unanimously! I work with (sub) intelligent devices at work. One particular unit is a "self-healing" radio network: if radio A can "hear" radio B but not radio C, but radio C can "hear" radio B but not a, radio B becomes a bi-directional repeater for radios B and C; in other words, a mesh network.

 

I've only had my first Insteon devices for a few weeks now and it seems that Insteon isn't a true mesh network. I'm not bashing it in any way, just pointing out an observation. Besides, its 400% better than x10 I used to work with 8 years ago!

 

It sounds like you're describing a "cacophony" issue. For instance: if you're in a canyon and you yell, soon after, you hear an echo that would be rather intelligible. But if you had a tour group all yelling at the same time in the same canyon, the echos seem like unintelligible noise. Sorta like sonar but not...

 

It sounds like the error correction redundancies have their limits. Of course, with a technology that is fairly cheap (albeit, not x10 cheap) the limitations may be met rather quickly in large installations just like any other type of equipment. That's why there's residential, commercial, and industrial qualities...

 

Sorry for the ramble!

 

Back on track:

 

I have discovered that the 2476S I had an issue with was faulty! I went so far as to remove all the other Insteon equipment except for the 2413S and the 2476S, wired and plugged them into the same branch circuit, and shut down all the rest of the breakers. (Yes, a little excessive, but now I'm planning on buying a few FilterLincs, and x10 XPF hardwire filters to create a bench-test environment.) Using the ISY, I turned on and off the SwitchLinc, I manually turned it on and off and observed. It seemed to work about 7 out of 10 times using the ISY. And the ISY was able to read its status about 6 out of 10. I wired in the new 2476S I received today into the same test environment and the results were 10 out of 10 on both ISY switching and status updating. And yest, I did default the SwitchLinc, before, after and during this ordeal.

 

So, that's that. Its all packaged up and ready to go back to Amazon. Thank you for all your help and insight! Now I wanna experiment more!

[wwebb]

Same problem with new 2476S V6.0 1209 switches. I just bought four for a new installation. I initially tested the ISY (99i/IR Pro, running v3.2.6) with a LampLinc Dual-Band which had been linked to the ISY before it was installed here - the LampLinc continues to work fine in this installation. The ISY and all devices I am working with are on the same phase.

 

Installed the first 2476S yesterday, which I linked to ISY successfully the first time but then had the "ISY turns light on but loses connection" problem. ISY did see when I manually switched it on/off, but failed whenever attempting to control with the ISY. Today reset it to factory defaults and tried again, but no improvement. After deleting it from the ISY, now the ISY will not even see it to link it again, so I can't do any more testing. This switch controls two CFL lights, hence the purchase of the relay switch.

 

Today I installed a second one of the 2476S switches for another exterior fixture, on a different circuit but same phase. It linked to the ISY successfully, but then has the same "ISY turns light on but loses connection". I tried removing the single CFL bulb, but no change in behavior was seen. (I had the other switch turned off so no chance of noise there). ISY always sees manual on/off control of the switch, and if the switch is off then ISY will turn it on sometimes, but reports failed communication after a delay.

 

I need to have these lights (and two more similar) automated before leaving on a trip Friday morning (not optimistic at this point). No time to order different switches!

 

Thanks for any further help, beyond what has been written above which has not resolved it for me.

[wwebb]

Forum Error adding this reply, so I added again (see below) then found this one posted so edited with this note.

[wwebb]

Progress! I moved the LampLinc Dual-band close to the second 2476S and now the switch operates normally (so far).

 

I then swapped the LampLinc for an Access Point and that 2476S continued to operate normally (so far). So that appears to have been a weak powerline signal problem.

 

Later today I will have time to install a third 2476S on another exterior light (2 CFLs) to try that out.

 

I cannot do anything with the first 2476S I installed, as the ISY does not seem to see it when I attempt to link it again. I suspect an ISY database issue, not fully cleaning up when I deleted this device during troubleshooting. How might I go about resolving that, short of factory resetting the ISY and starting over?

 

Thanks!

[LeeG]

wwebb

 

The 2476S has no RF capability so the device must be reached over the powerline. If you have any question about the SwitchLinc working correctly connect it to an appliance cord and plug the appliance cord into the PLM plug point. I do not believe this is a SwitchLinc problem.

 

How was "on the same phase" determined? It is always a concern when no phase coupling is installed because of the belief all devices are on the same phase. Some are under the mistaken belief that all the breakers on the same side of a power panel are on the same phase.

 

It could be the PLM plug point has something that is attenuating the Insteon signal at the source. PC/UPS and other power supplies are common sources of signal absorption. Plugging the PLM into a surge/noise suppressor power strip will kill the PLM signal. If the PLM powerline signal is reduced in any way powerline only communication will be compromised. This can be tested by plugging the PLM into a long extension cord and plugging the extension cord into another circuit in another room. This has to be done carefully when trying to run Insteon on a single phase only.

 

Just saw the last post. That could also indicate the assumption they are on the same phase is incorrect.

 

PLEASE DO NOT FACTORY RESET THE ISY. That will not resolve the powerline problems you are having.

[LeeG]

wwebb

 

I really don't think an inability to communicate with a particular device indicates an ISY database problem. Run Tools | Diagnostics | Event Viewer with Level 3 selected. Use New INSTEON Device specifying the Insteon Address, Name and Auto Discover. Post the trace of the device add attempt and note any error popup messages received.

[wwebb]

Lee,

 

That could also indicate the assumption they are on the same phase is incorrect.

Thanks for your wake-up call on this - I neglected to recall what I know, that in this type of home panel, single circuit-breaker slots in each of the two columns alternate between phases. I will revisit this when I'm back at the location later today.

 

I really don't think an inability to communicate with a particular device indicates an ISY database problem. <...> Use New INSTEON Device specifying the Insteon Address, Name and Auto Discover.

Agreed that the communication problem is NOT an ISY database problem. However, the observed behavior when I attempted to re-add this previously-deleted device to the ISY (using the method you describe), after I added the access point, indicated something outside a communication problem in this particular situation. The ISY did NOT report any error - it simply showed the box indicating it was working, then after a few seconds that box went away with no reported communication or other error. I repeated this process a few times.

 

I'll post what I find out later today, including the suggested trace if I need further help to figure it out.

 

Also - thanks for your information on sources of attenuation and interference which may be a factor in this or any additional Insteon device installations!

[LeeG]

That sounds like the device was not actually Deleted. If a device is added with an Insteon address that already exists in the My Lighting tree there are no messages about already exists, etc the device add just ends in seconds.

[wwebb]

That sounds like the device was not actually Deleted.

Yes, that is what I suspect, that although it is not showing in Admin, the address was still in the database somewhere.

 

Back to communication troubleshooting...

 

I did find that the second switch (now working fine with an Access Point on that circuit) is on the other phase from the controller. My bad.

 

The first switch appears to be on the same phase as the ISY. I turned off the breaker to confirm if the adjacent receptacle was on the same circuit. I also flipped the breaker to reconfirm the receptacle the ISY was on, thus power resetting the ISY. I then discovered that using Start Linking found and added the switch to the ISY again, so either the ISY power-restart and/or the switch power-restart cleared it. However, the switch still had the earlier-described control problems.

 

I next tried the suggestion to disconnect the load (2 CFLs) to see if it was an electrical noise issue. That appears to be the issue - the switch is now controllable from the ISY. I'll next get out the ladder and try replacing the CFLs with standard bulbs and if that works, buy some LED bulbs to try as these lights run all night - after I search this forum for any feedback on using LED bulbs with INSTEON

 

Thanks again for your help Lee!

[wwebb]

For the first installed switch, I replaced the two CFLs with standard bulbs, eliminating the apparent line noise issue. However, I found that the switch is only ISY-controllable if I have the dual-band LampLinc plugged in on that circuit. I will order one of the new dual-band relay switches to replace that 2476S, so I don't need any other dual-band on that circuit which has little use other than those lights. Fortunately the switch box is a deep one, so I won't have the concern others have expressed re: the dual-band relay switch.

 

I installed a third 2476S on the additional exterior lighting which has two CFLs. Thankfully it works without any further effort.

 

My simple program for controlling all three exterior lighting switches is now running (sunset/sunrise with offsets and 15-minute random waits). It appeared to work properly overnight for the first two switches, so I'm confident it will work with the third switch added today until I can come back to swap out the one switch and do final testing.

[oberkc]

after I search this forum for any feedback on using LED bulbs with INSTEON

 

I believe you will find most folks having better experience with LED bulbs than with CFLs, regarding noise or interference. This may not be the case with LED-based fixtures that include an integrated power supply.

 

Agreed that the communication problem is NOT an ISY database problem. However, the observed behavior when I attempted to re-add this previously-deleted device to the ISY (using the method you describe), after I added the access point, indicated something outside a communication problem in this particular situation. The ISY did NOT report any error - it simply showed the box indicating it was working, then after a few seconds that box went away with no reported communication or other error. I repeated this process a few times

 

My anecdotal experience is that adding and deleting items while experiencing a marginal communication environment can result in some strange behavior. If you believe the access points have improved your communication, I suggest to delete and re-add your trouble device, if you have not already done so, with the access points installed.

[Raynentye]

Progress! I moved the LampLinc Dual-band close to the second 2476S and now the switch operates normally (so far).

 

I then swapped the LampLinc for an Access Point and that 2476S continued to operate normally (so far). So that appears to have been a weak powerline signal problem.

 

Later today I will have time to install a third 2476S on another exterior light (2 CFLs) to try that out.

 

I cannot do anything with the first 2476S I installed, as the ISY does not seem to see it when I attempt to link it again. I suspect an ISY database issue, not fully cleaning up when I deleted this device during troubleshooting. How might I go about resolving that, short of factory resetting the ISY and starting over?

 

Thanks!

 

The new 2476S is having a similar problem than the last on my end. I'm just going to get a few phase couplers (one for the main breaker and one for sub panel). It seems this series has a signal reception problem. Could be a manufacturing issue but its more of an annoyance at the moment and doesn't seem to effect the performance of the switch. Its just annoying to see it pop up on the screen! Funny thing is, Mobilinc has no issues communicating a status change to the ISY, and the ISY in turn, has no problem changing the status of the switch! This is either in a wireless or WiFi circumstance, too!

 

When you deleted it, you deleted it from the ISY's database. Try to default the 2476S manually and then try reassigning it to the ISY from there. I had to do that when I took my original problem switch out and attempted to relocate it. When I deleted it from the ISY, I backed out of the Admin screen for a good 30 minutes or so, I wasn't really keeping track as I was rewiring the new switch in and wired up the old switch in another location for experimentation purposes. I'm really not a fan of these flying leads of these Insteon units. I like positive lock screw down terminals! Ended up using screw-type terminal blocks instead of wire nuts (or wire caps or whatever they're called in other parts of the country!) But now I'm getting a bit sidetracked...