WONKY performance issues

[trocks797]

Something is going on with our system and it all started early last week. We have added no new appliances. We have 8 dual-band devices spread thoughout the house. The rest are regular powerline.

 

-First off is the reliability. Every night all lights are supposed to come on at dusk. There is always one light that no longer gets the message. The same can be said when dawn comes. One remains on. Maybe two.

-The Insteon outdoor heavy duty on/off switch, when on, turns off then back on whenever there is **ANY** activity on the network.

-Simple programs, like "bedroom lights off when bathroom control off" now take up to 10 seconds to complete.

-The garage lights refuse to turn off. Turn them off via MobiLinc, at the switch, at the ISY portal, whatever, they turn right back on.

 

I did modify the "auto-off garage lights" program a few weeks ago, but I can't understand how that would cause the lights to turn right back on. I attached the level 3 event viewer when I turn the garage lights off. They come right back on. Note: the outdoor on/off module also turns off then right back on per usual (since it is on at the time).

 

Any idea which direction I should take to figure out the source of the problem? I'm hoping the garage lights are the result of programming issues but the reliability and the outdoor module (which is not on any program, as it is indoors) really stumps me! Thank you in advance.

[LeeG]

Device 22.DA.F5 was explicitly commanded to turn On. This is not a communication or device error.

 

Look for a Program(s) that is being triggered by the Status of the device changing to Off which is turning the device back On. Disable all the Programs, enabling one at a time if necessary to identify the Program. From the Programs | Summary tab the device node name can be searched to identify all the Programs that reference the device.

[trocks797]

Device 22.DA.F5 was explicitly commanded to turn On. This is not a communication or device error.

 

Look for a Program(s) that is being triggered by the Status of the device changing to Off which is turning the device back On. Disable all the Programs, enabling one at a time if necessary to identify the Program. From the Programs | Summary tab the device node name can be searched to identify all the Programs that reference the device.

 

Thank you! I took care of that issue, but I am still dealing with the severely degraded network performance. Tonight a brand new dual-band switch failed for the first time. Is there any way to tell if the problem is between the ISY and the switch or the ISY or switch itself?

[LeeG]

It is very unlikely an ISY issue. Could be PLM, more than likely the Insteon Mesh Network in general. The Event Viewer running at LEVEL 3 can be useful in determining how well various aspects of the Insteon Network are working. Send some Direct On/Off commands from the Admin Console and watch the Hops Left=x counts. 2 is best, 1 is okay, 0 is as bad as it gets and still function. A Hops Left=x count that is jumping around when controlling the same device is indicating an Insteon network issue.

 

Some basic things

Electronics at the PLM plug point should be isolated with a FilterLinc

Run the 4 tap Set button test to insure the Access Points are in reliable RF range of each other and on opposite 120v legs.

Do the problems occur when certain loads are On such as CFLs, fluorescence’s, and the like.

 

Providing detail about a problem rather than a “Dual Band device failed†is always helpful.

[trocks797]

It is very unlikely an ISY issue. Could be PLM, more than likely the Insteon Mesh Network in general. The Event Viewer running at LEVEL 3 can be useful in determining how well various aspects of the Insteon Network are working. Send some Direct On/Off commands from the Admin Console and watch the Hops Left=x counts. 2 is best, 1 is okay, 0 is as bad as it gets and still function. A Hops Left=x count that is jumping around when controlling the same device is indicating an Insteon network issue.

 

Some basic things

Electronics at the PLM plug point should be isolated with a FilterLinc

Run the 4 tap Set button test to insure the Access Points are in reliable RF range of each other and on opposite 120v legs.

Do the problems occur when certain loads are On such as CFLs, fluorescence’s, and the like.

 

Providing detail about a problem rather than a “Dual Band device failed†is always helpful.

 

Wow! Thank you for your wealth of knowledge.

 

We have no more access points around the house. Both quit working but historically it was okay since there are now many dual-band devices in the house.

 

Running direct commands, turning on and off the garage light always results in 2 hops left. It's a 2477D. In the same gang box is the driveway lights (they fail the most out of all the outside lights). Max hops left on that is anywhere from 0-1. What would cause this? Like I said, that switch is in the same gang box. It is a 2476D. Trying other lights around the house, the dual-bands seem to return the most 2's, but the patio porch light has everything from 0 to 2 just turning it on and off over and over again.

 

Direct commands from ControLincs results in even worse results. Most are 0 hops left.

 

How can I fix this? More devices? Thank you!

[Roman]

I noticed the same happening on my end... I did have a power failure, which resulted in a few brown outs during that time.. on and off power.. I have since then tried to power off (unplug for 10 seconds) the PLM plug as well as the main ISY99i unitl... Noticed that a few devices needed several cmds sent to them before they started to work reliably once again.

 

With that said, is there a proper procedure to run say just after a power outtage to ensure all devices are working properly, especially the PLM device/plug??? I still have X10 cmds coming from my ALARM panel, and noticed that the PLM seems to be once again not receiving the X10 cmds... Unplugged the PLM plug for 10 seconds, plugged back in then sent 2-4 X10 cmds from the ISY99i and it seems to be working once again?? Luck or proper reset??

 

Roman

[trocks797]

I noticed the same happening on my end... I did have a power failure, which resulted in a few brown outs during that time.. on and off power.. I have since then tried to power off (unplug for 10 seconds) the PLM plug as well as the main ISY99i unitl... Noticed that a few devices needed several cmds sent to them before they started to work reliably once again.

 

With that said, is there a proper procedure to run say just after a power outtage to ensure all devices are working properly, especially the PLM device/plug??? I still have X10 cmds coming from my ALARM panel, and noticed that the PLM seems to be once again not receiving the X10 cmds... Unplugged the PLM plug for 10 seconds, plugged back in then sent 2-4 X10 cmds from the ISY99i and it seems to be working once again?? Luck or proper reset??

 

Roman

 

 

That's one portion of my system that has been reliable- X10 signals coming from the alarm. The X10 power supply for the alarm is plugged into a receptacle that is on the opposite side of the wall, back-to-back with the PLM. It's once the Insteon system has to do stuff that things get unreliable. I just don't get all the 0 to 2 hops left on the same devices seconds later. I guess I need more devices to solidify the network? Not sure how many more can fit in the house!!

[Scottmichaelj]

@LeeG Can you explain if you have two of the same devices in a dual gang box why one device would return a 2 and the other switch a 0? Secondly why would I see my KPLs main button have hops=2 but then the A,B,C or D buttons show hops=0? Is it not really the keypad button but the actual device the buttons are connected to?

[Xathros]

Are the two switches in the same gang on the same circuit?

 

On the Keypad, the difference is likely you are issuing a direct command to the Main button and Scene commands to the secondary buttons.

 

-Xathros

[Scottmichaelj]

On the Keypad, the difference is likely you are issuing a direct command to the Main button and Scene commands to the secondary buttons.

 

-Xathros

 

I get that but those devices I am controlling have hops=2 when queried directly.

[Xathros]

I am unclear then where you see the 0 hops remaining vs 2 hops remaining. Can you post a trace that shown an example?

 

-Xathros

[Scottmichaelj]

I am unclear then where you see the 0 hops remaining vs 2 hops remaining. Can you post a trace that shown an example?

 

-Xathros

 

When I do a query on the keypad the secondary buttons show lower hops. Then if I directly query the controlled device the buttons are controlling those are showing hops=2 - So am I to think maybe the route from the keypad to the actual device is causing the lower hops?

 

As for another circuit you mean are the two switches on different breakers? Sorry for the noob response but how would I know if they are on the same circuit? Both switches are on the same breaker.

 

The communication issues are frustrating me. When I finished building my new home I did everything around Insteon. I have a hardwired phase coupler right in the panel, I have installed the recommended Levitron Surge Protection, I have filterlincs on all my UPS devices, and my PLM is plugged into an outlet I had also installed right by my panel. IMHO I don't think there should any reasons for low hops but maybe I am being naïve. The only thing I dont have is any access points.

[Xathros]

 

When I do a query on the keypad the secondary buttons show lower hops. Then if I directly query the controlled device the buttons are controlling those are showing hops=2 - So am I to think maybe the route from the keypad to the actual device is causing the lower hops?

 

Sorry, I'm still not fully understanding this. The devices the buttons are controlling are different devices so the route very well may be different. The characteristics of the KPL circuit may be different than the circoits that the remote devices are on.

 

As for another circuit you mean are the two switches on different breakers? Sorry for the noob response but how would I know if they are on the same circuit? Both switches are on the same breaker.

 

Same breaker=same circuit. May be the load attached to one of the switches is noisy, may just be a crappy

PLM in one of the switches. Could be they are dual band communicating via RF and one is blocking the others antenna.

 

 

-Xathros

[Scottmichaelj]

 

Sorry, I'm still not fully understanding this. The devices the buttons are controlling are different devices so the route very well may be different. The characteristics of the KPL circuit may be different than the circoits that the remote devices are on.

 

I think you are right, its probably due to the route being different from the controlled device vs just the route to query them.

 

Same breaker=same circuit. May be the load attached to one of the switches is noisy, may just be a crappy

PLM in one of the switches.

 

Again I think your right, maybe something on the actual switch load is causing more noise then another. One of the family room switches is a 2477D and controls 6 recessed can lights with 6 Ecosmart/Cree LED bulbs with a hop=2 - But using the same 2477D (same firmware) switch installed right next to it in the same box has a Hop=0/1. It controls the outlets that has three lamps plugged into it with Philips LED bulbs. Maybe the actual items that is being controlled is affecting the communication? LED bulbs?

[trocks797]

System has performed disastrous today. 3 out of the 4 outside lights failed to turn on at sunset like they were supposed to. Is this X10?!

 

Like I said, direct commands always work. The problem comes when the ISY tells 2 lights to turn on at 50% and the other two to turn on at 85%...all at the same time. I can't remember the last time it worked right. What is the difference in me sitting here at the command center and turning the lights on/off from the ISY and the ISY doing it for me every day at sunset? Can it not handle telling 4 devices to do something at the same time? Should I put "wait" commands between each light in the program to make sure the ISY isn't overloaded?

[LeeG]

The difference in using the Admin Console to control devices directly versus using an ISY Scene (multiple devices) is the Insteon retry mechanism. Direct commands are ACKed and automatically retried by the PLM if an ACK is not received. An ISY Scene command has no ACK and therefore no retry.

 

When Scene devices fail intermittently with different devices not responding each time it indicates a poor Insteon Mesh Network environment.

[trocks797]

The difference in using the Admin Console to control devices directly versus using an ISY Scene (multiple devices) is the Insteon retry mechanism. Direct commands are ACKed and automatically retried by the PLM if an ACK is not received. An ISY Scene command has no ACK and therefore no retry.

 

When Scene devices fail intermittently with different devices not responding each time it indicates a poor Insteon Mesh Network environment.

 

Did not know this. Thank you! So please, what do I need to do? The house isn't that big...I have dual-bands all over the house (more dual-bands than powerline only), and the PLM is centrally located. Should I try FilterLincs? Replacing the 2476Ds with dual-bands? More devices in general? I have one dual-gang box in the garage that I mentioned earlier. In this box is 1 2477 and 1 2476. The 77 has 2 hops left quite frequently. The 2476 rarely has even 1. I put the 2477 in this box specifically to help the reliability of the 2476 but it appears it hasn't helped all that much. While I would love to keep the 2476's part of the system, I have invested far too much money now to deal with an unreliable network, so if replacing them with 2477's is the best answer, I would do that. I just keep remembering how iron-solid reliable it used to be...until a couple of months ago.

[LeeG]

Sorry but I cannot tell you what is wrong with the system. If it worked reliably two months ago what has changed? Appliances can degrade over time but it is more likely something new was added or moved.

 

By using the Event Viewer at LEVEL 3 and watching the hops left counts from various devices it may be possible to isolate the problem to a particular circuit or circuits where the problem is worst. Is it better on one 120v leg or the other.

 

Some basics such as insuring other electronics on the PLM circuit are isolated with a FilterLinc to insure the PLM signal is at its best. Being physically central to the house may not be the best location. If there is a long run back to the phase coupling it may be difficult to get a signal on the other leg. Could also affect moving the signal out to other circuits on the same 120v leg.

[ELA]

 

It controls the outlets that has three lamps plugged into it with Philips LED bulbs. Maybe the actual items that is being controlled is affecting the communication? LED bulbs?

 

Hello Huddadudda,

What Philips bulb?

 

http://forum.universal-devices.com/viewtopic.php?p=69957#p69957

Not the one in red there I hope?

[Scottmichaelj]

 

It controls the outlets that has three lamps plugged into it with Philips LED bulbs. Maybe the actual items that is being controlled is affecting the communication? LED bulbs?

 

Hello Huddadudda,

What Philips bulb?

 

http://forum.universal-devices.com/viewtopic.php?p=69957#p69957

Not the one in red there I hope?

 

I dont know that the number means but the Philips website indicates its a 11w (60w equiv) Medium Base (E26) Daylight (5000k) Dimmable. #046677425265 - However those bulbs or switch never give me issues.

 

I also use the Ecosmart Cree 6" LED for the whole house recess cans. So far been ok - just when adding the new PLM seen some issues with low HOPS.

[Scottmichaelj]

The difference in using the Admin Console to control devices directly versus using an ISY Scene (multiple devices) is the Insteon retry mechanism. Direct commands are ACKed and automatically retried by the PLM if an ACK is not received. An ISY Scene command has no ACK and therefore no retry.

 

When Scene devices fail intermittently with different devices not responding each time it indicates a poor Insteon Mesh Network environment.

 

Did not know this. Thank you! So please, what do I need to do? The house isn't that big...I have dual-bands all over the house (more dual-bands than powerline only), and the PLM is centrally located. Should I try FilterLincs? Replacing the 2476Ds with dual-bands? More devices in general? I have one dual-gang box in the garage that I mentioned earlier. In this box is 1 2477 and 1 2476. The 77 has 2 hops left quite frequently. The 2476 rarely has even 1. I put the 2477 in this box specifically to help the reliability of the 2476 but it appears it hasn't helped all that much. While I would love to keep the 2476's part of the system, I have invested far too much money now to deal with an unreliable network, so if replacing them with 2477's is the best answer, I would do that. I just keep remembering how iron-solid reliable it used to be...until a couple of months ago.

 

Trocks- I see the same results you are. In dual gang boxes if there is a 2477D and 2476D the 76 is lower hops than dual band. But I also have seen in boxes with two dual bands one is lower too. Xathros mentioned one device possibly blocking the other device antenna. Not sure but its a thought.

 

As for unreliable scenes. I have had success in my last house with adding repeats and waits. Also try making a "Scene" with all the devices versus adding each device line by line in the program. When I did my sunset scene line by line what lights to turn on I had more lights not work then putting them all in a scene (including on levels). Then for good measure do a "wait" of say 5seconds then a repeat 1 or 2.

 

Its not going to solve the underline issue but it may help with getting the scenes to work right. I do agree with you, I spent a lot of money on Insteon and sometimes its a pain. I have about 25 2476Ds with about the same amount of 2477D Dual Bands for my network and I dont want to swap/invest in more 2477Ds! Nor do I have 100% reliable system either after spending thousands on my current one. It shouldn't be this hard.

[ELA]

Again I think your right, maybe something on the actual switch load is causing more noise then another. One of the family room switches is a 2477D and controls 6 recessed can lights with 6 Ecosmart/Cree LED bulbs with a hop=2 -

 

But using the same 2477D (same firmware) switch installed right next to it in the same box has a Hop=0/1. It controls the outlets that has three lamps plugged into it with Philips LED bulbs. Maybe the actual items that is being controlled is affecting the communication? LED bulbs?

 

Sorry but I took those comments to mean you were questioning whether or not LED bulbs could be an issue.

 

The link was indicating that yes they can.

 

I agree that getting Insteon communications to be reliable should not be as hard as it can be. My experience has been that on a circuit with a lot of Insteon devices that reliability can suffer due to reduction in the power line signal level because of the heavy signal loading of many devices. Insteon claims that the more devices you add, especially dual bands, that the network gets more robust. Many have experienced just the opposite.

 

When you have such a situation, you have to be sure that all possible signal suckers are filtered. It can take a lot of effort sometimes and you have to be willing put time into testing and experimenting, or punt.

 

Consider the relationship of the PLM to the devices having issues. Is the PLM close enough that it can the communicate directly via RF with the one 2477 in the dual gang?

If so then it might be that replacing the neighboring 2476 , with issues, with a dual band may help.

 

The reason adding dual bands to assist a "neighboring Power line only device" does not necessarily help is that you are then relying on retransmits and simulcasting. My experience has been that there are, or can be, issues with this aspect of Insteon. If the power line signal is very weak on the initial receipt ( or hop #0), I have often seen where subsequent, very strong hops 1,2 & 3 still fail.

 

My goal has always been to get a strong enough initial receipt, either via power line or RF to each device when possible. This is not to say that retransmissions don't ever work, because they do. It just gets complicated when you are somewhere between. When you have marginal initial receipt and many devices attempting to simulcast on subsequent hops.

 

Best of luck to you and to trocks797

[Scottmichaelj]

So I replaced my PLM and have restored it twice and keep rebooting the ISY however my programs are not running correctly still. They work then they don't. When I switched the PLM I physically removed the old one, plugged in the new one, rebooted the ISY, did "restore PLM" from ISY and then restarted the ISY again. Could the new PLM be bad or should I think its my setup? I am not getting any communication errors when I query the system. Its running the scheduled programs but the triggered programs based on devices are not. I am getting fed up now with this and about ready to dump my whole system for something else!

[trocks797]

Sorry but I cannot tell you what is wrong with the system. If it worked reliably two months ago what has changed? Appliances can degrade over time but it is more likely something new was added or moved.

 

By using the Event Viewer at LEVEL 3 and watching the hops left counts from various devices it may be possible to isolate the problem to a particular circuit or circuits where the problem is worst. Is it better on one 120v leg or the other.

 

Some basics such as insuring other electronics on the PLM circuit are isolated with a FilterLinc to insure the PLM signal is at its best. Being physically central to the house may not be the best location. If there is a long run back to the phase coupling it may be difficult to get a signal on the other leg. Could also affect moving the signal out to other circuits on the same 120v leg.

 

Thank you for your continued help. Moving the PLM would cause more trouble than it's worth. It is plugged in next to the cable modem and ISY, and is plugged in back-to-back to the Honeywell X-10 transformer for our security system. Needless to say, the X-10 response is 100% reliable!! It sounds like FilterLinc's are in order. I added 3 more devices tonight (only one being dual-band), and it seems to have helped on table light in the basement...I am seeing 1 hops left compared to 0 before now!

 

Am I correct in thinking I should be spacing the dual-bands out instead of clustering them? I realize I will never know the path of the home wiring, but if I keep the dual-bands scattered evenly, does this help more than say, 2 dual-bands in 1 box?

 

Trocks- I see the same results you are. In dual gang boxes if there is a 2477D and 2476D the 76 is lower hops than dual band. But I also have seen in boxes with two dual bands one is lower too. Xathros mentioned one device possibly blocking the other device antenna. Not sure but its a thought.

 

As for unreliable scenes. I have had success in my last house with adding repeats and waits. Also try making a "Scene" with all the devices versus adding each device line by line in the program. When I did my sunset scene line by line what lights to turn on I had more lights not work then putting them all in a scene (including on levels). Then for good measure do a "wait" of say 5seconds then a repeat 1 or 2.

 

Its not going to solve the underline issue but it may help with getting the scenes to work right. I do agree with you, I spent a lot of money on Insteon and sometimes its a pain. I have about 25 2476Ds with about the same amount of 2477D Dual Bands for my network and I dont want to swap/invest in more 2477Ds! Nor do I have 100% reliable system either after spending thousands on my current one. It shouldn't be this hard.

 

After thinking about it, a 76D in a box with a 77D still won't be at 2 hops left. If the 77D helps the 76D out, that is a hop right there...correct? But then again if I already had 1 hop left on the 76D before the 77D and I am still seeing 1 hop left on the 76D after installing the 77D, was the 77D necessary at all?

 

Anyway...about the outside lights...You're right. Back when I integrated the security system into the Insteon network, I had a program that, when the alarm was tripped (full alarm, not entry delay), all the outside lights would flash repeatedly. Originally, I had each light listed (like I do for the sunrise/sunset). Well, when testing, I noticed the response was terrible. The outside house looked like a twinkling Christmas tree. It was sloppy! The ISY had to do too much and the network was stressing. THEN I added all outside lights to a scene and made the program flash the SCENE 80 times, and it works wonderfully! I didn't use that method for the sunrise/sunset because I was unaware you could set different on levels for each light in a scene. Thank you! I will try that! And while I am on the topic, I still don't know what the "else" function means in the programming....

 

So you're suggesting something like this:

 

IF:

Sunset

 

Then:

Scene - outside lights on

Wait 2 min

Scene - outside lights on

 

correct?

 

I like your thinking, but you're correct when you say this doesnt fix the underlying problem. Here's to hoping the FilterLincs work magic! It sounds like you could use a few as well!

[ergodic]

A difficulty with Insteon is there isn't always an easy way to distinguish between ordinary comm and signal loss issues vs. more difficult firmware and device failure madness. The other diagnostic hurdle is that a single device failure can 'pollute' the whole network, masking the source of the problem.

 

I had a very similar problem to yours about 4 years ago. All of a sudden things just wouldn't talk reliably. It was cured after I replaced the PLM with one that had a newer (V91) rev. Though not before I drove myself half-insane taking off plates everywhere. But you've already tried that. It's possible your new PLM is faulty, but not likely.

 

I first recommend you try placing your PLM behind a Filterlinc, which will force it to communicate with RF only. This is how I run. This configuration often improves comm reliability.

 

Make 100% sure you don't have any V.35 switchlinc dimmers anywhere in your network.

 

You can get the free HouseLinc software, which has "ping"-type diagnostics http://www.smarthome.com/houselinc.html . This can help you to isolate the issue.

 

Is it possible something external nearby is injecting noise to your powerline? Powerline HF noise won't travel far so it would have to be something close (construction site, MRI scanner, Google data center, proton accelerator, etc.). If you get really desperate and have a scope you can construct a probe to take a look the zero-crossing and see how much noise there really is.

 

I'll also mention that I've not found that dual-band devices alone a decent substitute for a pair of APs on both power legs. But the newer devices may be better so possibly that's old thinking. I have quite a few dual-band devices now (about two dozen), but if I pull out either AP I still will get dead zones.