Lights Flicker caused by Insteon Dimmer switches

[Krusty66]

I'll start by saying that I have a lot of experience with Insteon products and have been running them via ISY in my house for several years.  The issue I'm now having has been observed by an electrician and an Electrical Engineer.  The E.E. also uses an ISY/Insteon setup in his house and has hundreds of insteon devices, including the ones I'll mention, in use.  Nobody has any explanation other than what I'll put forward -

First, my system uses the ISY 994i with the PLM and I have dozens of insteon devices and some pretty simple programs running (turn on lights at dusk, turn on a light if Motion sensor is tripped, etc.).  I use many of the insteon Plug In Lamp Dimmers (2457), Plug-in On/off (2634, 2635), many of the mini-remotes (2432), a few motion sensors, some of the bulbs (2672), a couple Toggle Dimmers (2466).  I have never had any sort of issue like what I will describe here:

I installed a 2334-222 Keypad Dimmer switch in my bedroom.  It controls nothing but a single incandescent bulb overhead.  Shortly after installing, I noticed that the bulb in question had some flicker going on. In short, it was a randomly timed, quick duration flicker and it appears that its not just an attenuation of the power.  It's as though there is a surge sometimes (making the light brighter than normal) along with a trough (making light dimmer), then a return to normal.  Sometimes just a trough, sometimes just a surge, sometimes both.  Always in a split second and always returning the light back to normal.  It could happen a few times in 10 seconds, or it could go a minute without happening.  But rest assured, over a 5 minute period, it would happen several, or many,  times.   Went thru all the normal procedures to address such a problem and the flicker would not stop.  Eventually yanked the switch and went back to the passive cheapo switch that was in there before.  Flicker went away, just like it had been for the 17 years I've lived here.  Chalked it up to a bad insteon switch.

New part of the house - installed a 2477D Paddle Dimmer for my back patio (4 bulbs).  Again with the flicker.  Much testing, much observation, no solution.  Could not stop the flicker.

Now the biggie - doing a massive renovation on the basement and already had purchased several paddle dimmers (2477) to control 4 separate banks of ceiling lights.  This is brand new wiring, direct from the breaker box 6 feet away.  My house is old (1949), but this is all brand new equipment (can lights, wiring, circuit breakers, etc.).  Long story short - exact same flicker problem as seen in bedroom and patio.

Like I said, my electrician and E.E. were both stumped.  I, too, was stumped.  So here's what I did to analyze it, and bear in mind this took many, many hours and I am still in process.  I've got at least 10 to 12 hours in this so far...I turned off every breaker in the box except one in the basement with a bank of lights and a 2477 dimmer paddle controlling them.  There is nothing else on the circuit.  In this state, no flickering.  That is, the circuit by itself, with nothing else in the house powered up, does not produce the flicker.  I then brought up other circuits in the house one at a time then watched for flicker.  Brought up the bedroom circuit, sat and watched for the basement lights to start flickering...nothing.  All clear, so bring up another circuit.  Turned on the dining room circuit, sat and watched for flickering...nothing.  And on and on until I bring up a circuit and observe the return of the flicker.  And sure enough, the kitchen (by itself) and the living room (by itself) being on at the breaker will cause the flickering on my circuit in the basement.  Let me re-iterate to be clear - I can take down every single breaker in the house except the one with my basement lights and there will be no flicker on those basement lights.  If I then turn on my kitchen breaker, the basement light flicker happens.  Ditto the living room breaker.  When those two are shut down, I can turn on every other breaker in the house and the flicker still does not happen.

So clearly we have a situation where (at least) one device in each of the two offending circuits (kitchen and Living room) is causing some kind of havoc with the electricity in the entire house.  And whatever that havoc is, insteon paddle and keypad dimmers are sensitive to it and cause a flicker to happen on the bulbs to which they are carrying power.  The flicker can even be seen on the dimmers themselves (the LED lights and lit buttons will flicker in unison with the lights).  That said, the other devices (plug in dimmers and toggle dimmers) are not sensitive to this issue.  Again, I have been using those latter devices for years and have never seen any kind of flicker.  The paddle and keypad dimmers take one minute of observation and the flicker is clearly there and enough to drive one insane.

Continuing my process, I've observed that when I remove certain devices from the offending circuits, the problems decrease.  The first one was a 24 volt power supply for some LED light strips in the kitchen.   Just to be clear again, took down the entire house except basement lights and kitchen, observed flicker happening, removed 24v power supply, flicker stops.  Now, I wish I could say that its just that device, but we already know there are more in the living room.  That, of course, contains A/V equipment, telecom stuff, fish tanks, UPS battery backup devices, you name it.  Surely many of these devices are part of this problem.  And it is clear that there is a cumulative effect.  That is, when the whole house is up and running, the flicker is more frequent and more violent.  As I remove devices, things get better.  But as of now, until I remove almost every damn thing in that living room, the problem is intolerable.

My E.E. tells me that it certainly sounds like I've tracked down the problem and it seems reasonable that cheap Chinese devices like these could be causing the havoc.  What I'd like to know from the forum is - 1) has anyone observed this problem?, 2) why do only the keypad and paddle dimmers do this?.  They are obviously using some different (and inferior) circuitry to the plug in modules, and 3) is there a solution?  My E.E. suggests using some FilterLinc devices, which I have ordered and will try. 

[Teken]

1. Seen this all the time in various degrees.

2. From experience you would see similar results with other brands and dimmers but not a relay.

3. Replace all of the noise makers in the home.

4. Filter Linc is a 50/50 chance so don't bet the bank.


Sent from my iPhone using Tapatalk

[Krusty66]

Replace all what noisemakers?  My TV? My refrigerator? Everything with a power supply or transformer?  What, in your experience, are the noise makers?

[Brian H]

UPS units have a power line conditioner in them. Many will absorb the Insteon power line signals. Even if is not a bad noise maker. You may want to FilterLinc it any way. Unless your equipment uses over its 10 amp fuse rating. There are still a few 15 amp X10 ones {FilterLinc was originally an X10 one} still made that can be used.

I suspect the power supplies in the dimmer switches and plug in modules are a different design. Both look like small switcher. The plug in are 120V 60Hz and the wall switches 100V-277V. 50/60Hz. That may contribute to the level of sensitivity to power line noise. Between a plug in and a wired in wall dimmer.

 

[mwester]

ANYTHING with a switching power supply (which is basically anything manufactured in the last couple of decades) can be a noise source.  Some are less, some  (often the cheapo junk devices) are more so.  Size or power-draw is irrelevant in my experience, although quality of the device is an indicator -- I've observed that a $9 Chinese knock-off phone charger was capable of taking every Insteon device on the first floor effectively out-of-service.

Insteon's protocol was designed back when transformer-based power-supplies were the norm, and it has not scaled well in today's world, particularly given the "race to the bottom" in terms of price for electronic gizmos.

Filterlincs will help -- I have my house littered with filterlincs.  They sprout on every wall, often connected to power strips so I can use one expensive filterlinc to cover several devices -- which results in the house also being littered with extension cords and power strips, despite have plenty of unoccupied wall outlets everywhere.  For a few troublesome wired devices, I use the 20A X10 in-line filter.

Despite all those efforts, I still have marginal comms.  But at least no flickering of lights.

 

My advice:  switch to Z-Wave.  Works well with the ISY, and unlike the Insteon protocol (and I fear, the company itself), Z-Wave has a future.  It's not perfect, and not a panacea, but once you have a "critical mass" of always-on devices to form the mesh, it's reliable and just works, and far faster than Insteon, plus it has a lot more "cool" devices.

[KSchex]

Would like to know what Insteon says the problem is.  My guess is the output circuit driving the light is of a poor design.  It probably uses a triac device or similar that is being turned on an off (modulated if you will) by the line noise.  My guess is that later manufactured devices have a more stable circuit with some minimal filtering.

Getting rid of the noise makers (switching power supplies, etc.)  or adding filters in your home is a solution but in todays digital age this may be impossible.  Nobody uses good old fashion analog power supplies any more or worries about harmonic distortion of your home power.  Cost effective designs that are quick to market is whats important.

It blows me away how a company can sell a device on such a grand scale that is so susceptible to power line noise, hampering the output and/or communication.  Sorry, just had to add my 2 cents to a problem I have spent days working on.  One reason I switched to Z Wave. 

[Krusty66]

Great info from everyone.  I can't just switch to z wave at this point, but I will try the filters this week.  It is more than a bit disappointing that Insteon has a formula for a circuit that is not susceptible to this problem (in the lamplinc and togglelinc units), but hasn't bothered to fix the issue in the keypad and paddle dimmers.  Making it all the more annoying is that those latter units are more expensive and more difficult to install (than the lamplinc).  You make a much bigger commitment to them (in money, and more importantly, time), just to find out they're inferior.  I'm also a bit surprised this issue isn't documented or mentioned anywhere.  Look it up on this forum or in any product reviews and you won't find it anywhere (until now).

[Teken]

36 minutes ago, Krusty66 said:

Great info from everyone.  I can't just switch to z wave at this point, but I will try the filters this week.  It is more than a bit disappointing that Insteon has a formula for a circuit that is not susceptible to this problem (in the lamplinc and togglelinc units), but hasn't bothered to fix the issue in the keypad and paddle dimmers.  Making it all the more annoying is that those latter units are more expensive and more difficult to install (than the lamplinc).  You make a much bigger commitment to them (in money, and more importantly, time), just to find out they're inferior.  I'm also a bit surprised this issue isn't documented or mentioned anywhere.  Look it up on this forum or in any product reviews and you won't find it anywhere (until now).

Line noise has been present since electricity was invented.

Anyone who has used X-10 (The God Father) of power line home automation is very aware how power line noise can impact their systems. Anyone who uses power line from X-10, UPB, Insteon, Power Over Ethernet, etc is aware of the very same.

The long term solution is to identify all noise makers / signal suckers and remove, replace, filter, isolate.

 

[Brian H]

I have seen reports that even the higher end Lutron  manual dimmers can flicker on line noise and certain model bulbs.

They make a AC input filter to remove the incoming noise from the manual dimmer. Cost in line with the Lutron name and too expensive for my tastes.

 

[lilyoyo1]

54 minutes ago, Krusty66 said:

Great info from everyone.  I can't just switch to z wave at this point, but I will try the filters this week.  It is more than a bit disappointing that Insteon has a formula for a circuit that is not susceptible to this problem (in the lamplinc and togglelinc units), but hasn't bothered to fix the issue in the keypad and paddle dimmers.  Making it all the more annoying is that those latter units are more expensive and more difficult to install (than the lamplinc).  You make a much bigger commitment to them (in money, and more importantly, time), just to find out they're inferior.  I'm also a bit surprised this issue isn't documented or mentioned anywhere.  Look it up on this forum or in any product reviews and you won't find it anywhere (until now).

The funny thing with noise is how it can effect things. You can have the exact same switches side by side and one has issues while the other doesn't. I wouldn't say they are inferior as I've seen many more issues in togglelincs than in the kpls and 2477d's. 

This is one reason why I wouldn't tell someone to switch from insteon to zwave or vice versa. Using one or the other doesn't guarantee a trouble free system. Most likely you don't have a lot of things causing issues. Since you've isolated what room is causing the problem, I would start with unplugging things there until you find the culprit. Once found, put a filter on it. 

[Krusty66]

Unfortunately, its not just a matter of finding one culprit.  I can see there are at least 5 different culprits.  And there might be more, and every time I bring a new device into the house I have look out for that, too.  I can filter the heck out of everything and hopefully it'll work. 

It seems that this issue should be present in the vast majority of applications.  My house is 2000 sf with just two adults.  I don't have kids (with their dozens of devices) or a large house (which would equal more potential culprits).  It would seem that virtually everyone should have this problem.  Who doesn't have a million wall wart adapters and power supplies all over their house these days?  So I can't help but wonder why my E.E. friend has done massive Insteon installations in 3 different houses now, has never had this issue. He doesn't have filterlincs all over the place and he has many, many paddle dimmers.  It sounds like there is some other variable in play, but I don't know what it could be.

[lilyoyo1]

Every environment is different which is why he doesn't have issues. Wall worts may not be the culprit. When people give tips they start with the most likely and go from there. It could be a single wall wort. Maybe 2 or 3. Maybe it's something else entirely like a wiring issue.

Now that you've isolated the problem to the circuit, you simply go a little bit further to find the actual culprit.

[Teken]

9 minutes ago, lilyoyo1 said:

Every environment is different which is why he doesn't have issues. Wall worts may not be the culprit. When people give tips they start with the most likely and go from there. It could be a single wall wort. Maybe 2 or 3. Maybe it's something else entirely like a wiring issue.

Now that you've isolated the problem to the circuit, you simply go a little bit further to find the actual culprit.

^ Spot On . . .

Of the hundreds of installs that I have been involved in only a hand full required any filters. Those that did needed only a couple shared filters via power strip. As noted up top this is easy accomplished when a home is in the design and building stage.

- Dedicated outlet / circuit for known noise makers - signal suckers: Fridge, Freezer, Sump, HVAC, etc

- Dedicated outlet / circuit for the 2413S PLM

- Purchase and test all lighting loads; CFL, LED, Magnetic Ballest, Florescent with the Insteon hardware.

- Wiring noise makers / signal suckers on opposing electrical leg and closest to service feed.

- Install a hardwired phase coupler at the service panel.

[Krusty66]

2 hours ago, lilyoyo1 said:

. It could be a single wall wort. Maybe 2 or 3. Maybe it's something else entirely like a wiring issue.

Yes, it could be something other than wall warts, or any other device and perhaps a wiring issue.  That was point exactly.  I was looking for the possibility that there is something specific to my wiring or electrical service and if that was the case, what could be done about that at a macro, rather than micro, level.  If there is something about the wiring and the electrical service at the 3 houses that had no issues, vs. my house, I'd sure like to know what it is.

[Krusty66]

Drum roll please...Two FilterLinc devices arrived here an hour ago and I've been testing with them since then.  They will indeed solve this problem.  I put my entire A/V system on one and my entire Telecom closet on the other.  Then killed the known culprit in the kitchen (the 24v supply for LED lighting), and absolutely no flickering at all.  So there's the answer.  Much thanks to everyone that chimed in.  I hope this helps future users.

[lilyoyo1]

41 minutes ago, Krusty66 said:

Yes, it could be something other than wall warts, or any other device and perhaps a wiring issue.  That was point exactly.  I was looking for the possibility that there is something specific to my wiring or electrical service and if that was the case, what could be done about that at a macro, rather than micro, level.  If there is something about the wiring and the electrical service at the 3 houses that had no issues, vs. my house, I'd sure like to know what it is.

Even if it was something big like a panel issue, the smaller things could still be part of the problem. It's cheaper to solve the micro problems than trying to fix a major problem that may or may not be there. 

[Brian H]

Glad you found the issues.

Thanks for keeping us informed.

 

[widnerm]

8 hours ago, Krusty66 said:

So I can't help but wonder why my E.E. friend has done massive Insteon installations in 3 different houses now, has never had this issue. He doesn't have filterlincs all over the place and he has many, many paddle dimmers.  It sounds like there is some other variable in play, but I don't know what it could be.

As that referenced E.E. friend, I will say that I have learned two important lessons over the years that I strive to pass along to others whenever possible to save them from heartache, confusion, and endless torment.

1) you can't mail a letter without a stamp (except when it says "no postage necessary if mailed in the US")

2) filterlincs are the Michael Cohens of Insteon communication problems:  you send them to take care of the noise you brought on yourself and you don't ask how or why.

Edited May 8, 2018 by widnerm

[Krusty66]

I originally thought FilterLincs were the Leonard Cohen of Insteon products.  In that they were ridiculously over-rated one-trick-ponies.  But I was wrong.  And I think on that note we wrap up this thread.

[Teken]

10 hours ago, widnerm said:

As that referenced E.E. friend, I will say that I have learned two important lessons over the years that I strive to pass along to others whenever possible to save them from heartache, confusion, and endless torment.

1) you can't mail a letter without a stamp (except when it says "no postage necessary if mailed in the US")

2) filterlincs are the Michael Cohens of Insteon communication problems:  you send them to take care of the noise you brought on yourself and you don't ask how or why.

[larryllix]

11 hours ago, Krusty66 said:

I originally thought FilterLincs were the Leonard Cohen of Insteon products.  In that they were ridiculously over-rated one-trick-ponies.  But I was wrong.  And I think on that note we wrap up this thread.

Just reading your OP and this sounds like you have a tuned circuit on that leg that the noise being generated (somewhere) causes a resonance at that tuned frequency. I worked in metering for years and we found certain customers at the end of 14kV lines would blow high voltage fuses on the metering transformers twice per year. Once a new high-voltage customer was added to the end of the line the problem would typically go away. Some electrical noise will resonate in certain length of line and load combinations that tend to amplify the noise and it will focus on one spot as seen by standing wave demo tanks where the reflected waves synchronise in one particular place. ie. where the amplitudes become additive.

As an experiment you could try adding a temporary piece of NMD 14/2 cable onto the end of the circuit (hanging in the air or on the floor) to see if you can change the tuning of this circuit. Alternatively plug in a heavy load on a receptacle in the same circuit and see if you can change the whole dynamic.

 

Good luck. It sounds like your troubleshooting is better than your EE and electrician.

[Krusty66]

My E.E. is obviously a complete bozo as evidenced by his nonsense response about stamps, or whatever the hell he's talking about.  What you're suggesting here in terms of testing for a tuned circuit is exactly what I was looking for.  The problem has gone away using the filterlincs, but there's obviously something different with those of us that have this issue and those that don't.  Your explanation may be the answer, or at least part of it.

[dsstrainer]

I feel it's sometimes more the bulb/dimmer fault than it being an insteon problem. I have some LED bulbs from Homedepot in my basement that flicker with a normal wall (non-smart) dimmer, even when set to the highest position. I recently moved some of those bulbs into 2 lamps upstairs that are on insteon dimmers, but still set to 100%. I see the same flicker. So It's the bulb.. not the insteon.

[larryllix]

19 minutes ago, dsstrainer said:

I feel it's sometimes more the bulb/dimmer fault than it being an insteon problem. I have some LED bulbs from Homedepot in my basement that flicker with a normal wall (non-smart) dimmer, even when set to the highest position. I recently moved some of those bulbs into 2 lamps upstairs that are on insteon dimmers, but still set to 100%. I see the same flicker. So It's the bulb.. not the insteon.

OP stated "incandescent" but Yes. LEDs can be a huge problem with dimmers. I think we are hearing less and less of that with newer bulbs here though.

Edited May 8, 2018 by larryllix

[larryllix]

1 hour ago, Krusty66 said:

My E.E. is obviously a complete bozo as evidenced by his nonsense response about stamps, or whatever the hell he's talking about.  What you're suggesting here in terms of testing for a tuned circuit is exactly what I was looking for.  The problem has gone away using the filterlincs, but there's obviously something different with those of us that have this issue and those that don't.  Your explanation may be the answer, or at least part of it.

We always tried to keep the design EEs in the closet. Usually two different coloured pencils could keep them out of trouble, Everytime they got out there was trouble and we had to call the EE control to net them again.  To be fair, troubleshooting is not their training or experience, typically and we relied on some of them to haggle  with management financial types over what colour RAM should be installed in the new systems. (Hint! The blue ones have slipperier electrons, making them much higher speed)  

IIRC Insteon was designed around dimmer noise. Based on X10 they changed the signal from the front porch after the powerline zero crossing to the back porch. This was to avoid the dimmer noise on the voltage rise up where dimmer switches switch, and onto the voltage drop down just before the zero crossing where most powerline waveforms are quieter.

This has it's own  challenges. ie. The sinewave frequency has to be known (N.Am. vs Oz. and Euro, accent dependant :)) and it has to be timed from the previous zero crossing.  2nd harmonics can shift the zero crossing making timing problems. These are huge in HID lighting (sodium vapour, mercury vapour).   3rd harmonics can do the same but can't remember where you would get them anymore, except from electronic powers supplies. Switching powers supplies can produce all different problems depending on how they rough control the higher voltage input.

In the end the simulcast broadcast repeat has it's own problems. They all have to be synchronised to not cause cancellations and if each Insteon repeating device has a different zero crossing distortion device near it, they can make their own problems with each other.

 

 

ahhh new feature! Mesh lock. When you get one that worked, press the button is an S-O-S pattern.

 

Edited May 8, 2018 by larryllix