Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[-Bill- (of wgjohns.com)]

Henry,
Assuming your A/C unit isn't the culprit, here's another theory.

You have a noise maker (CFL's or other) on one phase of your powerline.  Since you don't have a phase coupler (assumption?) much of this noise isn't being communicated to the opposite phase.  When your A/C kicks in, it effectively becomes a phase coupler and the noise begins causing problems.

I just noticed something from your original post:

...Recently I installed a CF outdoor light not on the remote control (uses light sensors for sundown/sunrise light control)...

Is the light sensor that you're referring to a X10 sensor or a built in (fixture mounted) photoelectric sensor?

Photoelectric sensors and CFL's do not play well together.  The output stage of the sensor (triac) can oscillate like mad with a CFL load.  I'm not sure that a filter will correct this.  Other forum members may have direct experience(??).

Good points all!   

[steven r]

Am I the only one that seems to remember a similar situation that turned out to I believe an electrostatic filter or something like that related to the furnace or am I just experiencing Deja vu. So far I haven't found it yet.

Ok I'm not losing my mind or at least you can't use this for proof... This is what I remembered. http://www.x10community.com/forums/index.php?topic=12125.0

[-Bill- (of wgjohns.com)]

Am I the only one that seems to remember a similar situation that turned out to I believe an electrostatic filter or something like that related to the furnace or am I just experiencing Deja vu. So far I haven't found it yet.

Ok I'm not losing my mind or at least you can't use this for proof... This is what I remembered. http://www.x10community.com/forums/index.php?topic=12125.0

Yes, I remember that one too.  I didn't reference it because he says he already disabled the air cleaner.  In retrospect, I probably should have.

Thanks for tracking it down steven r!   

[Boiler]

Am I the only one that seems to remember a similar situation that turned out to I believe an electrostatic filter or something like that related to the furnace or am I just experiencing Deja vu. So far I haven't found it yet.

Ok I'm not losing my mind or at least you can't use this for proof... This is what I remembered. http://www.x10community.com/forums/index.php?topic=12125.0

Steven, you had me really worried when I started reading the tread you referenced.  Electrostatic filters generating 300 MHz RF?? - I'm really very happy this turned was determined to not be the case (problem was traced to conducted powerline noise rather than RF).

I have a Honeywell F300B2012 electrostatic.  So far I have not have had not problems (RF or conducted).

[hmseiden]

The outdoor lights have sensors built into the fixtures. The bulbs that came with the fixture is a CFL (Compact FLourescent). I assume therefore that these should work OK without interference, and they have until recently. So that doesn't really correlate.

I scoped my neutral compared to ground and there is considerable interference (>200mV) on the neutral at least on one phase. It doesn't however interfere at all with X10 controller operation so I'm not sure if my interference pattern is normal, has always been present or what the source is.

With the AC running I see the same thing. I can see the X10 signals clearly on one neutral (one unfiltered wall plug) but not on others (with GFI). The lamps that are on X10 are not GFCI circuits and they all work. I have not scoped another circuit that corresponds to 'different phase' since the house is single split phase 220VAC incoming. Home wiring, not commercial. So I don't see correlation.

I have however looked for other X10 type signals during daytime hours and saw none other than mine.

[Boiler]

The outdoor lights have sensors built into the fixtures. The bulbs that came with the fixture is a CFL (Compact FLourescent). I assume therefore that these should work OK without interference, and they have until recently. So that doesn't really correlate.

I agree that if the product was sold with a CFL it should work without excessive noise.  It's possible that your CFL's were designed to work with photoelectric sensors or the sensor itself has a modified output stage.  That does not mean that they don't interfere with X10. 

As far as the system working previously - you've made changes.  What was working before may not work now.  I'm not sure what type of circuit the CFL manufacturers use to excite the bulb, but you can be sure it's low cost and probably subject to drift (both thermal and over life). Since these lights are on when your problems exhibit themselves, I'd consider disabling them as a test.

I scoped my neutral compared to ground and there is considerable interference (>200mV) on the neutral at least on one phase. It doesn't however interfere at all with X10 controller operation so I'm not sure if my interference pattern is normal, has always been present or what the source is.

When you say you're using a "Scope", are you referring to an oscilloscope or a X10 troubleshooting tool (ELK ES-M1)? 

I'm also confused by the statement that you are checking between "neutral and ground".  This may have been a simple typo on your part - if so ignore the following (I'm not trying to be insulting). 

Your switches operate between supply (black) and neutral (white).  Ground is used for safety (grounding fixtures, appliance chassis, etc) but your switches really aren't referenced to it.  Your ground line should not be carrying current unless you have a fault in the electrical system.  The neutral line is carrying all of the return current from all of the devices in your house (yes it will be noisy VS ground). 

If you are actually using an oscilloscope, you would need to monitor noise between supply and neutral.  Please don't attempt to measure across the 120V line unless you know exactly what you're doing (in which case you shouldn't be reading this anyway).

With the AC running I see the same thing. I can see the X10 signals clearly on one neutral (one unfiltered wall plug) but not on others (with GFI). The lamps that are on X10 are not GFCI circuits and they all work. I have not scoped another circuit that corresponds to 'different phase' since the house is single split phase 220VAC incoming. Home wiring, not commercial. So I don't see correlation.

I have however looked for other X10 type signals during daytime hours and saw none other than mine.

• A split phase 220V system has two 110V (hot black) phases (90^o apart) and a single neutral (white common).  220V is generated when you bridge across both hot phases (110 + 110).  When you bridge from a single Hot phase to neutral you get 110V.
• X10 signals are sent over the Hot side of your wiring system (black).  This is why you need a coupler to bridge the two phases of your electrical panel.
• If you were actually using an oscilloscope to monitor X10 activity between neutral and ground I wouldn't expect you to see much (if any) signal other than on the actual circuit sending the information.  Again - your monitoring the wrong side of the system.  The signals are on the hot side.  You can only see them on the neutral side because of line resistance and distributed inductance.
• I have many X10 modules on GFI circuits with no problems (as do other members of the forum).  I do not have experience with the new Arc fault interrupter.
• You've looked for extraneous X10 signals during daytime hours (signals from outside your house).  Unfortunately you need to look at the exact moment that your neighbor goes to the bathroom and activates his motion sensor.  Don't assume that the signals don't exist just because you can't see them at that moment.  If the signals are from outside your house, changing housecodes should at least change the problem.
  

Please don't take any of the above as a slap in the face.  You may understand all of the above and were testing from neutral to ground because it was safe and convenient.   

When troubleshooting problems like these you often need to think of multiple, additive problem sources.  I've added CFL's (more noise), a new AC unit (better coupling between phases?), the noise frequency of my outdoor CFL's has shifted (sum = invalid X10 reception).

[Brian H]

I have an outdoor CFL type fixture with a dark on sensor. It has a very small relay in it and not a solid state switch. I can hear a real small Click when it goes on.

[Boiler]

I have an outdoor CFL type fixture with a dark on sensor. It has a very small relay in it and not a solid state switch. I can hear a real small Click when it goes on.

Brian,
That makes a lot of sense.  It prevents the consumer from accidentally installing the incorrect type CFL and causing all kinds of problems.

[hmseiden]

It looks like it's got to be coming in from outside as it was on as something kicked the circuit (channel A4) I'm having trouble with on during daylight hours when the CFLs are not on and I know visually they are not!! More importantly another light on the same channel but a different phase did not come on!!!

I'm going to try replacing the switch module. Do you think it could simply have gone bad?

On another question, yes it was an oscilloscope I was using. The correct way to measure neutral to hot is to use two channels one neutral to ground (for safety and noise), and the other on the hot side to ground, then add the two channels together if your 'scope has these options. Mine does. But I see the signal even on the neutral side is what I was trying to relate. It is, of course, there on the hot side also.

Attaching a scope ground to neutral may be dangerous and/or affect your measurements. If your scope is ungrounded, there is even greater risk. if wiring is wrong in your house you become the ground path if you touch metallic portions of the scope frame. Be careful out there...

Yes, thanks Boiler, I do have a bit of knowledge in this area. Unfortunately I don't have one of the fancy meters that can chart the problem.

[Boiler]

It looks like it's got to be coming in from outside as it was on as something kicked the circuit (channel A4) I'm having trouble with on during daylight hours when the CFLs are not on and I know visually they are not!! More importantly another light on the same channel but a different phase did not come on!!!

I'm going to try replacing the switch module. Do you think it could simply have gone bad?

OK, you've eliminated your CFL's and the electrostatic filter from the list of offenders.

If I understand the above correctly -

• You have lamps activating (un-commanded) during the day on one phase.
• Lamps on the same house/unit code on the other phase do not activate.
• Your CFL's are off at this time

Assuming that your TM751 reliably activates these lamps when you use a RF command - it should not be the culprit.  If you were receiving external RF I would expect the TM751 to drive both phases and turn all your lights on (as it does when you command it).

I have a hard time believing that you have a switch module that has failed in a manner where it turns on with spurious noise.  I suppose you could have a capacitor breaking down (I haven't seen a schematic - do they use electrolytic's in these?) causing a loss in noise immunity.  If replacing the switch corrects the problem, please  let us know.

The other option is conducted X10.  An X10 signal coming through your supply transformer will likely be low level and could activate selected modules (depending on individual line loading) as you are observing.  This could easily be an address A all-on command from your neighbor which would explain why multiple addresses are activated.

Try changing your house code.  I like D, J and K - not often used from what I've seen.

On another question, yes it was an oscilloscope I was using. The correct way to measure neutral to hot is to use two channels one neutral to ground (for safety and noise), and the other on the hot side to ground, then add the two channels together if your 'scope has these options. Mine does. But I see the signal even on the hot side is what I was trying to relate. It is of course there on the hot side also but not on the ground.

Attaching a scope ground to neutral may be dangerous and/or affect your measurements. If your scope is ungrounded, there is even greater risk, if wiring is wrong in your house you become the ground path if you touch metallic portions of the scope frame.

Excellent - happy to meet another sparky - sorry for the lecture.  I use an isolation transformer to measure across the line (I'm old and visually challenged - don't want to worry about polarities).

Be careful out there...

Hill Street Blues?

Unfortunately I don't have one of the fancy meters that can chart the problem.

If you are considering expanding your system in the future (might not be a good time to talk about this) you really might consider one.  Scopes are great tools, but you can't use one to decode an X10 transmission or see if a particular bit was "stepped on". 

[hmseiden]

Summary of the effect I see:

• Certain channels on 'A' House Code, and only one light switch on that channel comes on all by itself. Daytime and night-time occurrences but most often over night between 3-6AM, exact times vary day-to-day.
• Always the same switch and the channel # changed on the offending switch from A5->A4 makes no difference to the problem. That one in-wall switch goes on.
• It happens at all hours, even when dusk-dawn light controlled CFL outdoor lamps are off.
• TM751 controls any of the lights including this one affected switch perfectly.

Just thought of something else that's possible and seems to fit my findings. Maybe pulling at straws but what do you think? Although I don't know for sure, here's my (new, latest, greatest) theory:

• FPL (my local utility) has a program to control major devices and save electricity (it's called Power Saver, or somesuch). You put in their system and it switches these appliances according to it's need and you never see the effect of them shutting off your hot water heater or the AC for an hour at 5AM(!).
• What if they use the same technology as X10 to control the relay devices on these appliances? The box they put on the outside of the house probably uses data sent via power lines. The relay boxes on the appliance might use X10-like commands.
• One of my neighbors is a technology geek. Maybe he put it in or maybe his parents did and he knows about it. There could be other neighbors that have this. I only speak of this next door neighbor, because he made some major renovations when he moved in a year or two ago. However this would have to be more recent because it's only been weeks since the trouble began.

I am going to put the house over to a new house code, if I can get the problem to quit. I'll first start with this switch. It's making me nuts!



Henry

[steven r]

...Always the same switch and the channel # changed on the offending switch from A5->A4 makes no difference to the problem. That one in-wall switch goes on. [/li]...

Skimming back through the discussion before tossing in another 2 cents of mine, I wasn't sure if you had tried swapping the switch. i.e. Exchanging it for one that was working on a different circuit and seeing if the problem swaps places. Also try changing some of you other switches in the house to "A5" and see if they come on.

...(my local utility) has a program to control major devices and save electricity (it's called Power Saver...
...What if they use the same technology as X10 to control the relay devices on these appliances?...

I have that service on my AC. In addition to the few bucks a month they credit me, I've never had a problem with it interfering with X10. I believe they use a far more dependable signal than the X10 protocol.

[hmseiden]

Tried swapping other switches to A5, in fact I also tried (and mentioned early in this thread) swapping the switch to A4 and that still has another switch on the same channel (but possible on a different leg of 120VAC) which is sorta what you recommended. The latest move was to swap the offending switch to a different house code 'O4' (that's the latter 'O') and that also triggered falsely.

So I'm back to my original premise that it's the switch as no other switch has turned on mysteriously lately, although they did in the recent past. On the new channel, when the 751 controller is changed to appropriate house code, the offending switch works OK. But it continues to trigger on by itself for no apparent reason.

Thanks for all your suggestions guys. I do now have another switch and will try that.

[Noam]

A few years ago, I had a problem with a wall switch turning on by itself, with no commands showing up in the logs for it.
I found this online about it, and how to fix it: http://www.idobartana.com/hakb/modifying_wall_switch.htm#Fixing module that randomly turns on
In the end, I decided not to fix it, but I replaced it instead.
From what I've read, there is a history of these switches going bad. The Smarthome ones seem to have a better track record.

--Noam

[steven r]

Just to rule it out...

Do you have any macros running when the lights come on? If you do, do any of them have dim commands? AHP has a know bug of creating ghost on commands when a dim command is immediately followed by a non dim command. The delay command is kinda like fairy dust to work around this bug. Simply follow the dim command by a 2 or 3 second delay.

For what it's worth, C5 keep coming on for me till I added a delay. Since my C5 was a guest bedroom light, I didn't discover it till my mom was visiting and stayed in the "haunted room".