X10 reliability is always at risk

[bkenobi]

I've had my main X10 system unchanged for over a year without too much drama.  Over the summer I noticed that one of my outside lights was not responding reliably and assumed I'd need to replace the module at some point.  The module is a XPDF inline dimmer module that has no local control and is located in an attic off the side of a roof gable.  There's an access right next to it, so I was planning on fixing it soon as the days get shorter and I'll need it more.  Turns out, when I went to use it this time, it was working again.  I did some looking and the light is on the same circuit as the outlets in the game room below.  That room has a screen door on one side and my wife used a fan to pull air in after the outside temp dropped over the last few months.  As soon as the fan went away, the light worked correctly again.  So even a simple fan can cause X10 issues!

After thinking about it, I suppose this is a bit more predictable than initially thought as the fan has a remote control and a timer.  That means that rather than a simple mechanical switch, it has some kind of always-on IR/RF receiver as well as some kind of basic processor inside to control various functions.  This is an easy fix by simply plugging the fan into a filter instead of directly into the wall.  But it does go to show that even simple things can cause unforseen issues even for those who have been using this tech for a while!   

[Brian H]

Thank you for the information.
One other thought. Since it was always powered up to run the remote circuits.
The power supply may have had a AC rated capacitor directly across the power lines. To keep internal electrical noise from going out on the power lines. That could also absorb X10 power line signals as noise.

[brobin]

I have a weird one that just cropped up too starting earlier this week.  I have appliance modules on my router, cable modem and hot water recirc pump. Twice this week, for no apparent reason, one or more of those modules turned themselves off.  My Stargate shows no sent or received commands when this happens so the status shows that they're still on even when they're not.  The modules are C6, C15 & C16 and none of the other modules on the C (or any other) house code are doing this.  After using X10 for close to 40 years I thought I'd seen it all. I'll do some more investigating over the weekend.

[toasterking]

The module is a XPDF inline dimmer module

I've seen the same thing over the years, specifically with the XPDF.  As they age, they require a stronger X10 signal to respond, and stop responding intermittently sometimes, then resume later.  I haven't always been able to track down any related event, and I haven't bothered opening one up to try to determine the failure.  But I don't think you're the only one with intermittent XPDF issues.

[bkenobi]

I've noticed that sometimes that unit stays on when commanded off. I usually control it with a 4 button wall switch (XPT) by turning off units 1-3 in a row. I've always assumed the reason it stayed on was pushing the buttons too fast and crashing the last signal. I switched to a XPT4 module with indication LED's since i can't actually see the light which mostly solves the issue.

[toasterking]

pushing the buttons too fast and crashing the last signal

I've never used the original XPT (with the tiny code knobs), but FWIW, I've never been able to crash the XPT version A (with the standard code wheels which can be rotated with a dime); it is impressively resilient.  My testing shows that it buffers commands it is transmitting and implements both collision avoidance and collision detection/retransmission.  I had two installed side by side and could smash all the buttons on both of them at once.  All of the commands would eventually get through, though not always in the expected order.  I could also transmit a command from an impolite transmitter during that sequence and the XPT v.A transmitters would retry until all commands got through.  I don't know if any of this applies to the original XPT or the XPT4.  BTW, the Leviton keypads with status LEDs could also (unofficially) be used with the XPT v.A.

When I had XPDFs misbehaving, a check with an XTBM-Pro always revealed that the XPT v.A transmitters were doing their jobs, if a little off on signal frequency (they averaged around 124 kHz).  Though after 9-10 years of service, my XPT v.A transmitters did eventually start dying.  (One day working; the next day completely dead.)

[Noam]

I've had my main X10 system unchanged for over a year without too much drama.  Over the summer I noticed that one of my outside lights was not responding reliably and assumed I'd need to replace the module at some point.  The module is a XPDF inline dimmer module that has no local control and is located in an attic off the side of a roof gable.  There's an access right next to it, so I was planning on fixing it soon as the days get shorter and I'll need it more.  Turns out, when I went to use it this time, it was working again.  I did some looking and the light is on the same circuit as the outlets in the game room below.  That room has a screen door on one side and my wife used a fan to pull air in after the outside temp dropped over the last few months.  As soon as the fan went away, the light worked correctly again.  So even a simple fan can cause X10 issues!

After thinking about it, I suppose this is a bit more predictable than initially thought as the fan has a remote control and a timer.  That means that rather than a simple mechanical switch, it has some kind of always-on IR/RF receiver as well as some kind of basic processor inside to control various functions.  This is an easy fix by simply plugging the fan into a filter instead of directly into the wall.  But it does go to show that even simple things can cause unforseen issues even for those who have been using this tech for a while!   

Is it time for me to repeat my story about how the CFL across the street (my neighbor's porch light) would shut down half my system every night (*even with* my XTB-IIR)? 

[bkenobi]

Mine is a version A with a Leviton panel. I never tried watching the XPT output with my XTBR but that would be a good idea. The XPDF is on the same panel but different run as the XPT. I do have a few things plugged into that room/leg but everything of consequence is on filters.

[JeffVolp]

Though after 9-10 years of service, my XPT v.A transmitters did eventually start dying.  (One day working; the next day completely dead.)

We have four of the Leviton version with the LED indicators in our home.  After a dozen or so years one started being cranky issuing commands.  They weren't being picked up reliably by the XTBR.  It eventually failed completely.  The LED indicators still worked, so power supply was OK.  It turned out the film capacitor that couples the transmitter to the powerline had degraded.  It has been working fine again after I replaced that capacitor.

Jeff

[toasterking]

Though after 9-10 years of service, my XPT v.A transmitters did eventually start dying.  (One day working; the next day completely dead.)

The LED indicators still worked, so power supply was OK.

Thanks for the tip.  I had 3 fail, but my failure mode was different as the LED indicators stopped working also.

[bkenobi]

I still have my X10 version of the panel so if XTBR suggests issues i could swap back for testing. If the cap is faulty is it something that a novice could repair? Most X10 stuff is through hole but newer stuff is SM which is trickier.

[JeffVolp]

If the cap is faulty is it something that a novice could repair? Most X10 stuff is through hole but newer stuff is SM which is trickier.

Mine are the through-hole version before X10 began the transition to surface mount.

Jeff

[bkenobi]

I dug out my original XPT push button plate.  The currently installed one doesn't reliably trigger button 3 off anymore.  Are the capacitors you replaced like the ones on this one?

[bkenobi]

The previ[us image was from the X10 plate. This one is for the Leviton unit which has the LED indicators.  I dont use button 4 so I'm not sure if the issue is overuse of 4 off or one of the other components but changing the capacitor looks easy enough to try.

[JeffVolp]

I dug out my original XPT push button plate.  The currently installed one doesn't reliably trigger button 3 off anymore.  Are the capacitors you replaced like the ones on this one?

This was several years ago, but it was a large film capacitor that connected directly to the AC input.  It was probably either .1uF or .22uF.

More recently another Leviton 4-button keypad failed.  Initially one of the LEDs was dimmer than the others.  Then after a week or so the LEDs started flickering and none of the buttons worked anymore.  I thought an electrolytic had deteriorated, but the DC power supply seemed fine with no significant ripple and power to the chip was fine.  So I don't know what caused the failure.  I just swapped in a spare unit.

Jeff