As regular readers will know, or anyone who has tried to do X10 control in a regular house environment, signal line noise is common, and can be incredibly difficult to track down and destroy! Since moving into this new house a year ago, I've spent literally months tracking down the reason why X10 signals simply stopped working at some times. The symptoms of this particular noise were more interesting that just stopping X10 signals. Lights would randomly turn themselves on and off throughout the day and night. I had to lock out my bedroom overhead lights each night or they would turn themselves on with some reliability and wake myself and more importantly my wife who was becoming less and less amused by the situation.
Even using one of these
I was unable for quite a long time to track down the source.
The procedure I used started with plugging that in and waiting for the large noise readings and then turning things
off around the house hoping that (as in the case of the CF bulb problems I would find
something that was generating the noise. I did not.
Step 2 was to go through the circuit breakers in the house 1 at a time until I localized the source to a specific circuit. One by one I switched off circuits and waited for the noise to return and each and every time it did. Though it appeared sometimes that I had switched off whatever was causing it, it would then be able to come back even with that circuit switched off. I began to think that the noise was coming in from outside the house.
Making one last desperate attempt to make some sense of it one day I had it plugged into a circuit on which the most lights were randomly turning on and off. I could quite plainly see the extreme levels of noise being dumped on the lines:

That second reading is .19 VOLTS! Thats 190 millivolts. A valid X10 signal can be under 30 millivolts, so you can imagine that this would cause havoc with the system! While playing around with things on that circuit I finally noticed something. The noise gradually ramped up with the furnace blower turning on. I verified this by turning the AC on and off a few times and indeed it was the FURNACE BLOWER!

This particular model (a Trane XV80) has a variable speed blower and a soft start and stop feature. Meaning that the motor is controlled by some kind of PWM or other system to control it's speed. And this feature seems to generate a HUGE amount of noise when the blower was running.
The reason that I was unable to localize the noise by going 1 breaker to another is that there are 3 of these furnaces in this house and they all run independently so there may have been more than one on at a time, or as soon as I thought I had found what it was, another one would click on. The reason that I was unable to correlate any specific lights turning on with the running of the furnace was to do with the nature of the effect the noise has on X10 units. It did not just make them come on when the noise started. The blower actually had to run for a minimum of 15 minutes or so before it would start to effect the lights. Even then, the noise would not make them come on, but the next X10 signal that was received would make several turn on, regardless of the address of the unit being spoken to. It left them in some state where they thought that they had received their address, even if they had not. So nothing actually correlated directly with the appearance of the noise, nor with the turning on of the specific heater/ac unit.
The fix has worked completely, and totally eliminated any measurable noise from the blower motors at all. I simply wired in 15 amp X10 noise filters in line with the furnace. Since there is no emergency heat and only the blower and electronics on the circuit, they are only 15 amps and a filter is reasonably priced for that load.

They also make a hard wired version for loads up to 20 amps if you have a larger blower that is causing a problem. I have had these in service for several months now without any difficulty. The house is once again functioning very well.
In the interest of saving some money on electricity and AC bills from running so many of those dangerous and cheap halogen tourcherie style lamps I purchased a GE, dimmable florescent model. It claimed to give all the light of a 300 watt incandescent but used only 30 watts. Even more important than a ten fold decrease in power usage was the 10 fold reduction in heat output! In these Southern climates anything you can do to decrease the amount of heat you're generating inside reduces your otherwise extravagant electric bill during the summer months. The lamp in question was very nice looking, high quality spun metal looking base and stand.


Unfortunately (or fortunately perhaps) the florescent color of the light was not appreciated by my family and the lamp saw little use, even after I complained that I spent $70 for it and that they had better get used to it! After experimenting with it a short time myself it was obvious that when this lamp was on, not a single X10 command could get through.
Since I had so much money invested in it and wanted to use it myself even if the rest of the family didn't want it I set about to try to filter it as well and see if I could make it usable. The first filter I added was a regular power line filter borrowed from an old PC power supply. It looked like this when I mounted it in a box and attached it to the cord of the evil lamp.

This made no difference to the noise at all. So I added a 5 amp X10 filter coil and still no change in the amount of noise. Finally I plugged this into my spare 15 amp filter as I had the furnaces. This finally started to put a dent into the noise and I thought that I would be able to use the lamp now, but as I let the lamp warm up and watched the signal analyzer I could see the levels increasing and increasing the longer I left the lamp on.
I never saw where it topped out, after about 30 minutes it was putting out more than 2 times the amount of noise the furnace had and this was through 3 different filters! The amount when not running through the filters was off the scale. So needless to say I cannot recommend these lamps for any use in a house with X10, or any other sensitive electronic equipment plugged in. There is no way to filter the amount of noise that is coming from this thing. This model is almost 3 years old now, so it's quite possible that they have fixed this ballast issue, or that this one is just defective. I have not purchased another to try it out. But if you've got one of these and are having X10 problems, unplugging this lamp is a great place to start troubleshooting!
UPDATE: 10/24/2004 Had another adventure recently with a hard to track down signal problem. It was very similar to the others in that after it had been going for a while, certain other units would respond randomly to other signals and turn on. So the worst of the symptoms, things like overhead bedroom lights turning on in the middle of the night, could happen hours after the noise had actually stopped and left them in a state where the next X10 ON that was received would turn several lights on.
I had the good luck one day of noticing that my evening script had run but only turned on about half the lights in the house that it was supposed to. So since the noise was active and blocking signals I spent the time to go around and see what was actually on at that time.
And it turns out to be the CF bulb that I had in a desk lamp over my workbench!

This bulb is a Consumer Electric brand purchased from Home Depot. I don't keep track of when I buy them but this had been in service in various places in the house for at least a year and never caused a problem. I had put it into a swing arm desk lamp several months before, but I don't use it every day, only when I'm soldering or other fine work at that desk. So the problem was very random. The light still works normally. It has not burned out, doesn't make any noise or flicker. From the outside it seems the same as all the other consumer electric lights I have in the house. But something has changed inside and it puts out so much noise that it can silence an entire phase of power in the house.
Overall I am very satisfied with the Consumer Electric CF bulbs. There are probably 30 of them of different size and shape in the house. The price is right, they are bright and I have never had a single one burn out yet or suffer the infant mortality that is so common with certain other brands. Also they are rated for use in enclosed fixtures unlike many others. They also switch fine on appliance modules and they turn on instantly with no flicker. This is the first one that I've had a problem with. So this is not a reason not to buy these otherwise very good bulbs. But if you have a problem, don't rule out the possibility that an otherwise normally functioning CF is the problem.
8/12/2005: the battle for a noise free powerline continues... I have now had 3 of those same 60 watt equivalent commercial electric bulbs start to generate enough noise to cancel out all X10 signals in the house. I've never had a problem with the 75 or 100 watt equivalent versions and the 85 watt floods and the new 95 watt equivalent floods have been excellent. I use many of the floods in the house, some running most of the day and evening every day and have yet to have a single failure or problem. But the 60 watt ones seem evil. At some random point in their life they start generating noise with no other obvious problems the light stays on, it's not flickering or anything else. But now that you all know about this you can try turning off any of these bulbs if you start having issues and check again.
Interestingly enough, it is only these 60 watt lamps that have been burning in a horizontal position here that have failed in this way. All the ones I have burning mostly base up or down are still going fine. It might be that the orientation causes heat problems inside the case? Or the electrodes to wear unevenly or something? I don't know if it's actually related to this or if it's just coincidence...
I recently replaced the old tired iMac in the kitchen that my daughter plays edumacational games on with an old powerbook of mine that has a broken screen. With an external monitor though it still works fine. At about that time I started again to have issues with X10 propagation. It turns out that powerbook and other switching computer power supplies can be very noisy. After plugging the powerbook into an X10 noise filter it's working fine again.
Lastly for this update, after purchasing a new APC ups for the office here I found that I wasn't getting but half of my signals through. Turns out that it's widely known that UPS's can be a "signal sucker" they don't make noise exactly, but rather just absorb the signals. Moving the X10 interface to a separate electrical circuit form the UPS has solved that problem.