I'm seeing a nearly continuous 2Vpp 120kHz signal on one 'phase'. The other 'phase' is clear. There are what resemble triac noise bursts approximately
1.6ms after each zero crossing and the 120kHz seems to be in sync with them.
I say nearly continuous because the signal is absent on what appear to be random half cycles (again in sync with the triac-like bursts, 1.6ms past ZC).
I say 'phase' because while the two circuits seem to have no communication, they are in phase when observed on a scope. My apartment was originally two apartments and there are two meters which might be a factor.
The signal is not coming from anything in my place but there are 4 apartments in the building and 4 more in another building across the street which shares the transformer.
A random snapshot of 24 half cycles is 101101101110111110110110 with 1 indicating 120kHz presence and 0 its absence. I've been unable to discern any clear pattern other than the OFFs never last beyond a single half cycle.
The triac noise pulses have been here for a few years but yesterday was the first time I saw the nearly continuous signal. It blocks X-10. It shows up as noise on my ESM1.
Can you determine whether the signal gets stronger as you get closer to the breaker panel? It's probably an external signal, but it would be nice to know for sure.
Don't baby monitors and intercoms that work via powerlines put out similar interference? I would assume that an apartment is much more likely to have "neighbor issues" then single family homes just because units are closer together. Are there any hallway outlets that you could check to get a signal level? You might be able to triangulate in on the source that way. Will your neighbors let you poke around their places with an ESM1?
This brings up an interesting question. How can an apartment dweller block such signals (assuming they are external) easily without inducing the landlord to modify the breaker panel?
Have you checked to see if any of your experimental setups are still on line? Is it possible that some piece of X-10 gear has gone wild and is just sending out random pulses?
My walker doesn't handle stairs or any but the slightest of grades so...
Even if I could get close to the (basement) breaker panel, signal attenuators can cause misleading measurements.
I can't navigate to the neighboring apartments. They are upstairs/downstairs from mine.
I don't think intercoms or baby monitors output a (somewhat) sporadic signal like this one. Most of those use FM so they output a continuous wave with no interruptions.
It seems to be related to the triac-like bursts which have been here for a few years. They (and I) predate all other tenants. There are some automated external lights but I think they are just on a timer.
I don't think there is anyway to block external signals except at the breaker panel. I think (but am not certain) its presence on only one 'phase' indicates it's probably not from a source external to the building.
The first thing I did was make sure there are no transmitters of any kind plugged into the problem 'phase'.
I wish this had shown up while I had the Insteon kit here. It would have been interesting to see how well (or not) it handled it.
Did you recently download or install X-blocker? Could be one that is malfunctioning and working on only one phase. You may not know that it may have been reported elsewhere that these are being distributing to neighbors of folks that have claimed that "X10 can be highly reliable if you understand the various gotchas."
Dave understands the various gotchas (and has an oscilloscope with a specialized interface connected for monitoring the AC line. So in his case, does "reliable" encompass being "reliably blocked" ?
There are not applicable enforced standards for powerline quality in the US, so the AC powerline is becoming a free-fire war zone with increasing random fatalities. This is a problem faced by all powerline carrier (PLC) technologies including X-10, INSTEON and CEbus.
INSTEON reliability is increased by signal repetition) with increasing number of INSTEON devices which simultaneously _decreasing_ X-10 reliability by lowering the impedance of the AC line at 120khz ("signal sucking'). This may not be a "friendly-fire" fatality ...
But back to the question:
Unlike the ubiquitous TRIAC-based lighting dimmers that turn the AC off during the AC cycle and are turned back on again only at the zero-crossing, dimmers based on IGBT (Insulated Gate Bipolar Transistors) devices can turn the AC on *and* off during the cycle, not just at zero crossing. IGBT dimmers are becoming more prevalent. As best I know, there is no reason why a manufacturer of a dimmer has to avoid 120khz (X-10) or 130khz (INSTEON) when choosing a frequency to control ("dim") an IGBT. And 120khz _is_ the maximum frequency reported for some popular IGBT devices. So Dave's new-found dependability may be from a newer, IGBT-based dimmer running at about 120khz for light dimming or motor control. Doesn't bode well for PLC in general.
Washing Machine Air Conditioner HVAC (Compressors, Blowers, Pumps) Refrigerator Small Motor Control Traction Inverter UPS Power Supply Servo Drive Medical Power Supply Motor Drives Inverter Welding
etc
formatting link
Lotsa new stuff potentially adding to the powerline cacophony.
Troubleshooting X-10 clearly is for the nimble. You've got a troublesome situation.
In the absence of any leads, it would be the one I would try to follow. It makes me wonder again if I shouldn't try to set up some ESM1's at different points in the house to continuously monitor the signal. It would make it easier to track down such anomalies because you'd at least have some baseline readings from different areas in the house. I recall you once saying that you were never able to reliably detect noise on your ESM1. I assume that wish has been granted in a round-about way. (-:
upstairs/downstairs
I thought it would be a problem. Yet it's plausible someone plugged in something in one of the other apartments that's generating the interference.
I wasn't sure how they functioned. Why would a triac-based device output so randomly? If they're chopping the wave at regular intervals, shouldn't their output be regular as well?
Have you noticed anything new? CF bulbs in the hallways?
That seems to be a good call. If it was outside, it would pass through the main breakers equally and show up on both phases. I suppose that leaves you with trying to determine whether the interference is 24/7 or whether there's a day/night pattern to it as you might expect to see if lightning was involved.
I was pretty sure you did, but it never hurts to touch on even the obvious.
The Electronic Demons don't work like that. (-: It would be too easy!
Have you unplugged stuff on the phase that is exhibiting the problem. The only other thing I can think of is that two transceivers stuck on at the same time causing collisions, but that's so unlikely an event it's hardly even worth considering. Unless someone's paid one of your neighbors to do it -- you do have one or two folks around who would love to see something like this (or worse) happen to you. I've heard of far stranger things done by unhappy newsgroupers!
IIRC, you were reporting pretty strong signal voltage of the interference. What kind of levels of X-10 signal do you normally see in your apartment? If it's in the 3-4 volt range it seems like another indicator that it's something in the next apartment because that seems to be about the right level of attenuation.
I guess absent any other options, I would execute the "X-10 Unplugging Dance" to make sure that it was nothing in your own apartment that went rogue. I had a X-10 Maxi controller that began spewing loads of garbage onto the line when it failed. That was before I got the Monterey so I can't say for sure what it was outputting but they weren't valid X-10 codes that could be captured by Activehome. All X-10 commands in the immediate vicinity of the failed Maxi wouldn't work - just things at the farthest point from the failed controller. I think I still have it in the junk box. It would be interesting if the Monterey showed a similar output to your noise.
My only suggestion at this point is to pretend to be the owner and hire a sparky to filter your apartment at the main breaker. It won't be cheap, though, because of the unusual configuration you have with two feeds from different phases.
RF has similar issues. I recall reading that lots of people living close to some military bases could not get their remote car lock openers and security systems to function because of the interference:
formatting link
[Keyless Remotes To Cars in Waldorf Suddenly Useless]
"Car entry systems, they have no rights at all," said Bruce Romano, who works in the office of engineering and technology at the FCC. "If they get interference, that's too bad for them."
IMHO, these car owners are in a lot worse shape than Dave, who could solve his problem with some $$$, an electrician and a couple of whole-house blockers. RF interference is a *lot* harder to block than powerline interference. In many cases, it's simply impossible.
Even though RF standards are allegedly defined and enforced by the FCC, RF controls suffer the same problems as PLC protocols. Every bit of the RF spectrum is getting more and more crowded. If military transmitters can overwhelm remote car openers, they probably can do a number on any RF home automation protocol. I'm assuming that even mesh networks like ZigBee will fail in the presence of a strong enough RF signal.
You and I share the belief that the only truly reliable system is a hardwired one. It's why the military still has wire-guided munitions in the arsenal. Pretty damn hard to jam. Most of the newer stuff is laser guided, but even that is not as reliable and immune from interference as wire-guided devices.
Some Sovs still think we blew up the Kursk by triggering one of their torpedoes inside the ship remotely. Modern torpedoes don't explode on impact anymore. It's much more effective to detonate directly below the enemy's keel and let the explosion bubble break a ship in half.
I have a great set of Powerpoint slides of a Brit test that show how effective it is. They show a torpedo detonating directly under the hull of a decommissioned destroyer. First, there's a giant black blast that blows out all the loose soot from the ship's funnels as the undership explosion moves upward. Then you see the black smoke turn grayish-brown from high explosive gases along with water spray as the ship lifts up in the water and its keel breaks. Within a very few minutes, the bow half sinks and the stern soon follows. Much, much more effective than simply blowing a hole in the side. To score such a precise hit requires complicated electronic triggering devices. But I digress.
Oddly enough, as more and more RF devices come on line, the interference problem will only become worse. My wireless cams, microwave oven, cordless phones and wireless network all step on each other in different ways. What's the name of that old disaster movie? "The Crowded Sky?" That's what we're looking at here; pollution of both the powerline and the RF spectrum.
I think Insteon is just a bridge that will allow people to perform a phased implementation of the new technology rather than dumping everything at once. In that respect, it's a good idea but, as you point out, it has its own set of issues.
zero-crossing,
Well, thanks for answering my question about why a triac-based problem would appear to be random. What are the advantages of using IGBT's? Are the cheaper? More efficient? And why wouldn't they sync with the zero crossing? It still seems odd that a lamp dimmer of any kind would output such random interference. I guess I'll have to Google IGBT's and learn more about how they work.
There's still hope! I've recently bought two switching power supplies that not only don't attenuate the X-10 signal, but appear to boost it slightly (the GBC video MUX PS's). It may be that customer feedback will initiate design changes. All it may take is a number of product returns marked "knocked out my home's X-10 system." I've returned quite a few items with just that information on the return form. It's possible that those forms might even make it back to engineering design. In this bean-counter society, manufacturers are falling all over themselves to prevent returns, especially to places like Wal-mart that will simply drop a vendor that has too many returns. They did that to a Chinese manufacturer whose name escapes me (Apex, I think) when the quality became an issue and return rates spiked. It nearly killed Apex.
You may have noticed that lots of new equipment comes with admonitions everywhere to "Please do not return this unit before you call
1-800-555-5555." From what I've read, that's part of the campaign not to get blacklisted by the big box retailers who simply despise having to deal with returns.
Returns cost stores *lots* of money in lost sales and employee time and vendors are being forced to strongly discourage them. Look at Smarthome and their "we've never seen THIS problem before." Uh-huh. That's one method of discouraging returns although it's a pretty poor strategy in light of internet users being able to compare notes.
Well, it's not the end of the world. Dave's in a peculiar situation where blocking the noise entering his dwelling is much more difficult than it would be for the average X-10 user. But it's not impossible as in the case of the poor folks who have to deal with military RF interference rendering their car key fobs useless. Dave's X-10 problem is just going to be a royal PITA. I'd rather have the PITA than an "impossible to correct" situation.
You're certainly right when you imply that X-10 users may have to face the fact that to keep things running smoothly, they'll have to fork out for a whole-house blocker. And you're certainly right when you say that the most reliable approach is hard-wiring. Israel obliterated almost all of Egypt's tanks during the Seven Day war and they did it with wire-guided anti-tank missiles. Friends tell me you can still find places in the Sinai where there the spent wires can be found.
If it was good enough for them, it's good enough for me, at least for security. Hardwiring my house for hardwired lighting control ex post facto would be incredibly hard. X-10 isn't as good as hardwire, but by God it's a hell of lot easier to install and uninstall and is usually cheaper than any hard-wired alternative.
Unless one installs an appliance that uses high-frequency waveshaping and is oneself unwittingly generating the noise ...
Did I? hmmm.. I wonder when ;-)
An IGBT used in a dimmer creates less noise than a random-phase TRIAC used for dimming. That's because in this configuration, the IGBT is turned on at the zero-crossing and off when enough power has flowed through the device at which time it is biased off. A TRIAC cannot be turned off during the cycle and so the dimming is accomplished by delaying conduction. IGBT's used as I described are dubbed 'reverse phase' (I presume) because a graph of voltage-time is the reverse of that for a TRIAC. Used in this mode, they are much quieter than TRIACS and don't require massive inductors to filter noise.
However the fact that IGBT's can be turned on and off during at any point of the cycle means that they can be used in other circuit designs. Specifically a modified sine-wave with lower amplitude can be created by switching the IGBT rapidly on/off (eg at 120khz). This is the mode that would be the problem wrt power line control devices.
And forego some sorts of devices which are becoming more prevalent.
lighting control ex post facto
... which is a good thing because one needs to do so very often ;-)
I'm confident I could run this down quickly if my mobility were not so severely restricted.
Powerline intercoms predate X-10. I used them in the early '60s.
I'm not sure this signal is from a triac-based device. It appears to be in sync with the triac like switching transients that have been on the line for several years (but which are not coming from any of my devices). The switching transients come ~1.6ms after each ZC. The endpoints of the periods when this signal is absent coincide with the switching transients but the transients are always there even when there is no change in this signal presence/absence.
That's not a reliable indicator since some X-10 transmitters output 10Vpp and others 5Vpp.
It's not worth that. I only have one X-10 device that's affected and it's in a walk-in pantry and is not critical. There are outlets for each apartment in the basement for washers/dryers. I suspect the tenant in the basement apartment may have plugged something into one of those.
Well, you need to train a monkey or build one of the Star Trek "Exo-comp" mobile troubleshooting computers to assist you. I'm just kidding, of course. I sympathize greatly because when I was laid up after surgery my X-10 went wild and I nearly did, too, trying to troubleshoot what turned out to be a simple problem (stuck Maxi button).
If the transients have been there for years it might be that the device causing them failed and was replaced by a newer and even noisier device. Were the pre-existing transients visible on both phases? (I assume they were since you've mentioned that they appear to be external to your apartment, but any differences between the phases might provide a clue to their origin.)
interference.
Agreed. I am just grasping at straws to try to figure out where the probably source is.
A good guess - but it would be nice to know what it is. I had a CF bulb that put out noise that could be seen on the ESM1 but I returned the whole lot of bulbs. They only output less than a volt right at the bulb, so it's hard to see how they could be a culprit. Still, I would lean towards something like that, which is why I suggested checking to see if the signal might not be present during daylight hours. Of course, apartment buildings often have hallway lights burning all day and night.
One other possibility might be one of the newer, low energy "Exit" signs. Some have CF bulbs and some use LED arrays.
I'd lend you my Monterey if I thought it would be any help. It's a heck of lot easier to check each outlet in the house than a scope setup. It would be interesting to see what it reported at the problem outlet and throughout the apartment.
Well, good luck. If I lived close by, I'd love to give you a hand tracking this down. Alas . . .
The Monterey wouldn't add anything. The scope is far better and the ESM1 can tell me which outlets are involved although I had long ago mapped all outlets to the two 'phases'. I also have a USB scope that is less capable but still adequate for this. I can use it with my laptop when I need something portable. But that really isn't going to tell me more than I already know. The hangup is my lack of mobility.
So, how much did you pay the basement apartment dweller to install your X-10 smasher? (-: As the Remote Sensor Master of CHA, you certainly have the technical skills to pull such a Promethean torture.
Well, finding something like that in your *own* apartment is a lot easier than asking your neighbors to allow you to roam about *their* apartments taking readings and inquiring about recent new electronic additions.
I completely agree with you that it's a serious X-10 liability - so much so that the first thing I do with ANY new device that plugs into the house wiring is to check it against both my ESM1 and my Monterey PLSA to see if it's going to be a problem. I also keep at least three X-10 filters on hand for devices that I feel I must use, even though they are X-10 unfriendly. I'm just more of an optimist than you are. It's clear that they can make dimmers, UPS's and power supplies that don't cause X-10 problems. They question is: will they?
When you explained how IGBT's worked.
So, without those inductors they could be spewing noise onto the line that an inductor-equipped TRIAC-based device might not. As you pointed out previously, X-10 interference is probably low-down on the list of potential problems. I assume manufacturers try to make sure their devices don't cause TV or radio interference first and foremost, since that type of noise is likely to affect the largest number of people.
Specifically
What types of devices would use such a circuit?
That's probably the crux of the issue. I, for one, have grown quite tired of forking over $20 to filter each new appliance that causes X-10 issues. Until the arrival of the GBC switching power supplies I was beginning to think that there would be no end to the parade of signal sucking devices coming to market. The GBC (and one other PS) gives me hope that manufacturers are aware of X-10 and the potential for their devices to inter fere with it.
Some of us have not yet settled into the home we'll live in for the rest of our lives. In addition, when my neighbor had her domestic problem it took less than half an hour to set up the entire house so she could turn on a light in every room at the touch of a button. In certain situations, ease of installation is a critical factor. Holiday lighting is another example of X-10's convenience.
X-10 indeed has a host of problems and in my next house I'll go hardwired for lighting control. For now, I'll just try to keep up with it until my wife finds us our next new home. Then I'll be glad X-10 is so easy to uninstall. HA doesn't seem to ever bring more money for a house. In fact, it seems to scare, rather than attract, potential buyers.
Dave, Considering how much help and information you've provided over the years, perhaps if you'd reveal the city you live in someone nearby would volunteer to stop by and be your "legs" to help track down the problem. Just a thought.
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.