When I started this diatribe this morning I had no idea I would carry on this long. Hopefully this won't be taken as the ramblings of an old man who's team just got their heads handed to them by Ohio State (shouldn't be any doubts here)....
X10, Insteon, and UPB all use your home wiring as a transmission line. As transmission lines go, home wiring systems (and the wide variety of connected loads) simply suck. I'm an EE weeny (over the past 30+ years I've come to accept the term from my ME "friends"), so I'm using the term "suck" in a very technical sense.
A 250' length of home wiring will exhibit a series impedance to the transmission frequency that can expressed in Ohms, uF, and uH per foot. Each junction box/termination on the circuit increases the impedance above the simple romex line values. If a circuit is unloaded, the effect of the wiring impedance is minimal. Turn on a load at the end of the circuit and you will form a voltage divider that attenuates the transmission amplitude. This is the "suck" that I referred to above - every time a load is activated, the signal levels on that circuit are affected.
My experience base has been X10 (30 years) and Insteon (since inception). Nonetheless, other powerline technologies must conform to the same physical properties of the home wiring system.
Published numbers for 12-2 nmb cable are (using the 130 Khz Insteon transmission frequency):
1) resistance - 0.0136 Ohms/foot, 250' resistance (2*250*0.0136) = 6.8 Ohms
2) inductance - 0.214 uH/foot, 250' equivalent impedance (2*250'*0.214e-6*2*PI*130e3 Hz) = 87.4 Ohms
3) capacitance - 10.4 pF/foot, normally not a concern due to low impedance terminations
Effect of turning on a 60 watt bulb at the end of the transmission line (voltage divider) = R bulb/(R bulb + Line resistance + Inductive impedance) = 240 /(240 + 6.8 + 87.4) = .718 (transmission signal is reduced to 72% or original amplitude). Using Insteon as an example (3.2 Vp-p) we would be down to 2.3 Vp-p at the end of the line (perfectly acceptable).
Effect of a 0.1 uF EMC capacitor at the end of the transmission line (equivalent impedance = 1/(0.1e-6uF*2*pi*130e3Hz = 12.2 Ohms)=
12.2/(12.2 + 6.8 + 87.4) = .114 (transmission signal is reduce to 11.4 % of original amplitude). Again using Insteon, we are down to .365 Vp-p (generally acceptable numbers are above 0.1 Vp-p).
Takeaways from the above -
1) Note that I am multiplying the 12-2 cable #'s by two. That's because the transmission signal must make it down the hot lead and back through the neutral. Hence the 2X numbers for resistance and inductance.
2) Most homeowners do not fully understand the tortuous route that their wiring must take. A simple 30' "as the crow flies" run can easily be a 100' foot cable run depending on obstructions and how it must enter the room. Multiply by two for forward and return paths.
3) The transmission voltage level that a particular device see's will depend on it's location on the line. Longer distance from the source = increased attenuation.
4) The above does not account for junction box connection impedance's. These can add significantly. I have my own "budgetary" numbers for powerline impedance. They are far more pessimistic than presented above.
So why do we have problems with communication when the above numbers appear good at 250'?
1) Most modern home circuits include a combination of signal absorbers and noise generators. These combine to lower the signal level (voltage division shown above) while noise will activate the Automatic Gain Control (AGC) circuit in the receiver to prevent false communication. When the AGC is activated the unit will require higher signal levels to recognize valid communication (the receiver goes deaf). Corollary - all house power distribution has noise. Impulse noise occurs at dimmer outputs. Motor brush noise is common on sweeper and other appliances. Switching power supplies and CFL's also generate noise at various frequencies. The challenge for the AGC circuit is to adjust it's input sensitivity to reject noise that could cause a communication problem while ignoring other noise. Some devices do this well, others - no so much. My favorite is the Leviton "intellisense" gated AGC. At one time I belive they had a patent on this approach. Not sure whether it is still in force.
2) The above ignores modern GFCI (wet areas - Bathroom, kitchen, basement, etc) and AFCI (bedrooms) protection devices. These devices use differential transformers to detect ground faults an voltage arcs. The same transformers absorb the communication frequencies (increased line impedance). The amount of signal absorption appears to vary by manufacturer/construction. Beware of any device plugged into one of these circuits. They will attenuate communications and, if combined with a noise source/absorber, will totally kill communications.
3) This is important for 2-way communications - whether or not a device (noise source/signal absorber) causes a problem depends on where it is installed in your home and which direction you are trying to communicate.
3A) A noise source close to a receiver may cause the receiver to miss incoming communications due to signal attenuation on the circuit (AGC activated). The same unit (talking insteon here) should be able to communicate through the noise source to other receivers (noise source is distant and does not activate their AGC).
3B) A signal absorber near a receiver may cause the receiver to miss incoming communications due to low signal level (long line length/GFCI/AFCI). The same unit will be able to drive the load presented by the absorber and communicate with other units.
The best thing a homeowner having a powerline noise issue can practically do is to take a Saturday morning or similar downtime when they have a few hours and start unplugging devices one at a time to see when the situation improves. It's often traceable to one or two TVs / printers / laptops/ battery backups that can be controlled via plugin filters. For example, I have a particular LCD TV that acts like a vacuum for my Insteon lighting signals, while three other LCDs in the house (one an identical model to the offender, purchased at the same time) seem to have no effect. The same goes for backup power supplies - I have one that seems to feed on powerline signals, while others of the same model seem to have no effect. Thus, a small number of plug-in filters will often solve the problem. If you want to be proactive, buy a few of them and put them on all your flatscreen tvs and backup power supplies, and you'll reduce your chances of line noise dramatically.
I realize one can also measure line noise via drops in signal voltage, etc., but a walk-around on a Saturday morning can do wonders, and costs nothing but a little time.
I agree completely with the walk-around approach. Unfortunately, problem devices are also portable and have a habit of moving to different circuits. When plugged into a different circuit they can bring it to it's knees.
I respectfully disagree in regard to your troublesome LCD TV. As an alternative, I'd submit that it is installed on a "problem" circuit. If your were to swap your "identical" LCD TV's, I would expect that the problem would remain on that circuit.
In my experience with X-10 and now INSTEON, it is well worth:
1) the time to develop a complete home wiring diagram
2)using #1 above to isolate INSTEON circuits where practical from utility circuits. In a modern home this is already the norm. In an older home, not so much ...
3) expense to purchase an Elk meter
4) the discipline to keep notes
You can then develop a strategy along the lines of [ I'm blocking on the name of the now-defunct (?) company that designed and sold a proprietary systems of X-10 filter/couplers near entrance panel to segment X-10 .... grrrr)] to isolate problems.
# 5) - 10) are: install an ISY
HTH ... Marc
Marc - I agree with all of the above, but would like to add/re-order some things.... The following is specific to Insteon.
1) Boosterlinc
2) Boosterlinc
3) Boosterlinc (and boosterlinc enabled X10 devices) - These devices will
actively interfere with Insteon communication. Contrary to SH doctrine, I have never encountered a Insteon friendly Boosterlinc device. They are virtually impossible to troubleshoot because they actively "talk over" Insteon communications. If you have one of these in an Insteon system, donate it to a landfill. Jeff Volps XTB-R plug in repeater has been specifically designed to co-exist with Insteon and will boost X10 levels to the 32 Vp-p range.
4) Circuit mapping per Marc's note - there really is no substitute for this. Know where your problem circuits (AFCI/GFCI) and problem devices are. Beware or plug-in problem devices (play-stations, cell phones, etc) - they move.
4,5,6) ISY - can't imagine using insteon without it. I used my Insteon devices in X10 mode for two years waiting for a credible controller to come out. When the ISY was born I jumped and haven't looked back.
7) The Elk meter was a boon for X10 - it's been discontinued. In general, signal meters have limited value with Insteon systems. The repeating nature of Insteon makes it impossible to determine what device is talking when... If you wish to monitor in-band noise, Jeff Volp has developed a noise/X10 signal meter that has received wide acclaim. The Insteon transmission frequency is close enough to X10 to make the noise function useful. While the device May register Insteon signal levels, it cannot validate Insteon communication.
Other -
Insteon has managed to overcome many of the disadvantages of X10 through it's repeating (hopping) protocol and CRC/command/response confirmation. The addressable unit feature and protocol is what drew me to the technology to begin with. While I believe it is the most flexible protocol (my opinion), it is also the most complex and difficult to troubleshoot. To that end -
1) Device programming/response errors/firmware - prior to assuming that you have a transmission issue, verify that your target device/PLM have the correct programming (link table entries). Without them, the best transmission line will do not good. As noted previously, V.35 firmware devices have know problems.
2) Modem location (this applies to Insteon/X10/UPB) -The best location for your modem/PIC/whatever is next to your primary load panel. Your modem is the brain for your system and the load panel is "grand central station" for all the wiring in your home. The load panel should also be a low noise location.
3) Accesspoints/Signalincs - SH pioneered the RF phase linking using these devices. In short - I'm not a fan. I use a single passive phase coupler at my main load panel in a 4500 sq ft home. Why? Active devices can be upset by noise, power surges, etc. The passive coupler always works. The APs/SLs also add a level of complexity to the system that I do not need. I can't tell what route signals are taking with these installed. With the passive coupler in place, I know exactly what route signals are taking. I can trace them and isolate problems.
Hope this helps,
IM
