Which One Should I Use #20. "Weird Coupling Situations" "Uncle" Phil Kingery Email Uncle Phil

Even though this is the twentieth article in this series, it really fits between #5 and #6. This one is going to try and explain the complex situations that often arise when trying to use multiple coupler-repeaters.

You would not believe the misunderstandings some people have about repeaters. (Now that I've said it that way, I suppose that was a little presumptuous. After all, you probably have some of the same misunderstandings, or you wouldn't be here reading this.)

Here are a few true stories that I have had to deal with in just the last couple of weeks! (The names were changed to protect me from a lawsuit.)

1. "Yes I know that a coupler-repeater gives the appearance of amplification, but in the strict sense of the word, an X-10 compatible repeater is not an amplifier. And don't try to argue with me, it just isn't and that's it!."
   
2. A large house with six sub-panels was in the final stages of construction so "they" (long story) ordered six coupler-repeaters. The house may have had a lot of sub-panels but its total ampacity was only 800 amps, so one CR234 should have been sufficient. (Plus, the electrician, who had no idea what he was doing, told me he was just going to install them all at the same place, because he had enough spare breakers at that sub-panel. (Wrong!)
   
3. One installer was having intermittent and unreliable operation, so he ordered a second repeater just like the one he already had (not an ACT repeater, by the way) and proceeded to install it on the same breakers as his first repeater. I guess he thought it would give him twice as much signal, kind of like putting two batteries in series. ("Buzzzzz" – nice try.)
   
4. An installer had twin-mains (two 200 amp panels side-by-side, acting like one large 400 amp panel). He installed a coupler-repeater on each side and installed a whole-panel blocking filter on each panel and then could not figure out why devices on one panel would not talk to devices on the other panel. ("Buzzz" – thanks for playing!)
   
5. Finally (at least for this list of examples), an installer was trying to improve the reliability in a large home, which had a separate garage and a separate guesthouse. A single repeater was installed at the main house but the signal was too weak for reliable operation in the guesthouse. So he installed a repeater in the guesthouse and connected the two repeaters with low-voltage SCC lines. (When everything crashed, he called me.)

Okay, as I write this, I can see that trying to go over this many examples will probably make for one very long article. I will try to do several before I decide if we need to make this a multi-parter. (Of course, by the time you see this, you will already know if it is the first of two parts, because I will have already written them.)

I know I have said this before, but it is worth saying again. A coupler-repeater is NOT an amplifier. I still get arguments from some people who say that since a repeater increases the signal strength, it is an amplifier.

No it isn't! It does not "increase" the signal strength! It may give that appearance, but what it is doing is creating a new signal. According to the " Dictionary of Electronics " (http://www.twysted-pair.com/dicta.htm):

amplifier - A circuit that increases the voltage, current, or power of a signal.

So, a repeater does not increase the voltage, current, or power of the original X-10 signal, it "repeats" the data frame, but (I will admit) since it repeats it at a level that is usually much greater than the original signal level, it gives the appearance of amplification.

I realize that in many situations, I am a purist and I am known to split hairs, but in this case it is important. We at ACT make several different "repeaters". If you need an explanation as to how a repeater works, go back and read "Which One Should I Use - Part III", which gives a simplified version of how all X-10 coupler-repeaters operate (yes, even our competitor's coupler-repeaters).

We also make a couple of true "amplifiers", and they are not the same thing as a repeater. There are several things that seem to fuel this continued confusion. One is that our true "amplifier" looks just like one of our competitor's repeaters. The other is that our competitors usually call their repeaters "Signal Amplifiers".

Okay, so what is the difference?

Let's talk about "amplifiers" first. In an X-10 sense, amplifiers do have their place. Often in a large PCC installation, it may be necessary to amplify the signal before sending it to other buildings over dedicated lines. Or perhaps, the signal needs to be increased and sent to a different transformer system within the same building. Perhaps the signal needs to be amplified so that it will travel down a 3-mile cable to an unmanned deep-sea submersible.

Yes, that really happened once. An oceanographic research facility wanted to add some features to an existing deep diving robot. Replacing the tether (the cable which connects the deep sea unit to the research ship on the surface) would be extremely expensive, so they did the job using PCC (using the basic X-10 protocol).

You may be asking yourself, "If an amplifier can do that, why do we need repeaters?". Well for one thing, an amplifier is unidirectional. It can amplify the signal in one direction only. Plus, amplifiers have one nasty problem.

All of us are familiar with audio amplifiers. When I teach classes, I sometimes have to use a microphone. (Even with large classes, I usually do okay without a mic. I do, however, warn those in the front row that I speak very loud, very fast and occasionally, I spit. However, for purposes of this example, we will pretend I use a microphone.)

In any "public address system", there are 3 parts; the input path (microphone), the amplifier itself, and the output path (speakers). As I teach, I tend to wander around the room. We all know what would happen if I were to walk too close to the speakers; the amplified sound from the speakers would re-enter the microphone and be re-amplified and round and round it would go. The resulting squeal is an example of bad feedback.

If you were to use a "true" amplifier (ACT part number CA000 or CA200) in most residential installations, you would not be able to separate the input from the output. Even so, I have seen installers try using one anyway. Unfortunately, as soon as they transmit that first X-10 signal, the CA000 amplifies that signal which then feeds right back into its own input. Literally, within a millisecond, the entire electrical system is flooded with constant 120kHz. It is the X-10 equivalent to that microphone in front of the speakers situation except there is no way to move the microphone.

That is why we need "repeaters". A coupler-repeater (in the X-10 world) is an intelligent device that will receive, process and then repeat individual X-10 data frames. It's input and output are on the same set of wires. That way, it can receive a frame of X-10 data, process it, and then transmit that frame of X-10 data on the same wires. Of course, for a repeater to work, the original signal must conform to the correct X-10 specifications.

I know of a few companies who demonstrated their misunderstanding of the protocol and ended up designing their transmitters which did not work very well. Many of them struggled with this for a long time. Some never got it and went out of business. Others finally figured it out. It was hard for them to understand that just because their equipment worked flawlessly in the lab, it could be very unreliable in the real world. (If you want to know more about the basics of the X-10 protocol, you might want to go back and read WOSIU#13.)

Okay, enough of "A Repeater is Not An Amplifier", so let's move on.

Old Chinese proverb: "Two birds: each fly beautifully. Tie them together and neither can fly worth a crap." (…or something like that…)

The same can be said of two (or more) coupler-repeaters. Sorry, but since each repeater has its own 120kHz oscillator, putting them near each other usually does more harm than good. Let's go over a few simple scenarios, some good, most bad.

First, lets look at a situation where two repeaters can be installed in a facility (residential or commercial building), and they kill the entire system.

Look at figure 6. For the moment, please pretend that the distances between the panels are quite a bit apart, maybe hundreds of feet. Perhaps the main panel is on one side of the house, one sub-panel is on the other side of the house and the second sub-panel is in the garage.

For the sake of this illustration, let's say that the installer had already tried one repeater at the main panel, but that didn't appear to work. There seemed to be weak areas at the distant ends of the house and the garage circuits. In an effort to increase his system's reliability, he decided to move one of the CR234's to a sub-panel and then install a new CR234 at the other sub-panel. He left the dip-switches to their factory settings and fired them up. At first, he did not wire up the SCC lines. (SCC= Signal Coupling Conductors, which are the low-voltage wires that appear between the two CR234's in figure 6.)

At first it seemed to work (and he was pretty darned pleased with himself). When he first tried it, the original signals were coming from his computer controller located on a circuit at the main panel. When it transmitted, the signal made it to both repeaters at the same time and both repeated together, like a chorus. Life was wonderful!

Although he had not really given it much thought, he was like most people. He believed that he should be able to install as many repeaters into one facility as he felt were necessary. He thought the repeaters would pass the signal along like firemen in a bucket brigade, passing water buckets to a fire. If he sent a signal from a distant part of the house, that signal should go to the first repeater which (he thought) would then pass it on the second repeater, which would then (theoretically) send it to the far distant part of the system.

Then he pressed a button on a wall-mount transmitter at the other end of the house. Suddenly, everything locked up. Nothing worked. He went to check the LEDs on the CR234's and was surprised to see that both of them were blinking "Receive" and "Transmit", over and over, in a never ending loop.

This can be very confusing even to experienced installers. There is a dip-switch inside the CR234 that tells it to ignore, or repeat, previously repeated signal. For the overwhelming percentage of installations, it is best that repeaters be set to ignore previously repeated signal. In this case, his two repeaters ended up passing the same bucket of water back and forth to each other, forever. We call this "ping-ponging".

There is a deeper problem aside from the likelihood of having repeaters repeating each other's data frames over and over, and that is "partially cascading data frames". For this example, we are going to say that our do-it-yourself-er recognized the ping-ponging repeater problem and turned off the breakers on one of them in order to get the cycle to stop. Then he set the first CR234 (the one closest to the transmitter) to "ignore previously repeated signals" (switch #3 = ON). He left the second CR234 (the one closest to the receiver to which he was trying to send the command) set to "repeat repeated signals". He figured that way the signal would be passed from the first repeater to the second repeater and on to the receiver, but the signal would not come back to the first repeater.

The result is unfortunately, lots of address data all over his house but no commands. The transmitter sends the first address frame (A01), which the closest repeater "hears". The closest repeater then repeats that address frame exactly in time with the transmitter sending the address frame the second time (A01). The distant repeater receives the "repeated" address from the near repeater, and repeats it again (A01) right over the top of the transmitter trying to send the command portion of the signal.

To try and illustrate this, I will use a non-proportional font so that the "timing" will line up:

A01,A01-AON,AON <-- original signal sent from the transmitter.
--- A01 --- --- <-- repeated signal from the First CR234.
--- ---A01 ---  <-- repeated signal from the Second CR234.

Unfortunately, the closest repeater cannot "hear" the command data frame because it is also being bombarded with the repeated address data frame from the distant repeater. (It's actually a little more complicated than what I have shown, but this is the basics, not a college course.)

There are some situations where this setup can be advantageous, but the entire system must be custom designed. The transmitter has to be either modified or it has to be computer controlled so that the data frames can be separated by enough time to allow signals to "cascade" down the line from the first repeater to the second repeater, without the second repeater "stepping" on some part of the signal that the first repeater is trying to receive.

A01,A01 (pause) AON,AON    <-- original signal sent from the transmitter.
--- A01         --- AON    <-- repeated signal from the First CR234.
    ---A01          ---AON <-- repeated signal from the Second CR234.

For this to work, the first repeater is usually set for "ignore" previously repeated signal (dip-switch #3=on), then the second is set to "repeat" previously repeated signal (dip-switch #3=off). This also means that the whole system must be controlled by a single controller.

Another problem should now be obvious. Signals can go out, but not back. In some installations, that is not such a bad thing, but modern powerline products are all 2-way.

One more piece of the puzzle and that will be enough for this edition. After failing to get his system to work, the installer finally tries to hook up the SCC lines. "SCC" stands for signal coupling conductors. These are low-voltage wires intended to transfer signals from one repeater to another. In large installations, it is common to have a dozen repeaters all talking to each other. However, for this to be effective, each repeater must be on its own transformer system, NOT on the same transformer as another repeater. And so, he finally learns that the SCC wires do not help either.

This (figure 8) illustrates a simple "system-to-system" coupling setup that does work. Any signals that originate in the building on the left are repeated onto all 3 legs of that electrical distribution system, plus they are sent to the building on the right.

The CR134 on the right repeats that signal onto all 3 legs of that electrical distribution system. The CR134 is smart enough to know that if it received the data frame from its SCC lines, it doesn't have to send it back onto the SCC lines. That prevents SCC line "ping-ponging". Calling one of the buildings the "source" and the other the "target" is no longer appropriate since now, signal can originate from either building.

So far, the largest system that I know of, connects 17 transformers, each with its own CR134 all connected with one daisy-chained SCC line. I say "that I know of" because there may be many other larger systems out there, somewhere, that are working just fine. If they work right, I rarely hear about them.

Okay, as I expected, this subject is too large for just one article so the next one will be "More Weird Coupling Situations". So I think it best that we stop here and come back later and add more. Don't worry. When we meet again, I will give you some ideas how we are going to fix the problem of the guy with 2 sub-panels.

Oh, by the way, did you notice that there are 8 dip-switches? And a jumper?? ACT coupler-repeaters are very flexible (if you know what you're doing).

Until next time,

To infinity…AND BEYOND!!

-"Uncle" Phil Kingery.