This article describes special issues that one is faced with when using X-10 technology outside the popular 110v, 60Hz single/double-phase US powergrid system. In addition, this article offers possible solutions to these issues.
Although X-10 was originally developed in Europe (by a Scottish company called ‘Pico Electronics’) its main market today is the United States. However, luckily, the original X-10 protocol was broad enough to handle other voltages, frequencies and triple-phase systems.
[begin of shameless plug]
For quite some time now, I maintain a web site dedicated entirely to X-10 technology. This web includes special information pertaining to 220v systems, so naturally, this article will be sparkled with links to the web for further information. Others pointers will link you to Phil Kingery’s excellent article series on X-10 technology, residing on the home toys web site.
[end of shameless plug]
Phew, that out of the way, here are the topics we’ll cover:
* Issues and solutions for X-10 in a 220/240v environment
* Issues and solutions for X-10 in a triple-phase environment
* Issues and solutions for X-10 in a 50Hz environment
* Differences between US and European X-10 wireless protocols
Issues and solutions for X-10 in a 220/240v environment
The X-10 protocol is inherently insensitive to the system voltage. All it cares is that the carrier voltage will alternate between positive and negative. The amplitude is not part of the protocol. thus X-10 theoretically and practically works on systems ranging from residential 110v to industrial 277v.
However, for many home automators X-10’s low module prices is a compelling reason for choosing this technology. The prices of 220v modules reveal an upleasant surprise, as shown by the following table:
Price for 110v module
Price for 220v module (roughly converted from GBP to US$)
Price Difference in %
Lamp module $7 $41 480%
Appliance module $7 $41 480%
Wall switch module $7 $44 528%
Eagle Eye motion detector $24.99 $72 188%
powerflash module $30 $51 70%
Wireless transceiver $16 $55 320%
Mini controller $8 $28 250%
Remote Chime $15 $35 133%
So if you limit yourself to purchasing 220v X-10 modules, you can easily expect to spend three to four times as much on your system components alone. Worst, the most popular receivers are priced close to 5 times as much. With such a huge difference, you may be thinking that the 220v modules are superior in technology and features. Flatly, they are absolutely not. The only difference between the two types of modules is the value of a very few components inside the modules. The difference in price is related only to the size of the target market.
One solution is to convert 110v modules to 220v yourself. This is not endorsed by X-10, of course, and will automatically void your warranty and UL listing, and is absolutely a ‘do at your own risk’ project. There are safety risks involving any electrical work, and particularly in this one since some parts of the module carry the powerline voltages. Know what you are doing! Always work in a ground protection fault circuit, stay clear when you apply power to a modified module for the first time and be familiar with the procedure and understand it before you start. Electronics is a fascinating hobby, but be careful.
Having said that, it is usually very satisfying to modify and improve a module, and after a couple of modules, becomes relatively easy and straightforward – about 10 to 15 minutes per module. Fortunately, the X-10 modules contain 80’s era electronics (read, very, very simple compared to today’s electronic integration), that can be easily modified and, in many cases, improved. The components used are very cheap. My web has general and specific details for each module conversion. These conversions have been tried and tested by numerous individuals (myself included) and I can attest that they work flawlessly. Furthermore, in many instances, once the module is open for surgery, one can easily add features to it, in addition to the voltage modification. For example, the appliance module can acquire local control. The wall switch can have local dimming (by pressing and holding the button), and this is just the tip of the iceberg.
In certain cases, an alternative solution to conversion is the use of a small 220-110v step-down transformer. In fact, it is sometimes a more economical solution than module modification. For example, suppose you wish to construct an X-10 6-zones irrigation controller using 6 universal modules (one per zone). Rather than converting all of them to 220v, one can simply cluster them in a box and feed the entire box 110v from a small step-down transformer. The transformer may need (actually, it usually does) a capacitor bridge to pass the X-10 signal. The following figure shows how:
In 220V ——–o——-)|||
0.1uF | )|||
600V — )|||
Out 110V ——–o——-)||| Transformer
Common 0V —————-)|||
Notice the 0.1uF (microfarad) 600v between the 220v input and 110v output. It is used to pass the 120KHz X-10 signal from the devices to the powerlines. The rating of the transformer can be very low. About 5W rating for each module driven is sufficient (modules typically are rated around 2W). Bear in mind that a transformer works even when it’s load does not, so take a margin, but don’t get carried away with it.
The transformer solution is also applicable to any device that does not feed a load (does not turn on a high wattage lamp or appliance). Here are more examples where using a transformer is warranted:
* Controllers. This is a great solution for controllers, since converting controllers involves more components.
* A cluster of powerflash modules that interface PIR motion sensors to X-10 (though there is an even cheaper solution for this specific issue)
Finally, there are 220v modules in the US. They are the ‘heavy-duty’ modules designed to handle larger loads. Some examples are the heavy duty appliance module (available for 15A and 20A load) and a heavy duty switch which is an appliance module in a switch envelope. These devices are superior in specification and built to the standard modules. However their prongs and socket are made for the US market, so an adapter or modification is needed to use them.
Issues and solutions for X-10 in a triple-phase environment
Triple-phase systems pose a unique problem. To understand why, we’ll need to understand a bit about powerlines and X-10 protocol. A great description of this issue is given in Phil Kingery’s article on Digital X-10 at the Hometoys web library.
All utilities generate electric power in three phases. You can see it for yourself next time you pass near a high-voltage intercity powerline. Look up and notice that the wires are grouped by triplets. Each wire in a triplet carries one ‘phase’.
wpe1.jpg (13144 bytes)
Now, the way X-10 controllers work, is that they detect the zero crossing of the powerline voltage, and on that event superimpose a short pulse of a carrier signal (120KHz) onto the powerline. This is repeated three times, to cover all phases as shown in the next figure.
wpe2.jpg (9548 bytes)
The presence of this carrier pulses indicates a logic 1, and the absence a logic 0. This is how X-10 controllers transmit commands. The X-10 modules (receivers) detect these zero-crossing pulses, and if the address match, perform the function specified in the command (ON/OFF/DIM etc). Obviously, the timing between those pulses must coincide with the zero crossings of the three phases, and there’s a catch here, but let’s leave this for now (we’ll get back to this in a bit).
Suppose that the controller is on the black phase. It will obviously be able to control all modules on that phase. However, modules on other phases will very likely not receive the controller commands. This is so because the X-10 signal can’t simply jump from one phase to another, without some kind of electronic bridge.
In order to solve this, the phases must be coupled, which is another term for bridged. A simple phase coupler (called a passive coupler) consists of a 0.1uF/600v capacitor that passes the X-10 carrier pulse from one phase to another. This usually works. You can also make a high efficiency phase coupler that also includes an inductor and a fuse. More details on some limitations of a capacitor bridge can be found on Phil Kingery’s “capacitor as a bridge” article.
On triple-phase systems, remember that two couplers are needed. In our example where the controller is on the black phase, a phase coupler must be installed between the black phase and the red phase, and another between the black phase and the blue phase. This sometimes poses a problem, because the original X-10 signal strength is reduced by the coupling. I suggest to simply try it and see. The result is largely dependent on the length of the wires and the type of loads on the powergrid, so there is no universal guideline here. Although breaker panels should not have components in them according to the National Electric Code, the best place to put a coupler is exactly there.
Another possible solution which is often overlooked is to move circuits around in the breaker panel. Thus, all lights and power outlets that are to be X-10 controlled are moved to one phase. The large uncontrolled appliances (oven, fridge, microwave etc) circuits to the second phase and the triple-phase appliances (HVAC, washing machine) feed as usual from all three phase. Of course, this work is for an electrician.
Issues and solutions for X-10 in a 50Hz environment
Remember our three pulses generated by the controller? In addition to coupling, a 50Hz environment pose another issue: the internal timing of the X-10 controller. By this I refer to the time difference between each of the controller pulses. For example, take a look at the following figure, describing the time points within a powerline cycle in which an X-10 controller sends its pulses after zero crossing:
wpe1.jpg (10617 bytes)
This controller was made for a 60Hz system. To cover all three phases, it is programmed to send pulses at zero crossing (ZC), ZC+2.778ms, and ZC+5.556ms. The second and third timing points coincide with the zero crossing of the second and third phases. Now if we take this controller and simply plug it into a triple-phase 50Hz system, only modules on the controller phase will receive its commands. Why? Because at 50Hz the zero crossings of the second phase occurs at ZC+3.333ms and that of the third phase occurs at ZC+6.666ms. The controller sends its pulses at the ‘wrong’ times. This problem is not limited to controllers. All transmitters (such as the powerflash modules, all two-way modules, mini-controllers) designed for 60Hz will have the exact same problem.
For controllers, one possible solution is to use a controller that can handle 50Hz systems. I am aware of three such controllers: the Marrick LynX-10, (priced at $50 in kit form) the Adicon CPU-XA/Ocelot (~$150) and the smart-america Thinkboxx (and the real price is…drums please…$1200). The LynX-10 and the Thinkboxx will self detect the powerline frequency and adjust their internal timing accordingly. The Adicon CPU-XA/Ocelot does not detect the powerline frequency, but it can be special ordered for 50Hz systems.
As for other transmitters, they must be on the same phase as the receivers they affect.
Differences in X-10 wireless protocols
The final issue is a bit of a weird one. For some obscure reason, European standard committees did not like at all the frequency 310MHz but found nothing objectionable with the frequency 418MHz (anyone can clue me in on this?). So, X10 Europe had to tune all their wireless products (transceiver, palmpads, remotes, motion detectors etc) to this magical 418MHz frequency.
This incompatibility means that European components will not talk to North American components. So purchase your entire collection from either the old continent or the new one, but not from both. You can, however, purchase a North American system and modify the wireless transceiver base to 220v and use the system in a 220v system. Luckily, the European committees did not object to battery sizes.