Saturday, April 25, 2009

How to Make an Insteon Phase Coupler from an X10 Passive Phase Coupler

Warning: Anything that you attempt after reading it will be at your own risk. Please don't try it if you don't know how to work with mains power. You could get electrocuted and dry solder joints could create a potential fire hazard. Believe me it is not worth saving a couple of hundred bucks.

I have been using X10 home automation for a long time in spite of the idiosyncrasies of this power line communication (PLC) protocol. As most of you already know, residential homes in US have electric power supplied by a neutral central tap wire and two 110V wires relative to the center tap. Since the two 110V wires are in opposite phases, the potential difference between those two 110V wires is 220V. This allows residential homes to use 220V for washer, dryer and air-conditioner and 110V for rest of the regular home electricity needs. Below is a simple diagram of residential power supply.

Now if you are using a PLC protocol like X10 for home automation, you will have problems sending signals from X10 transmitter on Phase A to X10 receivers on Phase B and vice versa. The solution to this problem is not that hard. X10 protocol uses a carrier frequency of 120 KHz. Therefore, all one has to do is to couple Phase A and Phase B with an LC filter having a resonant frequency of 120 KHz. The LC filter will allow 120 KHz carrier frequency to cross over from Phase A to Phase B and block everything else. Since power supply for dryer has 220V, X10 manufacturers have designed phase couplers that plug into 220V dryer socket and couple Phase A and Phase B for 120 KHz X10 carrier signal. These are called passive phase couplers.

An LC filter is a simple combination of an inductance and a capacitor connected in series. The resonant frequency of such a filter is given by the formula:

In this formula f is the frequency in hertz, L is inductance in henry and C is capacitance in farad. Using Microsoft Excel I can write the same formula as f = 1/(2*PI()*SQRT(L*C)).

Below is a simple LC circuit with an inductor and a capacitor connected in series.

Now I can use this LC filter to couple Phase A and Phase B as shown below:

This is exactly how an X10 passive phase coupler is designed. For X10 carrier frequency the value of the capacitor and inductor will be 0.1 microfarad (uF) and 17.6 microhenry (uH), respectively. In fact, I got these values by solving LC resonant frequency equation for L with f = 120 KHz and C = 0.1 uF. Here is how I can check the accuracy of this calculation:

Using Microsoft Excel you could plug in these values as follows:

f = 1/(2*3.14*SQRT(0.1*17.6*POWER(10, -12)))

This will give you f = 120,028 Hertz or roughly 120 KHz, which is the carrier frequency of X10. A 0.1 uF capacitor is also represented as 104K due to shortage of space on a tiny capacitor. Therefore, in order to design an X10 phase coupler, all you need is a 104K, 400V metalized polyester film non-polar capacitor and a 17.6 microhenry (uH) axial choke. If you open a passive X10 phase coupler, this is what you will find inside it. Generally, I prefer to use 5% tolerance on these components. It is likely that 10% tolerance components will weaken the signal due to lack of proper tuning. A word about inductor or choke. Inductors are available in several packages. It is generally preferable to use an axial choke. These axial chokes or inductors look very similar to resistors and in fact they have color coded bands on them in the same fashion as resistors. However, color code for inductors are different and you may want to make sure that you have the right inductor by checking against the color code.

Now we have figured out the calculations required to design a phase coupler. Armed with this knowledge, we know that designing a phase coupler for Insteon should not be a big deal. By the way, Insteon protocol is not without its own set of idiosyncrasies but it is more robust than X10 due to acknowledgements and dual mesh with RF communication complementing power line communication.

Insteon uses a power line carrier frequency of 131.65 KHz. Now I'll solve LC resonant frequency equation for L with C = 1 microfarad (uF) and f = 131.65 KHz. Solving this equation, we get L = 1.47 microhenry (uH), which I can round off to 1.5 uH. A one microfarad capacitor value is represented on the packaging as 105K. Now we've got the solution for Insteon phase coupler parts.

All I have to do is to replace 0.1 uF 400V capacitor inside the X10 passive phase coupler with a 1.0 uF 400V metalized polyester film non-polar capacitor. Second step is to replace 17.6 uH choke with a 1.5 uH, 100 milliamps choke.

Here are two front and back pictures of a passive dryer/washer type X10 phase coupler.

Add Image

Here is another picture of the guts in which I have made replacements of capacitor and inductor to make it Insteon compatible.

There are many vendors on the Internet who are selling passive couplers, which they claim are X10 and Insteon compatible. I don't know how they can make such a thing without either a variable capacitor or a variable inductor or a switch. Most likely these passive couplers are tuned for X10. Since band-pass filters have a narrow range, an X10 phase coupler will allow 131.65 KHz frequency to pass through but the LC filter tuned for 120 KHz X10 signal will definitely weaken 131.65 KHz Insteon signal substantially. Alternatively, these vendors claiming to have X10/Insteon compatible phase couplers may be using only a capacitor as a filter coupler instead of an LC filter coupler. Even this solution will have its own set of problems and such a capacitor will allow more noise to pass from one phase to another along with the carrier signal.

SmartHome makes an Insteon phase coupler that has to be installed at the mains power circuit breaker box. I haven't asked SmartHome why they are not making Insteon compatible passive phase couplers that could be plugged into a 220V dryer socket. I think I can guess some reasons. One of those reasons could be reliability. An Insteon phase coupler installed at the breaker box will be more reliable for two reasons. First of all coupling at the breaker box provides the shortest run of wires before they are coupled. Dryer will add another 20 to 50 feet of wiring. Secondly, if the dryer trips, the coupler will stop working until you unplug the dryer and reset the circuit breaker. Therefore, installation at the circuit breaker level makes sense. If you are aware of other reasons, I'd definitely appreciate your comments here.