X10 is a protocol for communication among electronic devices used for home automation where the signals involve brief radio frequency (RF) bursts over a home's A/C power lines.

Household wiring often has two different power circuits; each of the two 220V offers a single phase circuit to neutral. This means if the X10 controller is on one of those 110V circuits, the signals are not passed to the other of those circuits.

The solution to this is to bridge the two 110V circuits so the RF signals may pass between them. This question is about what the specs for the capacitor should be if we wanted it to be very reliable over time.

I bought and installed a capacitor like this (for a dollar) and it worked for probably five or more years:


But then it failed, and looked like this when I found it:

X10 Bridging Phase Coupler Capacitor Failed

Capacitors, when operating properly, do not conduct [significant] current across their leads, but this season (putting up Christmas lights), the signals were not passing to all outlets, and when I looked in the breaker box, the capacitor had dropped away and fallen to the bottom. The leads were gone (apparently melted). So the dielectric apparently broke down over time and the capacitor failed.

This got me to thinking, would there be a higher quality / higher spec capacitor that would be less likely to fail? If it meant I could count on it when I needed it, for decades to come, it would be worth spending more on it. But I don't know what to buy. I suppose I could install more than one, then if one failed, I'd have another. But I'm not going to be pulling the cover off of the breaker box to see if one or the other blew. Obviously not a critical problem, but just curious if there's a solution where a little more investment would mean no problems in the future.

Follow-up Appended Edit:

I ordered another one from the same place as I got the one the that had earlier failed:

X10 Bridging Phase Coupler Capacitor

  • 3
    \$\begingroup\$ Capacitors, when operating properly, do not conduct current across their leads Think about this one a bit more - if they take no current then what use are they? \$\endgroup\$ – Andy aka Dec 12 '19 at 15:28
  • \$\begingroup\$ Oh dear me @JackCreasey you have misread what I put. \$\endgroup\$ – Andy aka Dec 12 '19 at 17:25
  • \$\begingroup\$ I added the word "significant", so now it reads "capacitors...do not conduct significant current", but one could also argue with that, I suppose. The only reason I mentioned it is because when significant current passes, the whole thing blows-up! \$\endgroup\$ – Dale Dec 12 '19 at 23:12
  • 1
    \$\begingroup\$ 630v is a fairly low rating for this duty, when you consider "glitches" such a lightning strikes on the main line. I'd be looking for something closer to 1000v. \$\endgroup\$ – Hot Licks Dec 13 '19 at 3:06

I like X-10, use it myself, and have put it into several friends' homes. What has worked for me for decades is a 0.22 uF plastic box capacitor, rated X, X1, or X2. This is a film capacitor similar to the one pictured, but rated for continuous connection across AC lines such as Line-to-Line (240 Vac). Most are rated for something like 275 Vac.

This type of capacitor is called "self-healing". If an internal short happens, the transient current blows away the shorting material, often such that the part continues to have enough capacitance to function normally.


enter image description here

UPDATE: Seeing the added photos of the old cap with the "104" marking (0.1 uF) jogged a memory. Long ago I knew that was the recommended value, and used it for a while. But as I expanded the X-10 stuff around the house, some controllers could not "reach" some devices, so I increased the value to 0.22 uF and that solved it. That is why I recommend 0.22 uF over 0.1 uF. If you want to compromise, 0.15 uF also will work.

  • \$\begingroup\$ You put them AFTER a fuse? \$\endgroup\$ – winny Dec 12 '19 at 17:28
  • \$\begingroup\$ In both of my breaker boxes they are connected directly the two phase busses coming from the main disconnect breaker. In this way, any controller on any branch on either phase is connected to everything on both phases. Note: it is a self-extinguishing component inside an all-steel box. \$\endgroup\$ – AnalogKid Dec 12 '19 at 17:49
  • \$\begingroup\$ X-capacitors are unique in that they are allowed to fail to short and hence needs to be installed after a fuse. Y-capacitors are opposite, but unlikely to be available in 220 nF. \$\endgroup\$ – winny Dec 12 '19 at 17:53
  • \$\begingroup\$ Is there a range, above or below, 0.22uF that would be just as good? \$\endgroup\$ – Dale Dec 13 '19 at 14:51
  • 1
    \$\begingroup\$ 0.1 uF to 0.22 uF. Below 0.1 and I'd worry about the increase in the impedance at 100 kHz. Above 0.22 and I'd worry about the ESL (equivalent series inductance) increasing the effective impedance at 100 kHz. Of course, small differences in internal construction can outweigh everything. I use Panasonic parts because that's what I have left over from other projects. No problems ever, so far. \$\endgroup\$ – AnalogKid Dec 13 '19 at 16:15

Capacitors that are designed for direct connection across AC lines are designed to carry the resulting current at the line frequency. However they are not necessarily designed to carry the current resulting from the harmonic voltage that results from nonlinear loads distorting the power waveform. Every electronic circuit load from LED bulbs to motor controllers causes line voltage distortion. Devices such as the one pictured in the question may be designed to accommodate some harmonic current. If the voltage marked on the device is the rated RMS voltage, the margin between that voltage and the 240 volt line voltage represents some margin for harmonic current capacity. You should study the specifications for the devices that you can find.

Since the device pictured is marked "phase coupler" and lacks a capacitance marking, it is not clear what is inside. If it is intended for the purpose that you described, you should try to find more detailed information about it. Compare it with whatever you originally had installed.

  • 1
    \$\begingroup\$ The current through a small-value capacitor under 60Hz voltage is inconsequential. \$\endgroup\$ – Hot Licks Dec 13 '19 at 3:04
  • \$\begingroup\$ @HotLicks: Did you understand Charles' point about harmonics? These will cause increased dielectric heating. \$\endgroup\$ – Transistor Dec 15 '19 at 16:20
  • \$\begingroup\$ @Transistor - You are not going to see harmonics above maybe 300Hz to any significant degree, and the amount of these will be very small. \$\endgroup\$ – Hot Licks Dec 15 '19 at 19:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.