n w    w w w w

baner
You are here:   Home Content Home Automation Home Aut0mat10n and Other Data Corruption
large small default
Home Aut0mat10n and Other Data Corruption PDF Print E-mail
Written by Rick Swenton   
Monday, 11 July 2005 17:31
One of my hobbies is home automation.  I have X10 home automation modules to control lights and appliances in my house. These devices send and receive commands using a serial protocol at 120 KHz over the power wires in the house. They send and receive just slightly after the zero-crossing point of the AC voltage so the signals are sent when there is almost no voltage on the lines.  It has been a real challenge getting the X10 system to perform reliably in my second home.


Using an oscilloscope (my boss laughs at me because I have one)  I can see wildly changing noise on both phases of my incoming lines. One phase is more problematic than the other. They are short duration low voltage spikes. Sometimes they are time stationary. Sometimes you can see one drift up and down within the AC sine wave. These noise pulses are confusing the zero-crossing detectors in the X10 system. This causes random loss of the functionality of the system.

I know the source is not in my house because I turned off all the loads (except for the power to the scope) and the noise is still there. I thought it was a particular neighbor next door, perhaps with something like a sodium discharge lamp, because I seemed to have more problems at night when he was home. Lately there does not seem to be a pattern other than some nights are worse than days.

I know X10 is an old technology. Some people even call it unreliable. I have significant experience with X10 and I have never been unable to resolve X10 problems until I moved into this house. I installed a commercial quality X10 amplifier/repeater that monitors both phases for an X10 signal and then retransmits a strong signal on both phases. Even this device gets confused. My problem is not line loss or weak signals. It's the noise pulses that are confusing the zero-crossing detectors.

My house is on a cul-de-sac with two other homes. The High Voltage feed comes from the main highway into my street to a pole and transformer. The transformer feeds the 3 houses on my street.  Interestingly, 220V also returns to feed one house across the street on the highway. Maybe this saved the cost of another transformer on the highway.  Because the noise I see is different on each phase, that would suggest that the problem is between the transformer and the houses, because the transformer has a single phase High Voltage primary input. There are only houses on the transformer secondary. What could any of them be doing in their homes that causes such a continuous wide variety of changing noises?

I tried the commercial X10 filters. Because they are bandpass filters, they only provide for X10 signal pass through. They don't help with the zero-crossing problem.

I also know that the power company is only responsible for delivering reliable, continuous power - and this particular power company, Nstar Electric, does this well. I know they are not responsible for every noise pulse on the line other than the FCC requirements governing Radio Frequency Interference.  That's part of the challenge.  You can't hold them accountable for nuances in the power that cause highly specialized products to stop working.  After all, the power company never promised that you would be able to reliably send a 5 volt signal through your 110/220 volt power lines to remotely control lights and appliances and guarantee that it would work.

This is the kind of problem that will take a long time to solve and require skill and perseverance.  Don't you just love these kinds of problems?  More to come.

Jul 11 2005, 04:50:04 PM EDT (Transfered from my blogs.sun.com account)