@#$%^& X10 hardware

Sounds weird. What kind of modules and transmitters? What have you changed recently in your home? Any new devices?

-- Bobby G.

I don't think so, really need to double check. The Christmas lights went up but they're not the type with a power supply. I think I've got it down to being in my computer room as the devices that are side by side won't work (the garage light, which is on the other phase will).

I'm currently working with the TI103 and the CM11A.

One thing to note is that we took a power hit when this started and I've been seeing various house codes that I don't use being sent. So I have a nasty problem that's going to cause me to sit down and isolate circuits. I'll start by removing devices in my computer room. I think I need to order some chips from Jameco. I need to button up the X10 sniffer and put an opto on the serial port to make is safer and portable.

OK, I've a little further progress on this subject. Today I was trying out Mark Gilmore's ECS

software with my Insteon USB PLC. After some time of fighting with the PLC (it's behavior didn't follow the Insteon documentation, grrr). Anyway I had some of my Insteon working and I had some of the X10 working (more than previous) and it managed to get worse as the day progressed right back to the previous odd behavior. What I have discovered is that the signal is partially getting out. The first half seems strong enough at about 4V. But the second half is getting garbled or not sent at all. While the X10 looks valid the command D1 D1 DON DON gets sent as D1 D1 B5. As the day progressed the Insteon would work also. The Insteon PLC is now flickering, not blinking but flickering like traffic is being sent. This pretty much keeps the PLC from sending anything. The T1103 works but that seems to be it.

My next course of action is the start tearing everything out and add one thing at a time until the network fails. I have a feeling that the problem may be related to the cheap Geek Squad UPS units I have. They're behind filters but I still suspect them.

Not knowing where to start, I'll comment on what I think it most unusual about your situation - that the signal starts out well but deteriorates. Noise from a UPS doesn't seem a likely culprit since you would expect it to be relatively constant. The first frame succeeding seems very similar to a problem I saw with my repeater before it went back into the junk pile. Do you use a repeater or a Boosterlinc or any device capable of reacting to the first frame, even if it's a status response?

Do you have a Monterey Analyzer? It think the signal dissect mode might be useful.

Why would they allow just a partial send? My UPS's, unfiltered, just suck signals into the black hole. I had a CFL that was putting out noise that showed up at about half a volt on both the Monterey and the ESM1 but that disappeared behind an X-10 5A plug in filter.

-- Bobby G.

These UPSs are behind the filters, I have a funny feeling they're interactiing with the filter to change the impedance characteristics of the filters. To be honest that line of thought is week. Seems more likely that the signals are being stomped on but by what I can't be sure of.

Here's what I do know, I've tried removing the signal amp/repeater and that has had no affect on the problem. On the Insteon side, I can send commands, they are acted apon but fail to receive the ack back. This leads me to beleive that something on the same phase as the transceiver is causing the problem. I should be getting a new unfiltered powerstrip this evening. I'll be able to move the UPSs to the other phase plus a few other items. My computer room has both phases in the same room.

A couple of thoughts for you Neil:

Since you see the first half of the message get through, power supply droop in the transmitter combined with a lot of background noise is a possibility.

Another thing to consider is noise from compact fluorescent light bulbs. I have several generic CF bulbs whose noise drops into the X10 frequency band near the zero crossing. With multiple bulbs on the same circuit, the signals beat together and almost look like X10 signals. I have seen them morph one X10 command into another.

Jeff