Please login or register.

Login with username, password and session length

Author Topic: looked like a software bug but wasn't  (Read 5394 times)

marty619

  • Jr. Member
  • **
  • Helpful Post Rating: 3
  • Posts: 12
looked like a software bug but wasn't
« on: March 02, 2008, 09:42:25 AM »

I have spent a couple of years annoyed at my CM15A's failure to operate some of my circuits once a macro was run. Yesterday, I have found that the problem was not a software fault at all, and wish to share this info with the community.

The environment
I have 15 modules, and an XPCR repeater coupler. My TV's etc. are on line filters. I also have line voltage landscape spot lights, and two 40 Watt lamps on driveway columns, which are controlled by a photocell module.

The Problem
I have a macro which turns on the landscape lighting at dusk + 20 minutes and off at 11:00PM. Once this macro runs, commands from my palmpads did not easily control any other circuits or even the OFF command for the Landscape. I had to push the palmpad's buttons a dozen or more times to get any results, whereas during the day everything worked with a single push. Also, there are 5 circuits to be energized for the landscaping and not all came on or went off reliably.

What I thought was the problem
I suspected a neighbor's night time use of noise generating equipment. I suspected that something in the macro caused bad interpretation of the palmpad RF signals, and to that end I tried numerous changes to the macro by using appliance modules instead of the wall switch or soket modules. Appliance modules used in macros issue commands like A5 AON A6 AON whereas the house switches etc. will issue A5 A6 AON. Using the appliance module controls did help a little in getting reliable operation of the various zones. Putting a 2 sec delay between each module also improved reliability in getting each landscape circuit to be controlled, but neither helped with the control of other x10 circuits once the landscape came on.

What I think is happening
I did a test and filtered the line to the driveway column lights and I am now able to control other circuits reliably. My conclusion is that the triac and circuitry in the photo sensor started to broadcast noise once it was dark, the same time the landscape came on, a coincidence that threw me. In thinking about it, Triacs turn off at 0 current crossing and the X10 circuits issue commands at 0 Voltage crossing which are in sync with just resistive loads.

Additionally, in thinking about it, once the landscape is switched on ( by an appliance module.. a contact closure) I am adding 1000+ feet of cable to my whole system, so signal absoprtion increases in the ON state. So a given x10 signal which is say marginal in strenth, broadcast from it's sender might get a circuit to turn on if it's control module is located "nearby", but once the circuit is energized and a significant extra cable length is added, the signal strength drops due to absorbption, and you can't get it to turn off.

So my advice is to watch for the very loads you are switching as the source of problems once they are on.
Logged

Oldtimer

  • Advanced Member
  • Hero Member
  • ******
  • Helpful Post Rating: 45
  • Posts: 364
  • Line Noise & Signal Sucker Fatigue Syndrome
    • Oldtimer's X10 Forum Home Page
Re: looked like a software bug but wasn't
« Reply #1 on: March 02, 2008, 09:50:29 AM »

Very clear write up.  Thank you for taking the time to post this.  I had exactly the same problem with my living room branch circuit when the outlets were turned ON by X10.  It wasn't 1000 feet of wiring but was still enough of a rats nest to give me grief.
Logged
Having "fun" with X-10 since 1980.
 

X10.com | About X10 | X10 Security Systems | Cameras| Package Deals
© Copyright 2014-2016 X10.com All rights reserved.