Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[JeffVolp]

Jeff, does it make sense to you that an offending noise source could impact only the CM15A clock and nothing else regarding the X10 network of devices? And would you think that the offending noise source is most likely within our house?

If only the CM15A is drifting (and that is true for three different units), then the 60Hz powerline frequency is fine.  While I don't think we can totally rule out some sort of software quirk, my guess is that somehow noise is getting into its internal clock, and causing it to count extra ticks.

How that is happening really depends on the firmware inside the unit, and where in the cycle it might be susceptible to noise causing an extra count.  Most devices use zero crossings of the powerline for their time reference.  If the CM15A firmware does not have some sort of gate to prevent double counting, then a large transient near the zero crossing might cause an extra count.  That might be caused by a heavily loaded dimmer near minimum or maximum brightness, or a solid-state relay controlling a heavy load (switching on just after the zero crossing.)

There is really no easy way to identify whether the noise source is inside your home or coming in over the powerline.  The new washer/dryer set might be generating noise when it is running.  Can you correlate the time shift with days that equipment is running?

Jeff

[bvw]

Reply to Brian H:
I just shut my PC down and leave the CM15A plugged in and the USB cable plugged into both the CM15A & the PC.

Reply to Jeff:
We don't use the washer/dryer every day, yet the timers will trigger earlier & earlier each and every day (and not by any obvious constant factor that I can figure).
I plan on swapping out all our CFLs this weekend with incandescent to see if they can be ruled out. After that, I just don't know what else to test at this point.
Nothing else new has been added to the household so perhaps some electrical device might be going bad and causing interference that I am not yet aware of.

[bvw]

Current Status:
All CFLs were removed from household and all timers utilizing any form of variable time (dusk, dawn, security) were deleted.
This resulted in no change - all timers continue to trigger earlier & earlier each day.
If anyone has any further advice or direction for me it would be greatly appreciated.
Thanks.

[Dan Lawrence]

One other question:  Do you have a filter on the PC's line cord? If so, could the filter have failed?   I have a filter on the line that the PC and  our two printers are on.  The CM15A is plugged into an outlet that is fed to a different breaker than the one that powers the PC. 

I have had AHP since 2005, I had a CM11A before that, never had any time shift.

[pconroy]

and not by any obvious constant factor that I can figure

well that's the answer to the question i was going to ask.

You said three CM15s and all are doing this?
PC's clock is not drifting either?

Where were the CM15s bought from?

[-Bill- (of wgjohns.com)]

Bump!bvw,

From what you posted above, it appears you have "Security" checked on the timers, or am I misreading that?

If you do, then uncheck it.

The purpose of "Security" is to make the timers drift so they don't always happen at the same time and gives your house a more "natural" lived in look when you're away.

The variation is to make it appear that a person is controlling the lights and not a timer. 
 

[Noam]

... From what you posted above, it appears you have "Security" checked on the timers, or am I misreading that?

If you do, then uncheck it.

The purpose of "Security" is to make the timers drift so they don't always happen at the same time and gives your house a more "natural" lived in look when you're away.

The variation is to make it appear that a person is controlling the lights and not a timer. 
 

I would have agreed at first, but bvw reported that the timers keep getting earlier each day. I thought that "security" mode simply shifted them a bit, in either direction, but kept them within a certain window of the original time. It sounds like the timers are going outside of the window, as they execute earlier and earlier each day.

[bvw]

Reply to Dan:
I have a filter on the UPS that my PC is plugged into. I will swap it out for another in case it is defective. The UPS and CM15A are on different circuits but are on the same circuit breaker phase.  My AHP network has run flawlessly since I upgraded to it when it was first introduced (I have been on x10 for many years, three different homes), until a few months ago when the early triggers began.

Reply to PConroy:
The PC clock is not drifting.
All but one CM15A has been purchased from X10, one was purchased from TheHomeAutomationStore.

Reply to Bill & Noam:
Noam is correct in that the timers shift in only one direction, more and more each day.
I did, however, delete all timers using security, dusk, or dawn shifting during this troubleshooting phase in order to eliminate those time shifting variables.

What I have observed so far is that whatever noise that may be on the powerline is only impacting the CM15A clock. All timers and RF macros are being triggered and there are no phantom triggers occurring.  Each time I boot up the connected PC, it does in fact reset the CM15A clock.
After I shut down, the first few timers will trigger as programmed and then each one after that will trigger earlier and earlier. What could be generating just enough noise to impact the clock only? One would think that that would take a great deal of noise.
Anyway, even though this is a new install of the AHP software, I'm going to install the software on another PC and start out fresh just to see what happens.

[Dan Lawrence]

Reply to Dan:
I have a filter on the UPS that my PC is plugged into. I will swap it out for another in case it is defective. The UPS and CM15A are on different circuits but are on the same circuit breaker phase.  My AHP network has run flawlessly since I upgraded to it when it was first introduced (I have been on x10 for many years, three different homes), until a few months ago when the early triggers began.

Some UPS's will start putting various types of noise on the powerline after a while. Since the UPS and the CM15 are on the same phase, that may be the source.  Unless you really need the UPS, remove it and see what happens.

[MichaelHlubb]

Each time I boot up the connected PC, it does in fact reset the CM15A clock.
After I shut down, the first few timers will trigger as programmed and then each one after that will trigger earlier and earlier.

Based on the above, why not "fix"the problem by leaving the computer on and attached to the interface? In fact,  some of the addons to AHP require the computer to be running to work, i.e. they are not stored in / run from the interface. 

[AZDoug]

I moved "up" to the CM15 in Dec '09 after more than a decade of using CM11s.  Almost immediately I noticed that the clock was moving at an unacceptable rate.  I added a chime and set it off 3 times a day to track the clock movement.  The worst that I can remember was a 2 minute change from 8AM to Noon. 

The timers were triggering earlier, which is consistent with noise on the powerline.  I tried moving the CM15 to other outlets around the house with no success.   After going through about everything recommended here and other places, the clock still moved.

Some time in the Spring, I installed a CM11 with similar timers and a different chime.  The first thing I noticed was that the CM11 was rock solid.  After a while I noticed that the CM11 clock slowed slightly, while the CM15 ran fast.  I used an LCD WWV-synced clock as well as a regularly synced PC clock to check the clock drift. 

And then, only a short while into the experiment the CM15 clock quit moving.  NOTHING in the house changed (that we could perceive).  The chimes are still sounding and the CM15 clock accuracy is holding as well as I could hope for. 

While I was having the problem, I plugged in an older X10 LED clock/timer that I had bought (unused) at an estate sale.  The clock on it also advanced, consistent with noise on the power line. 
I wish I could say I had a good handle on the cause or solution, but I don't.  The only thing I'm certain of is that if it returns, I'm going to dig out my CM11s again.

[JeffVolp]

If this is indeed caused by powerline noise, you might want to try a beta version of my "tuned signal-sucker".  It significantly attenuates powerline noise except right at the X10 carrier frequency.  It does attenuate X10 signals slightly, and is best paired with a signal booster to insure the signal levels are well above the background noise level.  However, plugging one in near your CM15A may help with the clock drift.  Contact me through a PM if you want to pursue this.

Jeff

[bvw]

Thank you AZDoug for sharing your experience. Your note about the old X10 clock/timer also advancing was telling. Is it also true that the "noise" issue you were having at the time only impacted the clock and nothing else, like me? I will keep your CM11 experiment in mind as I continue forward.

Reply to JeffVolp:
This morning I installed AHP (v 3.271) and built a new .ahx on a laptop that had never had an AHP install previous. This is to ensure, I hope, once and forall that this is not a software issue.
I am very interested in your offer (I do not own a signal booster). I will contact you early next week after I know the results of my laptop experiment. Thanks.

[dave w]

I moved "up" to the CM15 in Dec '09 after more than a decade of using CM11s.  Almost immediately I noticed that the clock was moving at an unacceptable rate.  I added a chime and set it off 3 times a day to track the clock movement.  The worst that I can remember was a 2 minute change from 8AM to Noon. 

The timers were triggering earlier, which is consistent with noise on the powerline.  I tried moving the CM15 to other outlets around the house with no success.   After going through about everything recommended here and other places, the clock still moved.

Some time in the Spring, I installed a CM11 with similar timers and a different chime.  The first thing I noticed was that the CM11 was rock solid.  After a while I noticed that the CM11 clock slowed slightly, while the CM15 ran fast.  I used an LCD WWV-synced clock as well as a regularly synced PC clock to check the clock drift. 

And then, only a short while into the experiment the CM15 clock quit moving.  NOTHING in the house changed (that we could perceive).  The chimes are still sounding and the CM15 clock accuracy is holding as well as I could hope for. 

While I was having the problem, I plugged in an older X10 LED clock/timer that I had bought (unused) at an estate sale.  The clock on it also advanced, consistent with noise on the power line. 
I wish I could say I had a good handle on the cause or solution, but I don't.  The only thing I'm certain of is that if it returns, I'm going to dig out my CM11s again.

Very logical troubleshooting. Well thought out. It is too bad you did not discover what was causing the original drift. Since you did not find it, my hunch is it will be back.

[AZDoug]

[Very logical troubleshooting. Well thought out. It is too bad you did not discover what was causing the original drift. Since you did not find it, my hunch is it will be back.

I couldn't agree with you more.  I concluded several decades ago that there is no such thing as a "one-time-occurence" when it comes to glitches in electronic equipment.   Without some tools, such as Jeff mentions, it's pretty difficult to do more than guess at what might be going on.  My electronic troubling tools consist of a couple old VOMs and, of late, a DVM.

I was more than a bit irked when I read somewhere that this or other products of X10 were value engineered by removing extraneous filter caps here and there.  I don't know if it is true.  One of these days I'm going to open my CM15 and see how many blank component locations are present, expecially those that have a Cxx designator on the board. 

When the source of timing is a 60 Hz 100 volt signal, how expensive is it to wrap some filtering around it to increase the odds that it and only it gets counted?  I used 60 Hz-based digital clocks for many decades, and I don't remember ever seeing one do anything like this.  Of course, most are no longer operating in today's environment, but at the time, I always considered them almost as stable as WWV.   My mother's house still has two or 3 such clocks, and they never have a problem. 

Sorry I can't offer more.