X10 Community Forum

General Home Automation => Other Automation Systems => Powerline Control Reliability => Topic started by: JeffVolp on June 11, 2013, 05:53:12 PM

Title: XTBM Noise Monitoring
Post by: JeffVolp on June 11, 2013, 05:53:12 PM
As many of you know, powerline noise can be a serious problem for any X10 modules that do not incorporate AGC to raise the detection threshold above the noise level.  While the XTBM and XTBM-Pro are already good at tracking down noise generators, Im working on improving their noise reporting even more.

There are three main types of powerline noise:

1)  Random switching noise from modular power supplies and other electronic devices

2)  Transients from devices turning on or off during the X10 window 

3)  Continuous carrier noise from devices like a wireless intercom or baby monitor

An X10 transmission is comprised of a series 1mS long bursts of 120KHz injected onto the powerline after each zero crossing of the AC waveform.  Presence of a burst signifies a logic 1, and absence a logic 0.  Strong in-band noise can also be recognized as logic 1s, making it impossible for an X10 module to decode the command.

The XTBM presently does a good job detecting random switching noise and continuous carrier noise.  However, transients are difficult to quantify.

The X10 spec requires at least 48 cycles of signal inside the X10 detection window to register as a logic 1.  The XTBM presently ignores noise bursts that are significantly shorter.  Unfortunately, that can give strange results when a strong noise source is at the cutoff.  In that case it could appear as an intermittent source as the number of cycles within the window varies.

To provide even more information, Im working on a firmware mod that will identify transients with a T, and continuous or random switching noise with an N.  And unless the noise is relatively constant, it will display the number of cycles inside the X10 reception window instead of the noise frequency.  It will also display the peak value of the noise, and the user can decide whether there are enough cycles to be concerned.  That will prevent the noise from appearing intermittent when it is near the cutoff cycle count.

It is interesting how the XPCP and various X10 receiving stages respond to powerline transients.  A strong spike on the powerline can cause them to ring for a number of cycles.  So even though the transient itself only is a single spike, it could induce ringing in the input stage for many more cycles.  That is even more reason to be wary of powerline transients.

Another firmware change will prevent the longer transmission pulses from the X10 XPCR and similar Leviton repeaters from being reported as noise when they stretch out beyond the X10 transmission window.

What Id like to do here is create a dialog with some of you folks who have been using the XTBM or XTBM-Pro to see if there are any other issues that I should address.

I plan to complete this update by mid-summer, and it can be installed by just swapping the PIC chip.

Title: Re: XTBM Noise Monitoring
Post by: dhouston on June 11, 2013, 08:06:03 PM
Have you tested the 48 cycle rule using X10 lamp/appliance devices?

To the best of my knowledge the first mention of 48 cycles within the X10 acceptance window was in the X10dedtm Protocol documentation that was published about the same time as the CM15A was released.
Data is accepted bit by bit as the presence or absence of 120Khz carrier occurring after the positive or negative mains zero crossing. The acceptance window begins approximately 250 usecs. and ends approximately 900 usecs. after a zero crossing. In this window 48 or more cycle of carrier are accepted as a "1" bit and fewer than 48 as a "0" bit.
Interestingly, the CM15A ignores this rule.
Title: Re: XTBM Noise Monitoring
Post by: JeffVolp on June 12, 2013, 12:07:26 AM
Have you tested the 48 cycle rule using X10 lamp/appliance devices?

No I have not.

I read that 48 cycle spec a long time ago.  It might have been in the extended protocol document, or perhaps in one of the articles on X10 automation in magazines like Radio Electronics or Popular Electronics years ago.  I also recall a technology paper (not the one on the PSC05) written by Dave Rye years ago, but I can't find a copy on my PC.

A quick Google search turned up the 48 cycle number also quoted in the TesterLinc documentation.

Title: Re: XTBM Noise Monitoring
Post by: JeffVolp on June 14, 2013, 03:16:10 PM
The firmware update to improve XTBM noise monitoring has been completed.  I'd like to find a few people with noise problems who have the original XTBM, and would like to evaluate the changes.  Please contact me directly if you'd like a free update.

The update for the Pro is a month or so away.

Title: Re: XTBM Noise Monitoring
Post by: JeffVolp on July 09, 2013, 10:53:26 AM

The firmware for both the XTBM and XTBM-Pro has been modified to improve their noise detection capability.  As described earlier, they now display the cycle count for random and transient noise.  And they no longer report as noise the extended X10 signal bursts produced by the X10 and similar Leviton repeaters.

The XTBM-Pro update includes several other changes that were suggested by users.

Some of the noise processing code has been ported over into the XTB-232 and XTB-523 so they are better at handling pulsating noise that looks like an X10 command.

You may contact me directly to obtain an update for your unit.