Random events?

[Etienne]

Hi ... been using HG for a year or so ... anyway ... was rock-solid up to a month or so ago - after the latest couple of updates (not sure since when thou') I started noticing some weird behaviour - for example I would find one of my X10 devices in HG is displaying a status of ON - when in fact no-one turned it on ... and even stranger (i think) the switch will actually be off.

So .. TLDR:

1. Switch would be off both in reality and on HG
2. Once a while (max a couple of times a day) i check status and find Swich is on in HG .. but still off in reality

Code: [Select]

2015-02-12T18:29:52.9428760+01:00 HomeAutomation.X10 E1 - Status.Level 1
Just happened a few mins ago - log above!

Ideas?

[bkenobi]

You'll need to provide more details than the homegenie.log file.  If you capture the event in the mono output, you should have enough to debug yourself or post here for help.  That will give you a better idea of the source of the command.

[Etienne]

If you capture the event in the mono output, you should have enough to debug yourself or post here for help.

Any quick tips on how I go about doing that?

[bkenobi]

Depends on if you are running HG on Windows or Linux. 

For Windows you need to stop the HG service (select the icon and pick stop).  Then from a command line, start HomeGenie.exe.  It will stream the mono output.  You can redirect this to a file if you want to make reviewing the output easier.

For Linux, I modify the startup.sh so that it doesn't redirect the output to /dev/null by giving a log path such as /usr/local/bin/homegenie/log/mono.log.  You can also just run it directly by stopping the service and calling the HG executable but again, reviewing output from a file is a bit easier IMO.

[Etienne]

So ... only got the time & will to setup logging yesterday ... so - today it happened again ... at 1829 my X10 E10 switch was showing a state of on, when in fact no-one triggered it on and actually it was really off.

This time I have the log for that time thou'

Code: [Select]

[18:29:54.228952] X10 > 5B 00 97
[18:29:54.229004] X10 > MACRO: 5B 00 97
[18:29:59.380899] X10 > 5A 03 02 D2 65
[18:29:59.380960] X10 > PLCRX: 5A 03 02 D2 65
[18:29:59.380987] X10 > FNMAP: 01 00
[18:29:59.381004] X10 >  DATA: 65 D2
[18:29:59.381044] X10 >       0) House code = A
[18:29:59.381059] X10 >       0)    Command = BRIGHT
[18:29:59.908889] X10 > 55
[18:29:59.908946] X10 < 00

Does that mean anything to anyone ... please note the following ... when this happens it's not always X10 E10 that's switch to status on - some times it's X10 E10 ... others it's X10 E1 .. BUT never X10 E3 - the only difference being E3 is setup as a 'light' the other two as a 'switch'.

Furthermore - the log seems to be talking about house-code A ... which is not in use at all - i'm on E.

As mentioned before - I've been using HG for a year or so and this started happening a few updates ago.

Cheers!

[bkenobi]

The log says that the source of the command was a macro.  I assume that you must be running a CM15A since it can store macros programed into it from ActiveHome Pro.  Did you use your controller with AHP in the past?  Or, if it's a second hand controller, have you purged timers and macros from the module's memory?

[Etienne]

Yes - I'm using a CM15 PRO ... the controller was purchased new a year or so ago - and was only ever used with HG; initially hosted on a RaspberryPi, and then a few months ago moved to an x86 Debian server.... the random events started a few weeks ago.

Today it happened again - I've pasted the log file here ... that's 3 days of logs since when I started the logging - I've added labels for each day ... have a look at DAY 3 - at about 0030 I power cycled the CM15 and the logs shows loads of errors - not sure if this is normal.

At 0615 - as per my schedule an x10 switch was triggered (good)
at 0700 - some macro did something !?
at 0815 - schedule - turn off x10 which was enable at 0615 (good)
0930 & 1830 - again ... random stuff?

Any ideas? Maybe I should clear the CM15 just in case? Any idea how this can be done from HG and/or debian? Or should i connect it to a Windows machine?

Help appreciated.

[enterprised]

silly question, but do you have a close by neighbor who also uses X10 devices. It is a known issue that their can be cross-over from one house to another.

 

[bkenobi]

Close neighbors would be my next guess then.  A is very common for people to use until they run exceed 16 modules.  Also, 7AM seems like the time for an automated coffee maker to turn on.   

If this is neighbor's system interfering, there are 2 options that I'm aware of.

1) Make sure you don't use the same House Code.
2) If 1 doesn't work, source a whole house filter (PZZ01).

The only other thing this could be is a Smart Meter or some other interfering device (and it's not because this is not because the signal is called a MACRO...meaning not a random code but a designed signal from somewhere).

[Etienne]

I would be very very surprised it's a neighbour ... so I'm gonna rule that out - also I'm on house-code E ... not A (did that on purpose).

As for Smart-Meter ... yes - we do have one of those; but it was installed a couple of years ago!

So - couple of questions - is there an easy way to reset the CM15 from HG / debian ... and as a last option windows.

If that doesn't help - is there some filter which help cos' of smart-meter?

[bkenobi]

The only way I know of to reset CM15A is to use AHP.  There is a free version available from Marmitek that will work on Windows.

http://www.marmitek.com/

X10 also has one, but I believe it still requires some modification to the package to get it working due to server issues (long story).  It's also free and works in the US but if all you want to do is reset the controller, I'd go with the link above.

I don't know that your issue is Smart Meter related.  What I do know is that there was an issue with a certain meter used in Cincinnati's power district that caused a lot of issues when initially installed.  The PUD did not provide a solution to my knowledge but Jeff Volp did come up with a solution.  This is the solution posted on his forum that provides information about the issue and the solution.

http://jvde.us/forum/index.php?topic=9.0

From a brief review, it looks like the meter killed X10 more than created false signals.  I still suspect this is either something coded in your CM15A or a neighbor.