Re: X-10 Mister House Motion sensor problems

[nobody@xxxxxxxxxxxx (Dave Houston)]

Larry Moss <moss@xxxxxxxxxxxx> wrote:

>On 2006-06-07, Robert Green <ROBERT_GREEN1963@xxxxxxxxx> wrote:
>> May I ask you to start again from the top to explain exactly what sort of
>> behavior you want or expect to see from your HW and SW?  I'd just like to
>> get a resummation of the basic problem under a new thread heading so you can
>> get the maximum number of eyes reading about your problem.
>
>Sure.  The basic "problem" is that when I walk into my laundry/utility room
>with my arms full, I want the lights to turn on via motion sensors.
>There's about a 4 second delay before the lights come on.  It's not a
>serious problem.  I just take a few steps in the dark knowing the lights
>will kick on.  I'm just trying to figure out why it's taking so long.  It
>hasn't bothered me enough to remove the motion sensors, but I find it
>somewhat irritating.  I just want a faster response.  I could add another
>sensor at the top of the stairs so by the time I get down to the basement
>the light in the other room is already on.  But that isn't identifying the
>source of the delay.  That's just working around it, and not efficiently
>since I might not be going into the utility room.  (If I'm not carrying
>laundry or a lot of tools, I can flip a light switch easily, so I haven't
>added more motion controlled lights in other areas.)
>
>Here are the details:
>
>The motion sensor is an rms18.  There are actually two.  I don't think it's
>ever possible for both to see me at the same time.  The transceiver is an
>rr501.  The computer is connected to a cm11a.
>
>I haven't figured out a good way to accurately time the delay when I walk
>in the room.  Right now, just looking at a watch as I enter the room, I'm
>seeing about 4 seconds after I pass the first sensor before I see light.
>In the MH log, I'm seeing the computer recognizing the MS and turning on
>the light either the same second or the following second.  The timer
>resolution shown in the log is only a second.  Not sure if I can get it to
>show finer resolution than that, but I'm not sure that's necessary anyway.
>It clearly takes less than a second between the time the computer sees the
>signal on the power line and responds.  Hmm, I just added something else to
>the log.  It was printing to the log that it was turning the light on and
>then sending the command to the cm11a.  So I added a print statement after
>the ON command in case the delay was in executing that.  What I'm generally
>seeing now is MisterHouse sees the signal from the MS, reports that it's
>going to turn on the light and then that it has turned on the light all
>within the same second.
>
>So, the computer/software shouldn't be the delay.  Unless there's some
>oddity in the cm11a driver MH is using.  I suppose it's possible that the
>main MH program hands the command off to a driver that then sits on it.
>I'm not familiar at all with MH internals, so I don't know how it's
>handling that.  I'll have to post some of this to the MH mailing list if no
>one here has a good answer for me.

Motion sensors sometimes are slow responding to movement. The response time
is hard to determine.

IIRC, the RR501 waits until it sees the start of the next RF code before
sending to the powerline. This takes about 108mS. If the signal is a little
weak, the RR501 will use the first (or subsequent) copy (or copies) to set
its threshold. Each copy takes ~108mS.

It takes 94 half cycles of 60Hz for two PLC copies of two X-10 codes. (e.g.
A1, AON) The CM11A will wait for both copies plus ????. This takes at least
783mS.

Mr. House interprets the incoming code and sends the macro it triggers.
There's some handshaking with the CM11A before it transmits. It then takes
the same number of powerline half cycles to sent two copies of two codes.
(e.g. B1, BON)

Watching your ESM1 may give you a picture of when the delays occur.

The simplest way to speed things up is to have the RR501 turn on the light
and have Mr. House merely log the activity.