X10 Community Forum

🖥️ActiveHome Pro => ActiveHome Pro General => Topic started by: arf1410 on June 15, 2005, 06:10:40 PM

Title: Timers and Macros Quitting After "X" Days
Post by: arf1410 on June 15, 2005, 06:10:40 PM
There is another post (in the main 3.187
update post) regarding timers and macros
that stop working after a couple of days.
Scattered throughout these forums, there are
numerous people who have experienced this
same type of problem, including myself.  Can
anyone (X10 Pro in particular), come up with
a plausable explanation as to why the CM15A
would stop working after a period of time,
after successfully functioning for several
days - or longer ??????
Title: Re: Timers and Macros Quitting After "X" Days
Post by: arf1410 on June 16, 2005, 04:33:42 PM
X10 Pro - Any thoughts on this one?
Title: Re: Timers and Macros Quitting After "X" Days
Post by: murphy on June 17, 2005, 06:54:18 AM
If an external event causes the firmware in
the CM15A to go into a tight loop that
prevents it from processing it's normal
control loop, it would appear to have died.
A continuous stream of external events either
on the power line or via RF that arrive at a
rate that exceeds it's processing capability
would have the same effect.  I'm not familiar
with the microprocessor in the CM15A but if
it uses a push down stack and exceeds the
space alloted for the stack, due to a large
number of unprocessed events, it would
probably stop working until it was reset.
If you have a repeater I would turn it off to
see if it is involved in this failure mode.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: carmine pacifico on June 17, 2005, 09:50:54 AM
ARF1410

Why are you surprised?
The internal clock of the CM15a is software
based, noise on the power line will cause
it to speed up; RF command  at close
interval will cause it loose the clock, any
macros not depending on clock will still
function. This was the main reason why I
jumped ship. I installed the competitor
controller, set it up in mid April, it
never lost time, or missed any macros , and
we had power failure, thunder storm etc..
it keeps on ticking.
X10 Pro claims the CM15a is a solid
product, I posted several times the
hardware flaws of this product, in my line
of work I would consider this product in
Alpha stage( not even Beta) and the
customers are the unwilling tester. Having
software (and firmware up upgrade: Diag
units) every few weeks it’s not a sign of a
good product, but a sign of a desperate
attempt to fix the product with a lot of
Lemon-Aid on a badly designed product.
So far the following X10 product that I own
has lock up problem and require power down
to reset: TM751, RR501 and CM15a, repeated
RF command will do the trick every time.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: arf1410 on June 17, 2005, 10:36:18 AM
Though I don't regulary use RF, I do have a
coupler/repeater, and almost all my light
switches are switchlinc 2-ways, so are you
guys saying that it is all this incoming
activity that is causing my unit (and that
of others) to stop after "x" days?  If it is
this incoming activity, suppose I turned off
(unchecked) all the monitored housecodes
boxes in AHP - would that stop the CM15A
from attempting to process all these
incoming signals?
Title: Re: Timers and Macros Quitting After "X" Days
Post by: carmine pacifico on June 17, 2005, 10:55:10 AM
ARF1410

Unless they improved the hardware, the
CM15a I had was crashing and loose the
clock when it got too busy, this has been a
major problem with RF transceiver TM751 and
RR501 else, when I press the DIM or Bright
button on my palm remote. X10 Pro has
stated that “The CM15a +AHP + Macros works
as advertised”. The way I see it many
people in this forum use this product as a
hobby, and do not mind wasting countless
hours trying to get it to work, and
probably get a rush every time there is an
upgrade; my hobby is photography and
electronics, X10 products is just a tool to
automate some lights, it’s not reliable
enough therefore to me is useless even for
free.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: arf1410 on June 17, 2005, 10:58:34 AM
PS - I'm not surprised, but considering this
problem has been reported by numerous people
over the month.

/1/  I'd like to see X1- Pro's
acknowledgement that they have many reports
of this problem
/2/ I'd like to see X10's ideas of the cause
or at least tell us what they are doing to
try to resolve this issue.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: carmine pacifico on June 17, 2005, 11:09:19 AM
ARF1410

X10 Pro replied on your post:
”X10 Pro – Please comment on Post”.
The problems you are referring to are
related to the flaws I reported.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: SteveRF on June 17, 2005, 07:48:18 PM
ARF,
I also find it interesting that some of us
run for months and months with many macros
and timers firing right on time very day.
Although I may not have as many devices as
othersw, I do have bathrooms, utility rooms,
entrance ways and pool deck covered with
motion detectors that activate macros.  I
have timers for early morning, mid
morning..early afternoon, early evening and
late evening that are constantly changing
lighting.  All fire right on schedule.  The
only trouble I have had lately is response if
two motion detectors fires at or about the
same moment and an errant macro loop about a
week ago that took a clear reset to fix.  I
run the unit disconnected 24/7, run
XP/SP2/ATI Vid.  I wonder if I had a houseful
of kids triggering constantly if it would be
the same.   The disparity of problems is
interesting.
Regards,
SteveRF
Title: Re: Timers and Macros Quitting After "X" Days
Post by: arf1410 on June 17, 2005, 08:08:43 PM
I went back and reviewed that post, and i
don't think X10 Pro really did comment n
the question below...
Title: Re: Timers and Macros Quitting After "X" Days
Post by: paul mayfield on June 18, 2005, 10:25:38 AM
X-10 PRO: Is there going to be new hardware
released? I have been using the CM15A now
for 7 months with marginal success. I have
been late for work as the timers and macros
quit working (I usually get a week out of
them if I clear and download weekly). My
CM11 NEVER failed to get me up! I am now
getting a day to day and a hafl out of the
unit, timers and macros quit aqain last
night (good thing this is the weekend). I
can check the activity monitor and it shows
everything working as expected but the
lights are not responding. A manual
execution of a light on or off from the
program works ok, so the program is
responding. If I were to clear the interface
memory and leave the unit connected and PC
on (as I always do) would the system respond
and be more stable, or does it have to
reside in the interface? I have been very
patient but now I am getting annoyed with
this unit, I would think that by now some
major decisions would have been made. If the
unit becomes reliable I will be interested
in the I-Net plug in (I have smart macros),
if these MAJOR problems are not going to be
resolved I will give up and go to another
unit as some have. I will give it a few more
months, these problems have been identified
by many users over and over. What were the
results from the users with the diagnostic
units? Please post back! Thanks
Title: Re: Timers and Macros Quitting After "X" Days
Post by: paul mayfield on June 18, 2005, 10:41:57 AM
I answere my own question (see post below).
I cleared the interface memory and watched
to see if my macros would respond and they
don't! Would it be possible to allow users
to make a decision for the timers and macros
to operate from the software or reside in
the interface. Maybe this would increase the
reliability. Thanks
Title: Re: Timers and Macros Quitting After "X" Days
Post by: paul mayfield on June 18, 2005, 10:55:11 AM
See post below. With the interface memmory
cleared you can watch the program and see
the macros trigger so why do these not work
unless they have been downloaded to the
interface?
Title: Re: Timers and Macros Quitting After "X" Days
Post by: murphy on June 19, 2005, 06:41:45 AM
Did you click the radio button in the macro
designer for each macro that says "Run from
PC"?  
Title: Re: Timers and Macros Quitting After "X" Days
Post by: paul mayfield on June 19, 2005, 08:55:13 AM
No I didn't go back to the designer (I
didn't know that function existed), have
been running the same macros since Dec. Will
the timers still execute (maybe there is a
radio button for that too, I will look).
Thanks
Title: Re: Timers and Macros Quitting After "X" Days
Post by: carmine pacifico on June 21, 2005, 10:20:31 AM
To answer Steve RF question “ some are
running for months without any problems ….”
It’s not how many macros, lights or motion
control you have, it the complexity of the
macros that can cause problem. I only had 3
light and 3 motion on my system during test:
2 motions controlling the stairs lights
(top and Bottom)
1 motion for the bathroom.
The stair macros would use dim or bright to
the light and during night time would else
turn control the kitchen light using
Dim/Bright command.

The bathroom light else is using Dim/Bright
command.
I else had a recovery macros using the
commands from the motion photocell: in the
event the light where manually turn off.
This set up generate a large number of
commands on the power line and RF, unless
the CM15a (or RR501, V572) has a large
buffer to store more than 1 RF command, the
macros will not work properly.
Usually if I triggered the stairs motion
(stairs light and Kitchen light On  with
multiple Bright commands) and then trigger
the bathroom motion within 10 seconds, the
CM15a would fail and loose the clock( it
takes more than 10 seconds to send all
those Bright/Dim on the power line).
With the 1132CU I eliminate the crashes and
loss of clock, but to achieve the
performance I wanted I had to replace the
wall switches with “Preset-Dim” switches to
reduce traffic on the power line.
In a nut shell if you keep you macros
simple and minimize traffic on the power
line ad RF signals, it will probably work
flawless.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: arf1410 on June 21, 2005, 10:31:58 AM
X10-Pro -

Though my timers are simple, and I don't use
any macros or RF, is it possible all
the "activity" generated by my 2-way
switches (which send out a signal every time
they are manually operated) is what is
causing my CM15A to stop functioning after a
couple of days?
Title: Re: Timers and Macros Quitting After "X" Days
Post by: andre on August 20, 2005, 11:35:59 PM
I seem to be seeing a common problem where
the CM15A looses it's mind and no longer
properly processes timers.  A friend of
mine is also experiencing the same
symptoms.  Unplugging the unit from the
wall and removing the batteries is the only
way to restore USB communications to the
unit in order to restore the timers.
Assuming X-10 is monitoring this forum, I'm
challenging X-10 to provide a suitable
explanation as to why so many user's
patience are being challenged by this
product.  Is there a firmware update
available that we can use to reflash our
CM15As?
Title: Re: Timers and Macros Quitting After "X" Days
Post by: arf1410 on August 21, 2005, 07:27:28 PM
Andre -
my understanding is that the firmware in
this unit is NOT upgradeable ...
After 6-9 months of asking , X10 Pro has
come up with no explanation or potental fix
to the unit malfunctioning of a period of
time ...
Title: Re: Timers and Macros Quitting After "X" Days
Post by: SteveRF on August 21, 2005, 08:01:06 PM
Andre,
My CM15A did the same thing but was swapped
out for a diag unit (04J44 4 Mar) way back in
Mid Mar.  I have ramped up a bit since then
to include more timers and lots more If/Else
macros.  I use the unit disconnected from the
PC 24/7 and connect only to reprogram.  Even
with lots of recent power outages (some 2-3
hours) I have noticed all my timers and
macros fire on time (on the minute) and as
expected.  I still suffer from some several
second delays when several motion detectors
fire at the same time but I dont seem to
loose timer or macro control and definitely
do not loose PC connectivity.  The original
unit was awful...lost time, reset...stopped
working..you name it and it did it... now I
wonder why mine seems to be so stable when so
many have trouble.. I do set it up for No A
or P codes... only transceive codes I am
using B,C,D,E,and F and run the unit
disconnected fromt eh PC exclusively.  I have
been looking at adding Cams to the set up but
am not yet convinced of the cam quality...
Let me know if you find a resolution (other
than shooting the unit)...
SteveRF
Title: Re: Timers and Macros Quitting After "X" Days
Post by: andre on August 22, 2005, 06:45:45 PM
Thanks for the replies, ARF1410 and
SteveRF.  The loss of sanity can't be
attributed to transceived codes or complex
macros as neither myself or my suffering
friend run more than simple timers. I am
running timers that are located on 2
different house codes, H and I. It's quite
disconcerting to find the aquarium lights
going on at 2am instead of 7:30, or the
porch light going off at 8pm instead of the
programmed 6am.  As for power outages,
we've had stable power for months without
so much as a neighbourhood brownout.  I
have been able to tell because my Kenmore
microwave resets as soon as there is any
inconsistency in the line voltage.  One
note is that I run my CM15A disconnected
from the PC except for programming. I'm not
exactly looking forward to pulling out my
scope and debugging yet another poorly
designed consumer product.  Has anyone seen
schematics of the CM15A posted on the web?  
Title: Re: Timers and Macros Quitting After "X" Days
Post by: murphy on August 22, 2005, 07:16:09 PM
Go to fcc.gov and search for the type
acceptance number that is on the CM15A.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: andre on August 22, 2005, 10:25:15 PM
Hi SteveRF, I took a peak at my CM15A and
it has the same batch code as yours, 04J44.

To Murphy,
FCC, Industry Canada, UL and CSA filings
are confidential because they usually
contain proprietary data, so why did you
think that the FCC listing number would
yeild schematics?
Title: Re: Timers and Macros Quitting After "X" Days
Post by: murphy on August 22, 2005, 10:45:13 PM
Could it be because I downloaded the
schematic from there.  Note that fcc.gov
doesn't work.  It has to be www.fcc.gov

Look here:

http://makeashorterlink.com/?R1A4229AB

Title: Re: Timers and Macros Quitting After "X" Days
Post by: andre on August 22, 2005, 10:56:53 PM
Hi Murphy,
My appologies, and thanks for the link. I
see right off that the submitted schematics
are not totally true to the actual shipped
product.  The RF daughterboard has been
sub'ed out for a 2 board pair, each with
their own antenna.  It appears that the
module has been split into separate
transmit and receive modules.  The Cypress
microcontroller is consistent.  I'm goingto
start by having a deeper look at the power
supply section because that's where most
product's woes are located. It also appears
that there's no watchdog, but I still have
to download the datasheet for the
microcontroller.
Has anyone else explored the design?
Title: Re: Timers and Macros Quitting After "X" Days
Post by: carmine pacifico on August 23, 2005, 10:43:05 AM
Andre

The microcontroller is cypress Encore
series, it is obsolete and not recommended
for new designs. Apparently Cypress bought
the company that originally made the Encore
USB chip. This processor has some serious
hardware flaws in the silicon.
The Encore II is PSOC based  and should be
a replacement for the original Encore chip(
not pin and firmware compatible) but
unofficially my Cypress rep does not
recommend it, he suggested the new PSOC
with USB , this chip it’s not available in
production quantity yet.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: roger1818 on August 23, 2005, 12:47:06 PM
Andre:  It sounds like you are having
problems with powerline noise causing the
clock to run too fast.  X10 Pro has
admitted that some users suffer from this
problem, but despite the fact that I showed
him exactly where on the schematic the
problem lies, he claims that it isn’t a
bug.  I had this problem for about a week
about six months ago, but it seems to be
running on time now (I am thinking the
noise may have been coming from an HP
DeskJet but I don’t know for sure).

If you want to tinker with yours, you might
want to try putting a capacitor in parallel
with D1 or an inductor in series with R1.
You will probably have to try several
different values until you find one that
works well.  You want to make TR10 provide
a clean 60Hz square wave to U4.  Currently
if there is high freqency noise while TR10
is in its active region it can produce an
extra pulse causing the clock to gain time.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: Mike C on August 23, 2005, 08:14:14 PM
I have had the same problem with my unit
going to sleep after a number of weeks.
Somtimes days if I unplug it from the PC.
This past weekend I came home to a dark
house. I unplug the unit, pull the batterys
and wait a few minutes. Plug it back in run
the software and download the data.
Everything works again.  I have sent several
emails to x10 requesting a replacement unit
but get no response.  I guess my next post
will be sometime in Sept when it fails again!
Title: Re: Timers and Macros Quitting After "X" Days
Post by: andre on September 07, 2005, 06:49:13 PM
Hello everyone,
I've added the decoupling to my CM15A,
and it has retained its sanity for the past
week which is better that in the past.  The
components I added are 0.1uF 50V capacitors
across the following components:
U5: IN to COM
U5: OUT to COM
U2: pin 1 to pin 8
U1: pin 6 to pin 11
I checked the signal coming from U4 and it
needs no additional filtering.  The base-
emitter junction of TR10 only see the
botton 0.7 volts of the AC line so any
noise on the waveform is filtered out.  The
data sheet for the microcontroller is an
interesting read, it has the power-up reset
circuit and watchdog that I observed was
missing from the schematic of the CM15A,
and Cypress confirmed for me that they had
no issues in their database regarding the
use of either of those features.
If anyone has additional mods they'd like
me to explore, pass them along.  It's the
most fun I've had at work in months.
Cheers.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: roger1818 on September 09, 2005, 01:54:09 PM
Andre:  Thanks for posting your findings.
I can't believe X10 didn't put any
decoupling caps on U1 and U2!  That alone
could be causing lots of problems.  Did you
use tantalum caps?

I am curious why you needed to add caps
across U5.  Don't C11 and C12 serve this
purpose?  Are they not physically close to
U5 (I haven't taken my CM15A apart yet)?

I will have to give these mods a try on my
old (pre-diagnostic) unit.  Thanks again.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: andre on September 09, 2005, 06:31:51 PM
Hi Roger, The caps I used were ceramic type
X7R.  These are good high frequency
decoupling caps.  The electrolytic caps C11
and C12 have too high a series inductance
and are ineffective at high frequencies.
The linear regulator U5 has a really poor
filter characteristic so without the
ceramic caps around it, it's garbage in,
garbage out. Unfortunately the
microcontroller's reset circuit can't
respond to noise and doesn't react.  I
suspect that the watchdog is iether
disabled in the firmware, or wasn't given a
restart vector so it's ineffective as well.
My unit chalked up yet another night of
sane operation:)
Cheers.
Title: Re: Timers and Macros Quitting After "X" Days
Post by: roger1818 on September 12, 2005, 11:08:37 AM
Andre:  What you are saying about the
capacitor makes sense.  This could explain
why some CM15As seem to be more reliable
than others.  The ones that work better
could be using capacitors that happen to
have a series inductance that is low enough
to provide adequate filtering.  The
“diagnostic” units probably have capacitors
from an especially good batch.

As far as the watchdog is concerned though,
I believe it is enabled, but the problem is
that since it uses a software time of day
clock, a reset causes the clock to stop
until it has been set again.  As a result,
anything with time dependence (timers and
macros with time delays) won’t work.