Print

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

[joe]

Got sick of the infamous battery ritual to
reset the CM15A. So I installed a normally
closed pushbutton on the case and wired it in
series with U5. When pushed, it interrupts
power to the 5 volt regulator and resets the
controller. You dont even have to unplug it!

Joe

[SteveRF]

Now that was a GREAT idea !
Thanks for sharing !

regards,
Steverf

[Henri T]

Hi,

Any picture of that ?

Thanks

[joe]

You dont really need a picture to accomplish
it. Just get a small normally closed
pushbutton and drill a hole in the side of
the case to mount it. Then cut the lead from
the regulator input and wire the switch in
series with it. Nothing fancy, the switch can
go anywhere. I used a small one I had laying
around that required a 1/4 inch hole. I
drilled it in the side of the case about even
with the antenna but on the opposite side. A
quick push and the CM15A is hardware reset!
If you still want a picture, I will take it
apart this weekend and take a shot.

Joe

[Scott T]

This is a band aid to the problem and x-10
should just fix this issue, but I like your
end user innovations!

[joe]

Scott,

Yep it is a baindaid no argument there but
look , this battery thing has been around as
long as the CM15A and no fix from X10 so
sometimes you gotta take the bull by the
horns. Maybe the design engineer at X10 will
read this and have an epiphany, learn to put
reset circuitry in it next time!

Joe

[carmine pacifico]

Maybe if the design engineer at X10 had any
brain he would use the Watchdog reset(WDR)
instruction within the chip and self reset
the CM15a when his firmware hangs,
eliminating the need to power cycle the
micro-controller.

[joe]

Carmine, if they did that, it would just hang
when the reset occured, all the RAM would get
initialized. There is no way to reload unless
it is hooked to a PC.

Joe

[carmine pacifico]

Joe

Reload what?
Macros are stored in external EE_prom, only
the clock need to be reloaded,
During a WDR reset the internal ram is
intact, this means the clock is just
stopped.
Their reset routine should interrogate
the “Processor status and Control Register”,
If a WDR reset proceed without resetting
Clock, at the most the CM15a would loose 1
second or 2.
From Micro-Controller data sheet:
10.0 Reset
The USB Controller supports three types of
resets. The effects
of the reset are listed below. The reset
types are:
1. Low-voltage Reset (LVR)
2. Brown Out Reset (BOR)
3. Watchdog Reset (WDR)
The occurrence of a reset is recorded in
the Processor Status
and Control Register (Figure 20-1). Bits 4
(Low-voltage or
Brown-out Reset bit) and 6 (Watchdog Reset
bit) are used to
record the occurrence of LVR/BOR and WDR
respectively.
The firmware can interrogate these bits to
determine the cause
of a reset.

[roger1818]

Carmine:  I suspect (though I don’t know
for sure) that they are using the Watchdog
reset.  The problem is the Watchdog will
only cause a reset if CPU hangs
completely.  This often isn’t the case.
Since same CPU is used for composing &
interpreting powerline commands, composing
& interpreting RF commands, maintaining the
clock (for timed events), and talking to
the USB port.  I have never noticed all of
these not working at the same time.  I have
had one or more of them stop working while
others continue to work.  If any of these
still work, the CPU must still be running
and the watchdog would not trigger a reset.

I don’t know why some would stop working
while others continue to work, but that
seems to be what happens.  Regardless of
the cause, a reset seems to get everything
to work again.  With the diagnostic unit,
the only thing I have had stop working is
USB communications.

[roger1818]

Carmine:  BTW. I do agree that they should
check what type of reset occurred and not
reset the clock if a Watchdog Reset
occurred.

[joe]

Carmine, you are obviously much more familiar
with the capabilities of the processor they
are using. I was not aware that it had
resetting flexibility. By all means if that
is the way it works then X10 should have
implemented those features.

Hey, it sounds like at some point, someone
here will get a pic emulator and just
reprogram the CM15 with their own firmware. I
can just see it now on ebay, CM15A upgrade
firmware $49.95

Have a happy.

Joe

[carmine pacifico]

Joe

I hate to burst your bubble, there are not
using a PIC, it’s a Cypress Micro-
Controller and has less computing power
than the PIC used in older X10 products,
and to make matters worse the Micro is
flawed and new part(Encore II) it’s not
backwards compatible both at firmware and
pin-out.
I am to busy working on my smart sprinkler
controller to waste any time on it, why
waste any time fixing their hardware when
their PC software(AHP) it’s else sub-
standard.

[joe]

Hmmm, Cypress. I wonder if there is some sort
of emulator available for that one, maybe
from digikey? Anyway, I agree with you, it is
not worth the time to re-engineer it. If I
was going to do it, I would not use the same
device for USB and RF but what do I know.

Joe

[roger1818]

Joe: It was the RF integration that
attracted me to the CM15A.  I agree that
the CM15A needs to be in three locations at
once for optimal performance (near the PC
for USB connectivity, near the centre of
the house for optimal RF range and near the
breaker panel for optimal powerline signal
strength) but having them integrated allows
macros to be triggered directly from an RF
command instead of having to wait for it to
be transceived into a powerline command.

AHP originally didn’t support this feature,
but I accidentally discovered the other day
that it now does (I don’t think it is
documented anywere).  It could use some
improvement since the current rule seems to
be that if there is a macro for that
command it won’t transceiver it.  There are
times when you have a module at the same
address as a macro and would like it to be
transceived anyway, but it isn’t that
difficult to get around the problem.