Print

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

[Tigger]

 
Carmine -

You keep pointing out that the Cypress
controller has a known bug with the GPIO
function, but you also state that X10 doesn't
use the GPIO function of the chip.

Can you explain then the relevancy of the
chip being flawed, so I can understand why
you keep mentioning it?


T

[SteveRF]

I really do not care what kind of watchdog
timers are used, cpu, chipset, etc etc etc...
What I DO CARE about is ..If there is an EASY
and RELIABLE way to cause "safe" reset
without having to place the unit back on the
PC and reload program or time...then I am
definitely interested... I am not going to
reverse engineer something that cost $49.95
and works pretty darn well for weeks and
weeks... That does not mean I won't be
keeping my eye open for better solutions but
a quick and easy solution to keep me from
having to remove, unplug batteries, reload
programs and reinstall back to the utility
closet would be MOST welcome to a lot of
us...probably MOST of us !  Tips and Ideas
Welcome !

Regards,
SteveRF

[SteveRF]

JOE...
By the way, your straightforward way of
presenting a hardware solution was great..!
So many times all I see is run on and run on
discussing the same thing we can't do
anything about..over and over and over...
Thanks !
SteveRF

[carmine pacifico]

Tigger

In a nut shell, they can not use interrupt
on I/O pin, the other option is polling;
polling increase firmware over head and
slows down response.
Because Cypress new version of the Chip is
based on totally new silicon I suspect that
they may be more problem than just GPIO.

Steve

If you follow Joe advice you will void the
warranty.
All this communication it’s not wasted, if
more people are aware of the hardware
problems and complain X10 may improve the
hardware.
As far as reverse engineering I would not
waste my time, but some people have to much
time on their hands; who knows they could
come up with an improved version, has been
done with other X10 products in the past.

[donald mcmow]

Carmine:  If X10 is NOT using the GPIO then
I guess the chip is NOT FLAWED. You can't
keep saying it is when that function is not
being used. If you are talking about a
device that is using the GPIO and doesn't
work then there is a problem. X10 isn't so
that portion of the chip is just dead
silicon. I think it's time to drop the
issue and deal with something else.

[joe]

SteveRF,
The switch I described was just to keep from
having to unplug and remove batteries. When
the button is pushed, the unit is dumb for
all intents and must be reloaded, but it does
save some time and aggrivation.

As far as voiding the warrantee, most of
these folks have modified their units with
the antenna fix. The reason I decided to do
this was that this problem has been around
since day 1 with no fix in sight. Since X10
cant or wont address the problem, I threw my
solution out for those who are as aggrivated
as me that you have to go through this ritual
every time you want to reset the darn thing.

I noticed that in the schematic that there is
a 10 microfarad capacitor on the input to the
regulator so if you simply unplug the unit
from the wall and then remove the batteries,
the 10 mic cap has to discharge before the
processor actually resets. This is why people
are saying they have to unplug and let sit
for a while. My reset button is after the
capacitor so when you push it, the regulator
instantly loses input voltage and the 5 volt
supply collapses. So it is very quick, just a
"bandaid" but those of us who are hackers
will see the benefit.

Joe

[roger1818]

Joe:  Your idea is a good one.  It
certainly would be easier than the battery
ritual.

Not to burst your bubble though, the 10 uF
cap does not play a significant role in
holding the voltage.  There is a 470uF cap
just off of the rectifier that will have a
larger effect.  Even still, the capacitors
should be 98% discharged in less than a
second.

[carmine pacifico]

Mr. Donald McMow

You are right, X10 it’s not using the
flawed part of the chip; but it’s like
buying a car:
It’s a great car but because of a flaw in
the transmission you can only go up 3rd
gear instead of 5th.
You can still go from point A to point B,
but a lot slower.
Loss of Clock
Loss of RF signals
Crashes when using more macros memory
These problems could be resolved if they
could use GPIO.

[Tigger]

 
I must be lucky - I haven't had to go through
the battery ritual yet.

I have had one or two lock-ups over the last
month or two, but nothing I haven't dealt
with using the CM11 . . .

The lock-ups I experience are resolved by
unplugging the CM15 for about 1/2 second,
and plugging it back in. . .


T

[scootertrs]

Could you please show where exactly the NC pushbutton switch would normally be wired? maybe a pic?  thanks

[Brian H]

Someone has added bypass caps to the CM15A to prevent higher frequency noise that big caps don't handle from messing it up. If you have lots of powerline noise.

Most electronics have bypass caps as standard design.

I have not tried it but the report they gave was improved markedly.

They added a .1uf 50V cap as follows:
Across VCC to VSS [+5 and Common] U1 [the microcontroller] pins 11 to 6
Across VCC to VSS U2 [24C64 Serial memory chip] pins 4 to 8
C11 100uf Output side of U5 regulator
C12 10uf Input side of U5 regulator.
I asked about C2 the 470uf 50V on the Power Supply and they indicated it didn't make much difference there. If I did the mod I would add one there also.
Radio Shack carries the .1uf 50V caps last time I looked.

[scootertrs]

OK got my cm15a open and cannot find U5 component location, but do see a 5V jumper on the board.  would cutting this jumper and wiring the NC switch accross these leads accomplish the reset function?  If not, can anyone direct me to the proper place to wire in the NC switch? Thanks

[Brian H]

U5 is a small TO92 type package and looks like a Transistor.
With the component side of PCB Facing you.
USB Port on bottom edge.
Left of the power transformer and the large 470uf cap.
Leads of U5 are Silk Screened on the Board:  I Input [nearest 470uf cap]; C Common and O Output.
I don't know if any of the leads have a jumper on the PCB

[scootertrs]

Got it... Thanks Joe

You know, in my system, when the cm15A does its ugly thing, it not only quits working, but also keeps all of the RF remotes from working, even if I have other RF receivers in the system.  I guess it generates noise or some type of jam signal.  Only when I do the reset or unplug from outlet do I get back RF operation.  Hopefully this reset setup will make the resets less painful.  I was wondering, If any others had experience with the Powerlinc 1132CU.

[Brian H]

I have a 1132CU in service now. I only use the Free Smarthome Manager Essential and not the Pro that adds Conditionals.
I just have a set of times downloaded to it, so nothing fancy.
Mine has one problem that maybe related to my exact hardware. If I leave the USB Cable connected and the computer is off. The internal clock in the 1132CU can loose maybe 30 minutes to an hour over night. Unconnected it is just fine. I do have my USB ports set to have NO Standby Power on them, most have standby on them for things like Wake On USB Activity.