X-10 control chip
Andrew Beasley
port.
Is there a dedicated chip that I can use (via a suitable mains interface) that
will encode the data directly on to the main signal?
I am trying to make a small interface unit that will sit in a normal backbox
rather than an 'external' unit.
TIA
Andrew
Nothing bothers me - I have two kids :-)
Uwe Wolfgang Radu
of low end micros. Just do a search for info on X-10 firmware for
microcontrollers.
Uwe W. Radu
Dave Houston
Nearly all of which use a TW523 which is what he's hoping to avoid. ACT has
an OEM module but they don't want to talk unless you're buying in mass
quantities.
---
BX24-AHT All Housecode Transceiver
http://Commander-X.com/bx24-aht.htm
Dave Houston
>Every system I see uses the CM11 or TWxxx interface and talks via a serial
>port.
>
>Is there a dedicated chip that I can use (via a suitable mains interface) that
>will encode the data directly on to the main signal?
>
>I am trying to make a small interface unit that will sit in a normal backbox
>rather than an 'external' unit.
A few chipmakers used to make powerline modem chips. Most were FSK but some
were ASK (but not 120KHz). AFAIK, all have been discontinued. But you might
get some ideas by looking at the datasheets and application notes. Search
on...
LM1893
LM2893
ST7535
ST7536
ST7537
TDA5051
It is possible to accomplish the same thing using a PIC or other
microcontroller. After all, that is what X-10 does.
Uwe Wolfgang Radu
> microcontroller. After all, that is what X-10 does.
You're a funny guy. Correcting me and then suggesting the same thing <g>.
--
Uwe W. Radu
Dave Houston
And you're a stupid guy. I did not suggest the same thing. You were talking
about using the PIC to control a TW523 (which is what the X-10 functions,
that you cited, in various microcontroller languages do) while I am talking
about doing it at a much lower and more direct level with no powerline
interface other than the microcontroller. There is a huge difference.
Uwe Wolfgang Radu
Thanks, your insight is flattering.
> I did not suggest the same thing. You were talking
> about using the PIC to control a TW523
Which part of my following original reply suggests that?
> X-10 is a simple and slow protocol that can be implemented with any number
> of low end micros. Just do a search for info on X-10 firmware for
> microcontrollers.
Welcome to Usenet. With a bit for practice you too can be a successful
poster.
--
Uwe W. Radu
Dave Houston
you.
Look, I am not only willing to be proven wrong, I am eager for
enlightenment.
I and many others who have participated in this newsgroup for the past few
years without benefit of your guidance would love it if you would cite
specific URLs that point to "X-10 firmware for microcontrollers". We're all
a bit jaded and tired of the many bugs and glitches in the PL513, TW523, and
CM11A. And I am actively engaged in designing a product that would benefit
immensely from the magical "X-10 firmware for microcontrollers" that you
imply is common. Maybe you were thinking of "magic beans".
I am aware that PBasic for BasicStamps includes a function for sending X-10
commands. I am aware that PIC Basic & PIC Basic Pro (from MicroEngineering
Labs) have functions for sending (& receiving in the Pro version) X-10
commands. I am aware that BasicX for the BX-01 & BX-24 has a function for
sending X-10 commands. I am aware that PICMicro Basic has a function for
sending X-10 commands. I am also aware that _ALL_ require a PL513 or TW523.
I am not familiar with the various C compilers for PICs. Some may include
X-10 functions. I've never heard of any that do not need a PL513 or TW523.
I do have Myke Predko's latest book and it lists no PIC Assembler functions
for X-10 in the Index.
I even conducted a Google search on "X-10 firmware microcontroller" and come
up with about 9999 entries that refer to something other than the X-10
protocol for every 1 that does refer to the X-10 protocol. And, of those I
found, all that actually refer to the X-10 protocol are about the
PL513/TW523 dependent cases cited above.
Perhaps you overlooked the fact that Mr. Beasley wants to get away from
dependence on a PL513/TW523.
I know of no PIC or other microcontroller that can directly interface with
120V or 220V AC mains voltages. All seem to prefer 5V (or less) DC.
Regulars in this newsgroup know that I have little patience with fools. I
have zero patience with damned fools. You are a damned fool! Either supply a
URL to "X-10 firmware for microcontrollers" that does not require a
PL513/TW523 or shut the hell up and crawl back in to whatever hole you just
crawled out of!
No amount of practice will make you less stupid! Ignorance is curable;
stupidity is forever.
"Uwe Wolfgang Radu" <a...@at.com> wrote:
Dennis Heidner
powerline. But having looked at the note, I think it may almost just
marketing junk instead of a real application (maybe MP stands for Marketing
Proposal!). No firmware listing and not very detailed. I was pretty
sure that one of the Microchip seminars a few years back included an app
note with firmware showing how to build an X-10 receiving device - but I
sure can't find it.
I had seen a URL for somebody that had implemented an interface (without
TW523) using a 68C705, but when I went back through my list of
favorites.... I couldn't find it. But I did find lots of URL's no longer
valid.... darn :-( I'll look through my stacks of papers and see if I
printed a copy out.
I did find some PLC interface notes but all were referencing using the
ST75xx or 18xx chips that Dave has already identified as obsolete.
Dave Houston
better isolation from the mains for any commercial application. But, if you
combine it with the Philips TDA5051 documentation which does give specifics,
including part numbers and component values, you could soon "see the light".
You would have to create your own firmware for the PIC but I think that
would be the easy part.
If you could accomplish it all through a UL approved Class 2 transformer,
say 12 or 24 VAC, you would need no UL approval.
Actually, later versions of the Philips powerline modem chip are still made.
I think the TDA5051 and TDA5051A have been superceded by the TDA5051AT (?).
The Multi Purpose Power Line Micro-Module based on it that I cited will not
do X-10 "out-of-the-box" but they will replace the crystal and other filter
components to make it X-10 capable for a "nominal" charge. But, at $80 plus
a "nominal" charge, it doesn't appear very practical. Here's the URL
again...
http://www.michat.com/oem_micro.htm
"Dennis Heidner" <den...@heidners-no-spam.net> wrote:
Dennis Heidner
more details. I did see an interesting note in another appnote on the
12Cxxx stating that the designer should take extra precautions to protect
the chip against surges. The suggested a MOV... Hmmm I doubt they
actually ever tried the design.
I saw the specs for the Philips part when I searched their site I and been
hoping that perhaps they had a solution with a 80C51. No such luck.
I saw references to a PLC part on the International Rectifier site... but
didn't have time to chase that link.
Kevin Olalde
Kevin Olalde wrote:
>
> I'd love to see some specific URLs.
>
> There are plenty of things/examples on connecting to the existing power
> line interfaces, but I have yet to see anyone (aside from the recent
> Powerlinc) create 'X-10 firmware' that relaces something like a TW523.
>
> So, can you be more specific?
>
> Kevin
Dave Houston
http://www.michat.com/oem_micro.htm
However, you did not specify affordable. ;)
But, if you study the documentation and schematics you can probably see how
one might do this for X-10 with a microcontroller but w/o the Philips chip.
Kevin Olalde <kol...@home.com> wrote:
>I'd love to see some specific URLs.
>
>There are plenty of things/examples on connecting to the existing power
>line interfaces, but I have yet to see anyone (aside from the recent
>Powerlinc) create 'X-10 firmware' that relaces something like a TW523.
>
>So, can you be more specific?
>
>Kevin
>
>Uwe Wolfgang Radu wrote:
>>
Kevin Olalde
There are plenty of things/examples on connecting to the existing power
line interfaces, but I have yet to see anyone (aside from the recent
Powerlinc) create 'X-10 firmware' that relaces something like a TW523.
So, can you be more specific?
Kevin
Kevin Olalde
>
> Ask and you shall receive. ;)
>
> http://www.michat.com/oem_micro.htm
>
> However, you did not specify affordable. ;)
Yikes, I'm was just getting over how expensive I thought switching
regulators are.
> But, if you study the documentation and schematics you can probably see how
> one might do this for X-10 with a microcontroller but w/o the Philips chip.
Yeah maybe, but a tad much for this newbie. I'll have to wait for the
movie (or at least someone's circuit and sample code).
Kevin
Jay R. Ashworth
Dave Houston <dhou...@fuse.net> wrote:
> Look, I am not only willing to be proven wrong, I am eager for
> enlightenment.
Ok, I'll try to provide some. You get out of the wrong side of the bed
this morning, Dave? I read Uwe's posting, and it didn't seem to merit
this sort of response to *me*...
> I and many others who have participated in this newsgroup for the past few
> years without benefit of your guidance would love it if you would cite
> specific URLs that point to "X-10 firmware for microcontrollers". We're all
> a bit jaded and tired of the many bugs and glitches in the PL513, TW523, and
> CM11A. And I am actively engaged in designing a product that would benefit
> immensely from the magical "X-10 firmware for microcontrollers" that you
> imply is common. Maybe you were thinking of "magic beans".
Hmmm... this really *doesn't* seem like rocket science to *me*; if it
is so scarcely implemented, I'm forced to wonder why? Accept commands
on a serial port, queue them, and have a lower level interrupt routine
that pulls the off the queue and modulates them against a zero-crossing
interrupt...
I haven't written microcontroller code, but I can't see that this is
more than about a couple hundred lines of code...
Cheers,
-- jra
--
Jay R. Ashworth j...@baylink.com
Member of the Technical Staff Baylink
The Suncoast Freenet The Things I Think
Tampa Bay, Florida http://baylink.pitas.com +1 727 804 5015
Dave Houston
>On Thu, 15 Mar 2001 04:13:37 GMT,
> Dave Houston <dhou...@fuse.net> wrote:
>> Look, I am not only willing to be proven wrong, I am eager for
>> enlightenment.
>
>Ok, I'll try to provide some. You get out of the wrong side of the bed
>this morning, Dave? I read Uwe's posting, and it didn't seem to merit
>this sort of response to *me*...
Then maybe you need reading lessons.
>> I and many others who have participated in this newsgroup for the past few
>> years without benefit of your guidance would love it if you would cite
>> specific URLs that point to "X-10 firmware for microcontrollers". We're all
>> a bit jaded and tired of the many bugs and glitches in the PL513, TW523, and
>> CM11A. And I am actively engaged in designing a product that would benefit
>> immensely from the magical "X-10 firmware for microcontrollers" that you
>> imply is common. Maybe you were thinking of "magic beans".
>
>Hmmm... this really *doesn't* seem like rocket science to *me*; if it
>is so scarcely implemented, I'm forced to wonder why? Accept commands
>on a serial port, queue them, and have a lower level interrupt routine
>that pulls the off the queue and modulates them against a zero-crossing
>interrupt...
You seem to be missing the central point. Aside from the fact that you've
overlooked more than half of the task (receiving), the question was not
about whether it was possible to write such code, the question was whether
such code was already readily available for anything other than the
PL513/TW523 or CM11A. Having written such code myself for the PL513/TW523,
CM11A, PowerLinc, and "players to be named later", I know it can be written,
but aside from a couple of powerline modems intended for non-X10 uses, I
know of no hardware that is available. So, as I suggested, you first need to
design (or select) the powerline interface and then write your own firmware.
If you think that is the same as Uwe's claim that "X-10 firmware for
microcontrollers" for this task is readily available, then you are as
clueless as he is.
It isn't scarcely implemented - it isn't implemented at all - at least not
as "X-10 firmware for microcontrollers" for anything except the PL513/TW523.
I notice you haven't provided any URLs either.
As I pointed out, nearly everything that is available as "microcontroller
firmware" requires a PL513/TW523 and the originator of the thread is trying
to get away from needing a TW523 (he currently uses a BasicStamp & TW7223).
Even if such code were ubiquitous, it would have to be specific to some
hardware that provides the physical interface to the powerline (including
generation of the 120KHz carrier).
Until a couple of years ago when X-10's patent expired, X-10 only allowed
OEM's to interface using the PL513/TW523. Most of the existing PIC compilers
(including CCS) have limited X-10 functions that work only with the
PL513/TW523. It's the same for the BasicStamp and BasicX. I'm not familiar
with 8051 compilers but will hazard a guess that if any have X-10 functions
they are for the PL513/TW523. They are of no help to someone trying to avoid
using the PL513/TW523 which was the central issue missed both by Uwe and now
you.
>I haven't written microcontroller code, but I can't see that this is
>more than about a couple hundred lines of code...
So where's your code or pointers to code? How do you suggest we interface to
the powerline? I _have_ written microcontroller code (and am currently
writing some in this very area) and am still looking for the "enlightenment"
you promised. All I see is a bunch of clueless bafflegab.
If you have some facts, present them. I don't like wasting time with the
terminally clueless, especially before breakfast. And I have no respect for
someone who continues to argue on a topic about which he has demonstrated
cluelessness.
Jay R. Ashworth
Dave Houston <dhou...@fuse.net> wrote:
> You seem to be missing the central point. Aside from the fact that you've
> overlooked more than half of the task (receiving), the question was not
> about whether it was possible to write such code, the question was whether
> such code was already readily available for anything other than the
> PL513/TW523 or CM11A. Having written such code myself for the PL513/TW523,
> CM11A, PowerLinc, and "players to be named later", I know it can be written,
> but aside from a couple of powerline modems intended for non-X10 uses, I
> know of no hardware that is available. So, as I suggested, you first need to
> design (or select) the powerline interface and then write your own firmware.
> If you think that is the same as Uwe's claim that "X-10 firmware for
> microcontrollers" for this task is readily available, then you are as
> clueless as he is.
Ah. I see.
You know, Dave, even being a star doesn't always entitle you to go on a
star trip. Whether he's clueless or not, being a prick isn't going to
sell you a lot of kits...
> It isn't scarcely implemented - it isn't implemented at all - at least not
> as "X-10 firmware for microcontrollers" for anything except the PL513/TW523.
> I notice you haven't provided any URLs either.
>
> As I pointed out, nearly everything that is available as "microcontroller
> firmware" requires a PL513/TW523 and the originator of the thread is trying
> to get away from needing a TW523 (he currently uses a BasicStamp & TW7223).
> Even if such code were ubiquitous, it would have to be specific to some
> hardware that provides the physical interface to the powerline (including
> generation of the 120KHz carrier).
Indeed it would.
> Until a couple of years ago when X-10's patent expired, X-10 only allowed
> OEM's to interface using the PL513/TW523. Most of the existing PIC compilers
> (including CCS) have limited X-10 functions that work only with the
> PL513/TW523. It's the same for the BasicStamp and BasicX. I'm not familiar
> with 8051 compilers but will hazard a guess that if any have X-10 functions
> they are for the PL513/TW523. They are of no help to someone trying to avoid
> using the PL513/TW523 which was the central issue missed both by Uwe and now
> you.
No, I didn't miss it, as you can see from my comment. What my comment
missed what which part of the problem with Uwe's comments was troubling
*you*.
> >I haven't written microcontroller code, but I can't see that this is
> >more than about a couple hundred lines of code...
>
> So where's your code or pointers to code? How do you suggest we interface to
> the powerline? I _have_ written microcontroller code (and am currently
> writing some in this very area) and am still looking for the "enlightenment"
> you promised. All I see is a bunch of clueless bafflegab.
Well, perhaps your presentation is scaring off people with something
useful to contribute, Dave; has this thought ever occurred to you?
> If you have some facts, present them. I don't like wasting time with the
> terminally clueless, especially before breakfast. And I have no respect for
> someone who continues to argue on a topic about which he has demonstrated
> cluelessness.
Yawn. That's nice.
"How do we interface to the powerline?"
Hmmm... I'm not an EE, and it's been quite some time since I dabbled in
low-level... but it seems to me that a transformer would be required to
be clean at 110KHz (read: expensive)... I think I'd use a high-voltage
transistor oscillator to do the 110KHz, and opto-couple to it from the
low voltage electronics, myself. Broad brushstrokes, I realize, but
you've got to start somewhere.
Dave Houston
>You know, Dave, even being a star doesn't always entitle you to go on a
>star trip. Whether he's clueless or not, being a prick isn't going to
>sell you a lot of kits...
You know, Jay, you really don't have any idea what you are talking about and
keeping your head up your ass will not cure your cluelessness.
By "kits", you must mean the BX24-AHT as that is the only "kit" I am
associated with. So, just what part of "FREE" are you having trouble
comprehending?
I spent a lot of my time and my money developing the BX24-AHT and receive
absolutely nothing from it. The binary file is included in the distribution
and anyone with a BX-24 can make use of it. I have suggested that people buy
the necessary components from Peter Anderson (who sells the components at
bargain prices) in order to help support his educational activities, but I
receive nothing except the support burden (and crap from stupid assholes
like you). Did I mention that I funded my development efforts out of my
disability pension?
As I said earlier, I have zero respect for anyone who continues to argue
without any facts. Now, tuck your head back up your ass and go away.
Jay R. Ashworth
Dave Houston <dhou...@fuse.net> wrote:
> j...@dorothy.msas.net (Jay R. Ashworth) wrote:
> >You know, Dave, even being a star doesn't always entitle you to go on a
> >star trip. Whether he's clueless or not, being a prick isn't going to
> >sell you a lot of kits...
>
> You know, Jay, you really don't have any idea what you are talking about and
> keeping your head up your ass will not cure your cluelessness.
>
> By "kits", you must mean the BX24-AHT as that is the only "kit" I am
> associated with. So, just what part of "FREE" are you having trouble
> comprehending?
That I missed it entirely. That you're giving it away free, frankly,
*still* doesn't not entitle you to snipe at people in public. If
you're feeling put upon, go away.
> I spent a lot of my time and my money developing the BX24-AHT and receive
> absolutely nothing from it. The binary file is included in the distribution
> and anyone with a BX-24 can make use of it. I have suggested that people buy
> the necessary components from Peter Anderson (who sells the components at
> bargain prices) in order to help support his educational activities, but I
> receive nothing except the support burden (and crap from stupid assholes
> like you). Did I mention that I funded my development efforts out of my
> disability pension?
Nope.
> As I said earlier, I have zero respect for anyone who continues to argue
> without any facts. Now, tuck your head back up your ass and go away.
Well, Dave; on Usenet, the only currency one has to spend is
reputation. I'm sure those people reading this group will make their
own judgements on yours, and on mine.
I wasn't giving you any crap; you seem to have decided that you'd just
invent some, so you could stand in it's way, or so it seems to me. But
if you're really this oversensitive, might I suggest that Usenet is
about the *last* place you ought to spend your time? It's a jungle out
here.
Play nice, or don't bother to play at all, ok?
Uwe Wolfgang Radu
> missed what which part of the problem with Uwe's comments was troubling
> *you*.
Well, my only substantive post in this thread (other than disbelief at
Dave's behavior) reads:
> X-10 is a simple and slow protocol that can be implemented with any number
> of low end micros. Just do a search for info on X-10 firmware for
> microcontrollers.
He claims I said that "X-10 firmware for microcontrollers" (which for some
reason he keeps quoting like some sort of Dubbya-ism) is readily available,
while my statement does nothing of the sort. Basic grasp of the English
language would let one deduce that I merely *suggest* such firmware *ought
to* be available, given the relative simplicity of the problem (after all,
the CM11A sells for a trifle, so it must cost the manufacturer pennies). I
certainly make no claims that it *does* exist.
I guess Dave was in some sort of fighting mood, putting words in my
(metaphorical) mouth. He may or may not be knowledgeable in the field; it's
hard to tell, given that the only testament to his skills is coming from
himself. Also, the fact that he only seems familiar with micro BASIC
compilers would automatically put his skills in a certain category, as far
as I'm concerned. He seems to have to rely on built-in X-10 commands in his
tools, rather than bit-banging the data stream himself.
At my previous job we developed an entire powerline networking system for
in-house use. It was not X-10 compatible (very much faster, in fact), but
relied on an 8051 with suitable powerline interfacing electronics. No
off-the-shelf powerline modem, nothing. I am a software developer, so I
don't know the specifics, but I could ask the EE who developed it for the
schematics, since he's a good friend of mine and they ended up scrapping it
anyway. But attitudes such as Dave's aren't exactly conducive to
cooperation. In fact, if this were a moderated forum he'd have a tough time
being read at all. As it stands, I have to rely on the killfile. Oh well.
--
Uwe W. Radu
Dave Houston
firmware.
Having seen your posts to other newsgroups, I am not about to buy your "Me
no speaka da langwage" bullshit. Hmmm, you mostly post to VB newsgroups and
then try to belittle me because I only mentioned Basic compilers for the
PIC. The major C compiler for the PIC also includes X-10 functions for the
TW523 only.
You said I was suggesting the same thing that I had criticized you for. Not
only did I point out that you cannot speak of "X-10 firmware for
microcontrollers" without specifying the physical interface, I pointed out
URLs for some physical interfaces that do not rely on X-10 hardware.
You, on the other hand, speak of "suitable powerline interfacing
electronics" without specifying what they are and claim to have developed
such a system. Hmmm, but you say it wasn't X-10 compatible. Hmmm, maybe you
missed the fact that the originator of the thread was looking for an X-10
COMPATIBLE interface other than the TW523 or CM11A. Hmmm, you say it is was
much better than X-10. Hmmm, I wonder why it never made it to market. Hmmm,
maybe you've been spending too much time in the sauna and your brain is
water-logged.
How about posting some confirmable details instead of more bullshit!
"Uwe Wolfgang Radu" <a...@at.com> wrote:
Dennis Heidner
In the past I had looked at trying to make my own X-10 powerline interface.
I had previously searched for non TW523 line interfaces and found a few
references to receivers but NO X-10 transmitting (third party chips). I've
spent a chunk over the last several days checking out some of my old links
-- and looking at new links (600+). There still are no X-10 "do it
yourself" replacements for the TW-523.
Part of the reason for the lack of options was the requirement of using the
TW-523 to avoid patent infringement. But also simply the cost. A TW523
isn't that expensive. Even if you bought the parts to make 100 units, the
cost of the final product would still approach or exceed the cost of the
TW-523.
An PLC implementation that was done three or four years ago may no longer
be possible. As Dave has previously mentioned, many of the common
powerline interface chips have been discontinued. There are a few left,
but the ones used in designs of a few years back are gone (or going).
Marrick's web page has a X-10 TW-523 replacement interface displayed. But
it is apparently not yet shipping (even after a couple of years). Quite
unfortunate since Marrick claims to offer a device that handles collisions
AND extended codes.
Uwe Wolfgang Radu
the
> TW-523 to avoid patent infringement.
That's true, and I figured as much after reading some stuff on X-10's web
site I think. Since the patent only expired about two years ago, and many
people might not know that yet, I guess it's not surprising that nothing is
available yet.
> Even if you bought the parts to make 100 units, the
> cost of the final product would still approach or exceed the cost of the
> TW-523.
True, but for most hobbyists cost probably wouldn't be the primary reason of
rolling their own anyway, right? Many hackers would simply want something
that looks nice and integrated rather than just a cheap white box with a
dodgy serial protocol. I was writing a Delphi component for the CM11A a
couple of years ago but gave up in disgust after fiddling with it for a
while (losing interest in X-10 might also have had something to do with it
;-)
> An PLC implementation that was done three or four years ago may no longer
> be possible. As Dave has previously mentioned, many of the common
> powerline interface chips have been discontinued.
The point is though that you can create a powerline interface with discrete
components, without requiring an IC. An IC after all is nothing but a
collection of discrete components on a die, and 120MHz is slow enough to
handle off chip. The thing is, X-10 has so many other limitations, I'm not
sure it's worth expending the effort of actually doing something. There are
alternatives, after all, even if more costly.
--
Uwe W. Radu
Dave Houston
>> Part of the reason for the lack of options was the requirement of using
>the
>> TW-523 to avoid patent infringement.
>
>That's true, and I figured as much after reading some stuff on X-10's web
>site I think. Since the patent only expired about two years ago, and many
>people might not know that yet, I guess it's not surprising that nothing is
>available yet.
>
>> Even if you bought the parts to make 100 units, the
>> cost of the final product would still approach or exceed the cost of the
>> TW-523.
>
>True, but for most hobbyists cost probably wouldn't be the primary reason of
>rolling their own anyway, right? Many hackers would simply want something
>that looks nice and integrated rather than just a cheap white box with a
>dodgy serial protocol. I was writing a Delphi component for the CM11A a
>couple of years ago but gave up in disgust after fiddling with it for a
>while (losing interest in X-10 might also have had something to do with it
>;-)
The originator of the thread is not a hobbyist looking for something for his
own needs only. He wants to port the application he has developed on a
BasicStamp/TW7223 to a PIC and a small X-10 capable powerline modem that he
can integrate into the same module as the PIC and make it a commercial
product.
Why don't you go back to trolling the VB groups which is where you usually
hang out, complaining that VB isn't Delphi?
>> An PLC implementation that was done three or four years ago may no longer
>> be possible. As Dave has previously mentioned, many of the common
>> powerline interface chips have been discontinued.
>
>The point is though that you can create a powerline interface with discrete
>components, without requiring an IC. An IC after all is nothing but a
>collection of discrete components on a die, and 120MHz is slow enough to
>handle off chip. The thing is, X-10 has so many other limitations, I'm not
>sure it's worth expending the effort of actually doing something. There are
>alternatives, after all, even if more costly.
Which is exactly what I proposed. And is exactly what you did NOT propose
until now, long after you accused me of criticizing your proposal only so I
could then say the same thing. My first post to the thread was that the
firmware you claimed could be found in a search is all for the TW523. You
ignored that post and then made your false accusations only a few days later
when, after doing a bit of research, I did propose specific, concrete
hardware that could be adapted.
We're still waiting for you to point to such hardware or to non-TW523
firmware.
Uwe, you are an intellectually dishonest slimebag.
Dennis Heidner
Heathkit is no longer selling kits, because the demand dropped. The choice
of parts at RatShack is pretty horrible now days. The amount of effort to
make a safe device is well beyond the average experimenter. I'd thought
about it, but decided I have more important projects that I want to get
done -- and two teenagers do consume a little bit of time.
If you wanted to design one and perhaps publish the design through Circuit
Cellar, that we'd all enjoy the article and appreciate your efforts.
Jay R. Ashworth
> in-house use. It was not X-10 compatible (very much faster, in fact), but
> relied on an 8051 with suitable powerline interfacing electronics. No
> off-the-shelf powerline modem, nothing. I am a software developer, so I
> don't know the specifics, but I could ask the EE who developed it for the
> schematics, since he's a good friend of mine and they ended up scrapping it
> anyway.
[carefully trimmed]
Yes, We'd like it a *lot* if you did that.
It's time for something fresh.
Uwe Wolfgang Radu
> Cellar, that we'd all enjoy the article and appreciate your efforts.
Well, I don't have all the expertise myself, though with the help of a EE
friend of mine it would he doable. But as I said, I've kind of lost interest
in X-10 because it's just too damn slow and inflexible. I guess I'll wait
for IP-enabled light switches <g>.
--
Uwe W. Radu
Jay R. Ashworth
> components, without requiring an IC. An IC after all is nothing but a
> collection of discrete components on a die, and 120MHz is slow enough to
> handle off chip. The thing is, X-10 has so many other limitations, I'm not
> sure it's worth expending the effort of actually doing something. There are
> alternatives, after all, even if more costly.
Well, you can, but in the real implementer world out here, you *really*
want the controller to speak RS-232, if it's talking to a computer.
Realtime is a pain to deal with on computers.
I concur, to some extent, with your last observation, though, hence my
other inquiry. :-)
Uwe Wolfgang Radu
> want the controller to speak RS-232, if it's talking to a computer.
> Realtime is a pain to deal with on computers.
Of course. I thought we were talking about emulating the functionality of a
CM11A, only done right. So our box would contain a (custom) powerline
interface, a micro, and a decent ASCII protocol so we can talk (close to)
plain English to the thing.
> I concur, to some extent, with your last observation, though, hence my
> other inquiry. :-)
Frankly, I've seen minimalist UDP over PPP or even Ethernet designs that
could be built for as little as home-built X-10 devices, and it's only going
to get cheaper. Circuit Cellar had several articles on minimalist IP-based
devices: PIC web server (Oct 2000), Atmel web server (Jul 1999 online), and
a few others I forget. Dr. Dobb's Journal presented the Dallas Semi TINI
(Oct 2000), which is a Java-based PC-on-a-SIMM-stick for $50, with 10base-T
and RS-232 connections. Makes a great serial-to-Ethernet gateway, and is
trivially programmable in Java.
Any of these solutions really are more exciting than X-10. Of course, nice
also would be a set of devices that did high-speed IP (at least 115K) over
powerline, kind of like an X-10 on steroids with full two-way
communications. Having millisecond latency when hitting a switch would be
nice.
--
Uwe W. Radu
Dennis Heidner
"Every system I see uses the CM11 or TWxxx interface and talks via a serial
port.
Is there a dedicated chip that I can use (via a suitable mains interface)
that
will encode the data directly on to the main signal?"
The answer to that question is not yet.
As for X-10 v.s. Ethernet...
None of the Circuit Cellar designs ever tried to put Ethernet speed signals
onto the power line and THEN control a lightbulb in a socket. A nice
dream, but still several years away. Even using the Netmedia web server
kit in quantities of hundreds, a "smart" lightbulb would socket would cost
$50+ or more. Compare that to a cheap socket rocket...
There isn't anything that says a low-cost Ethernet on power line would be
anymore reliable. Ethernet does not guarantee delivery of packets. I've
seen some networks with bad connections and the result is the transfer rate
is almost down to a standard voice grade line. We often assume that if we
just make stuff Ethernet or internet it will be reliable... but there is
far more to it than that.
I believe you'd eventually here the same kind of complaints about Ethernet
on mains as you do X-10 on mains... unreliable, slow, etc... when often
many of the problems can be cleaned up by installing noise filters and
couplers (something that certainly would be required for powerline
Ethernet).
Dave Houston
>The Andrew Beasley original thread on 3/8/01 started with:
>
>"Every system I see uses the CM11 or TWxxx interface and talks via a serial
>port.
>Is there a dedicated chip that I can use (via a suitable mains interface)
>that will encode the data directly on to the main signal?"
>
>The answer to that question is not yet.
Gee, someone other than myself actually read and understood the original
post. ;)
Actually, there is a dedicated module that does X-10 as well as a faster,
proprietary protocol. I have a couple for testing. But they are not small,
need a power supply (18V & 5V), and need opto-isolation on all the control
lines as the circuits are not isolated from the mains. My initial evaluation
is that, in their current form, they have zero chance of passing UL or CE
certification so are not applicable for Mr. Beasley's commercial project.
(It might be that the ones I was sent are engineering samples and that
finished products will be different.) They might be applicable for DIY use
(although DIYers can accomplish the same thing with simple modifications to
a TW523) but I don't know whether the company that makes them would consider
selling them to individuals nor do I know the price point should they be so
willing.
Once I've had the time to test them, I may post more here depending on the
test results and the plans of the company that makes them. They are looking
for a company (or companies) to work with in the US.
BTW, on modifying a TW523, my guess is that Neil Cherry's article (to be
published in Circuit Cellar) will include the necessary details.
>As for X-10 v.s. Ethernet...
>
>None of the Circuit Cellar designs ever tried to put Ethernet speed signals
>onto the power line and THEN control a lightbulb in a socket. A nice
>dream, but still several years away. Even using the Netmedia web server
>kit in quantities of hundreds, a "smart" lightbulb would socket would cost
>$50+ or more. Compare that to a cheap socket rocket...
Plus the cost and effort of getting UL and/or CE approval for the "smart"
lightbulb or appliance. As you have already noted, the LynX-10 PLC is still
waiting for such approvals despite being announced and shown at a trade show
more than two years ago.
>There isn't anything that says a low-cost Ethernet on power line would be
>anymore reliable. Ethernet does not guarantee delivery of packets. I've
>seen some networks with bad connections and the result is the transfer rate
>is almost down to a standard voice grade line. We often assume that if we
>just make stuff Ethernet or internet it will be reliable... but there is
>far more to it than that.
>
>I believe you'd eventually here the same kind of complaints about Ethernet
>on mains as you do X-10 on mains... unreliable, slow, etc... when often
>many of the problems can be cleaned up by installing noise filters and
>couplers (something that certainly would be required for powerline
>Ethernet).
Besides, I think Uwe will be too busy to work on this anyway. His charge
that I denigrated his proposals and then claimed them as my own is an
outrageous and reprehensible lie and one that I cannot let go unchallenged.
My attorney will be here to discuss this matter today.
---
BX24-AHT All Housecode Transceiver is at:
http://www.geocities.com/dorothy1801/
Uwe Wolfgang Radu
I thought Dave kindly answered that one already mostly to the negative.
> None of the Circuit Cellar designs ever tried to put Ethernet speed
signals
> onto the power line and THEN control a lightbulb in a socket.
Not as far as I know. But then, it wouldn't need to be 10Mpbs; 115K or even
56K would suffice for most home uses. Intelogis did 350K with their p2p
powerline adapters, and now they changed their name to Inari and are
starting to peddle the chipsets directly--first 2M, later this year
supposedly 12M. Even 2M would be overkill for switching, but the price
remains to be seen.
> There isn't anything that says a low-cost Ethernet on power line would be
> anymore reliable. Ethernet does not guarantee delivery of packets.
Of course it does not. The main benefit of a fast(er) medium and a
bidirectional protocol is that you can implement handshaking, which does
lead to reliability. Unless I'm mistaken there's an effort afoot even for
X-10 to implement module feedback, which could make it reliable also. Of
course, couple a few lost packets with X-10 speeds, and you'll have some
mighty slow response times.
The main thing IP over powerline would buy you is that you wouldn't need yet
another protocol stack. All you would need is a powerline MAC and your
powerline adapter would look just like another Ethernet card in your system.
I have no doubt that this is the direction things are going.
--
Uwe W. Radu
Jay R. Ashworth
Dave Houston <dhou...@fuse.net> wrote:
> Besides, I think Uwe will be too busy to work on this anyway. His charge
> that I denigrated his proposals and then claimed them as my own is an
> outrageous and reprehensible lie and one that I cannot let go unchallenged.
> My attorney will be here to discuss this matter today.
The *first* thing your attorney will tell you is that you shouldn't
have said that in a public forum.
Have a nice day.
<plonk>
Uwe Wolfgang Radu
> have said that in a public forum.
The second is that he doesn't have a case anyway since he's the only one
exhibiting any agression in the entire thread. The third should probably be
to save his disability pension for more important things, like BASIC
compilers and medication. Sorry I still get to see his stuff through your
quotes :-)
--
Uwe W. Radu
Paul Beam
I created a TW-523 and CM11A replacement using a PIC 16C63. There are
really only two technical problems, but the biggest problem is that X10's
devices are basically "good enough" and very cheap.
There is a trick to connecting to the power lines. Working through a power
transformer does not work. The only thing I have found that works is to use
the same small-signal transformers that X10 uses -- they are manufactured by
Sumida. Whereas X10 uses a lot of analog circuitry, you can generate the
X10 signal digitally and receive it either with a comparitor/counter, PLL,
or frequency detection IC.
Unfortunately, what I have done, I do not own. From my experience I can say
this, X10's stuff has problems, but to compete with them is extremely
difficult unless you have the very deep pockets necessary to produce product
in the volume necessary to compete with them in price.
Dave Houston
I think this is Paul's first post to the news group. He is the designer of
the ESM1 meter that many of us have and love.
I am curious why you can measure the signal through a power transformer but
it won't go the other way.
Dennis Heidner found the 125MHz Toko transformer used by Phil Plunkett in
his X-10 compatible lamp module is available from Digi-Key.
The Newport 78250 pulse transformer used in the Philips TDA5051A application
notes is available from Mouser.
Jay R. Ashworth
Me! Mee!!
> I created a TW-523 and CM11A replacement using a PIC 16C63. There are
> really only two technical problems, but the biggest problem is that X10's
> devices are basically "good enough" and very cheap.
Yep.
> There is a trick to connecting to the power lines. Working through a power
> transformer does not work. The only thing I have found that works is to use
> the same small-signal transformers that X10 uses -- they are manufactured by
> Sumida. Whereas X10 uses a lot of analog circuitry, you can generate the
> X10 signal digitally and receive it either with a comparitor/counter, PLL,
> or frequency detection IC.
I so assumed.
> Unfortunately, what I have done, I do not own. From my experience I can say
> this, X10's stuff has problems, but to compete with them is extremely
> difficult unless you have the very deep pockets necessary to produce product
> in the volume necessary to compete with them in price.
Well, you tell me this:
If someone designed a system with 16 bits of addressibility (8
"housecode", 8 module address), direct dim presetting, *and active
acknowledgement*, with a more usable datarate on the wire, do you think
better would be the enemy of good enough, and it would sell?
(Especially if the transmitters and such could also be X-10
compatible?)
Specifically, if you could get silicon cut for a one-chip
RS-232-to-transformer-connections IC.
Is "X-20" trademarked yet? :-)
What else would such a system need to have to make people happy?
(Subject line adjusted)
Cheers,
Paul Beam
could never get enough power through it. The powerline is rather low
impedance, so you have to be able to pump a lot of current. The general
rule of thumb is 5 ohms for the powerline, so 5V requires 1A -- crude
approximation.
When I tried to crank up the current to a measurable voltage, I ended up
burning out the transformer. If the receivers were more sensitive, it would
be possible.
"Dave Houston" <dhou...@fuse.net> wrote in message
news:3ac098f7...@nntp.fuse.net...
Dave Houston
>I tried sending a signal through the transformer from the secondary side. I
>could never get enough power through it. The powerline is rather low
>impedance, so you have to be able to pump a lot of current. The general
>rule of thumb is 5 ohms for the powerline, so 5V requires 1A -- crude
>approximation.
OK. Anyway, European regulations would still require testing and
certification so it wouldn't be that much of an advantage.
Still, I wonder how the PowerLinc does it. Maybe they interface on the
primary side of the transformer.
Paul Beam
"Jay R. Ashworth" <j...@dorothy.msas.net> wrote in message
news:slrn9c1br...@dorothy.msas.net...
> If someone designed a system with 16 bits of addressibility (8
> "housecode", 8 module address), direct dim presetting, *and active
> acknowledgement*, with a more usable datarate on the wire, do you think
> better would be the enemy of good enough, and it would sell?
> (Especially if the transmitters and such could also be X-10
> compatible?)
>
> Specifically, if you could get silicon cut for a one-chip
> RS-232-to-transformer-connections IC.
>
> Is "X-20" trademarked yet? :-)
My experience has been that people talk more than they buy. Others have
come up with technically better solutions than X10, but they have not been
able to sell much. If you have the resources for an ASIC, the safety
approvals, and a large enough order to get the per unit cost down, then I
can provide you with a better solution! :)
I hate to be a pessimist, but my opinion is that the technology is not
lacking, but the killer application is.
>
> What else would such a system need to have to make people happy?
If I knew that, I would be wealthy.
Paul
Skip Carlson
I just joined this group and am definately paying attention to your post.
Is there a patent to be infringed if you create a TW-523 replacement? I'm
just curious because if there isn't this would be great news.
Did you require an emulator to program the PIC? Also, have you done any
research into what it takes to build your own receiver module for X-10 vs.
buying a lamp module from X-10?
Thanks!
"Paul Beam" <pb...@nxs.net> wrote in message
news:3ac0...@news.planetc.com...
Paul Beam
I did not do exhaustive investigation into the patent area, but I do know
the original X10 patent has expired. I made a device that was compatible,
but not identical. It was a completely new design and only resembled a
TW523 in its connection to a 3-rd party device.
The stuff I have done is 2-way, so I know what is involved in receiving X10
and doing the lamp function. I'm not sure it makes economic sense when you
can sometimes get lamp modules retail for $5 or less.
"Skip Carlson" <scar...@malachitetech.com> wrote in message
news:4aqw6.403$9d.1...@newshog.newsread.com...