Print

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

[jantman]

Hello,

I'm going to be doing a HA project with X10-based components and a Linux/CM11A controller.

Is there any list available of all of the 2-way modules (perhaps from all vendors) that will appropriately respond to status requests?

Thanks,
Jason

[Brian H]

I have not seen a list of 2-way modules. They are not too popular and I know X10 discontinued the LM14A; AM14A and AM15A. The only two way X10 module I know of is the RR501 Tranceiver. It has a appliance module type outlet and you can send it a Status request and it will answer back.

Smarthome use to make some two way X10 compatible modules also discontinued.

Their Insteon Modules can have an X10 address added to them and they will respond to a status request, but if it is a dimmer. It reports a %dim that I believe X10 chose not to support.

[jantman]

Ok, thanks.

I guess X10 is out (so much for the Firecracker kit I just bought) and I'll be looking around for something else...

[Tuicemen]

The X10 flood lights are two way as well.

[HA Dave]

I'm going to be doing a HA project with X10-based components and a Linux/CM11A controller.

So you HA project... is this for a class or something... or is there something your actually trying to accomplish?

[jantman]

So you HA project... is this for a class or something... or is there something your actually trying to accomplish?

Just for me. My original plan was to do lighting control for my den with the big LCD tv, and also build motorized blackout shades.

For initial testing, I decided to start with something I always wanted (as I'm a pretty heavy sleeper), a light-based alarm clock. I got the Firecracker kit, plugged a 60W lamp into a lamp module, plugged the CM17A into the computer, and wrote a little cron script (using heyu) to progressively brighten the light at a certain time in the morning.

The major issues started to become clear when I a) shut the light off with the remote I mistakenly left next to my bed, so the next "bright" command didn't make it as bright as it should have been at that time and b) changed anything from the remote instead of the computer.

Having programmed quite a few high-end remote controls (Universal MX-810 and 980) and dealt with TVs, DVD players, etc. that have one IR code for on and off, I've realized how important it is to have controls that can actually do what I want - in this case, adjust a device (on/off or dim/bright) based on what the current value is.

I haven't gone through all of the options yet, but given that I've seen intermittent loss of X10 signals in one room that's about 15x17, I think anything power line based may be a poor choice if I want reliable control (or, if power line based, it really needs to be something that can confirm that a command was received).

I'm going to evaluate all of the options out there, but cost is a factor. I'm also considering a custom system, using microcontroller-based relays/dimmers communicating via a serial protocol with an Ethernet-enabled controller in each room.

[HA Dave]

..... My original plan was to do lighting control for my den with the big LCD tv, and also build motorized blackout shades.

It was actually my Home Theater that got me into Home Automation.

For initial testing.............   wrote a little cron script (using heyu) to progressively brighten the light at a certain time in the morning.

I read a lot from linux users that start projects. But it must be more difficult than it would seem... I haven't read any success stories that I can recall.

... I think anything power line based may be a poor choice if I want reliable control

With the exception of battery powered devices.... everything electrical is powerline based. PLC works really well with HA too!

I'm going to evaluate all of the options out there, but cost is a factor. I'm also considering a custom system, using microcontroller-based relays/dimmers communicating via a serial protocol with an Ethernet-enabled controller in each room.

Good luck with that! What you want to accomplish is pretty simple HA stuff. Done with off-the-shelf items and windows based software it shouldn't be much more than a weekend job.

[jantman]

For initial testing.............   wrote a little cron script (using heyu) to progressively brighten the light at a certain time in the morning.

I read a lot from linux users that start projects. But it must be more difficult than it would seem... I haven't read any success stories that I can recall.

I've finished quite a few projects, though none in the HA realm yet. Mister House is one big linux HA project. One of the things that got me started on this idea was one guy's home power monitoing system, based on Linux. I'm sure there are many success stories. But like my simple time-based script (which, in terms of the software at least, works fine) it wasn't really a large enough undertaking to say much about once it was finished.

... I think anything power line based may be a poor choice if I want reliable control

With the exception of battery powered devices.... everything electrical is powerline based. PLC works really well with HA too!

I was referring to powerline communications... as opposed to communications via wireless or (non AC-carrying) wire.

Good luck with that! What you want to accomplish is pretty simple HA stuff. Done with off-the-shelf items and windows based software it shouldn't be much more than a weekend job.

Well... firstly, it may be difficult for people to believe, but I haven't owned a Windows-based computer in 5 years. I guess I should mention that Linux (administration and programming) is my day job, my night job, and my weekend job. I'm also quite a bit of a tinkerer, and the only thing keeping me from building a totally custom system using microcontrollers is the cost. Ideally, every device would be networked over IP, but the cost of building all of that hardware isn't worth it. I'd be very happy with plug-in relay "modules" (or even boxes with 6 or so of them) in a few places in each room, all hooked up via 4-conductor wire to an Ethernet-based box (one per room).

Mind you, if I were building a house, I would have implemented total "smart" control of every light and outlet at the breaker panel.

[HA Dave]

Mister House is one big linux HA project.

Yeah... I knew that.

I was referring to powerline communications... as opposed to communications via wireless or (non AC-carrying) wire.

Yeah... I knew that too. I was trying to explain that powerlines are really VERY predictable.

I'm also quite a bit of a tinkerer, and the only thing keeping me from building a totally custom system using microcontrollers is the cost. Ideally, every device would be networked over IP, but the cost of building all of that hardware isn't worth it. I'd be very happy with plug-in relay "modules" (or even boxes with 6 or so of them) in a few places in each room, all hooked up via 4-conductor wire to an Ethernet-based box (one per room).

Mind you, if I were building a house, I would have implemented total "smart" control of every light and outlet at the breaker panel.

Many of us here at the forum already have smart homes. Although I am sure you could invent much better systems than we currently use... I think I will just stick with what I have. I (and millions of other X10 customers) can assure you.. X10 works very well.

[hawk1]

Don't know if you would want to do this but, have you thought of using virtualbox with a windows guest?  I have Linux Mint with virtualbox running a windows xp as guest.  Currently i'm running an Ihouse client and it works great.
I just tried to switch to Linux about 6 months ago but, I haven't had much luck with automation using it. But I am still a Newbie at Linux.   

[jantman]

Many of us here at the forum already have smart homes. Although I am sure you could invent much better systems than we currently use... I think I will just stick with what I have. I (and millions of other X10 customers) can assure you.. X10 works very well.

I'm simply going on the theory that if I already have such unreliable communication inside one room, all on the same phase (missing quite a few commands) then perhaps it's better for me to test something with more robust communications (Insteon as a start) before I make a major investment. And consider whether I want to control some specific things with simple relays.

Don't know if you would want to do this but, have you thought of using virtualbox with a windows guest?  I have Linux Mint with virtualbox running a windows xp as guest.  Currently i'm running an Ihouse client and it works great.
I just tried to switch to Linux about 6 months ago but, I haven't had much luck with automation using it. But I am still a Newbie at Linux.   

Somewhere I have a VirtualBox machine setup with XP, mainly for snooping serial communications when I'm writing Linux-based replacements for proprietary software. At this point, I'm totally dedicated to Linux. Being that I work with Linux for a living, and am paid to solve problems, "it's easier with Windows" isn't an excuse I can use.

Windows and Linux (and any Unix, or even the core of Max OS X) were designed two totally ways, with two totally different thought patterns, and totally different goals. When you move from Windows to Linux, a lot of things simply don't make sense. Likewise, I've been using Linux long enough that Windows doesn't make sense to me. Things like having to update software through all different ways, rather than just typing two words at the command line and having everything updated... or with Windows, not being able to do *everything* I can do in front of my computer from a slow connection on a cell phone. Most importantly, especially for home automation, I want something that will run on a SMALL computer (think Linksys router, $100 single-board computer, etc.).

I appreciate all of the advice, but I'm already sold on Linux. At this point, the big decision is just going to be what technology (X10, Insteon, Z-Wave, etc.) gives me what I want, and which one will minimize the amount of custom hardware I want to build.

To tell where I'm coming from, I've done one HA project in the past - a few years ago, when I was in school and living in an apartment with a few roommates, I realized that our thermostat was horribly inefficient. So, I ripped it off of the wall, replaced it with a USB-controlled relay board (4x 30A relays), and hooked that up to a Linux-based computer which also had a Dallas 1-wire temperature sensor in each room. I wrote software that polled the temperature every 1 minute, and included all of our school and work schedules, in excruciating detail. The heat was only on when we were home. It understood that if I was the only person home over the weekend, it could ignore the temperatures of the other bedrooms. It even kept data on previous heating cycles, so that I could set the temperature to be 65 degrees at 8 PM on Tuesday, and it would kick on just early enough to heat the apartment to 65 degrees at 8 PM.

Bottom line - I want to be able to positively tell the status of everything that's controlled (even if that means using light sensors on every fixture), monitor everything, and have positive control. My personal feeling on HA is a lot closer to industrial control/automation than X10. If it weren't so expensive, I'd have everything running on SCADA boxes...

[HA Dave]

..........  I appreciate all of the advice, but I'm already sold on Linux. At this point, the big decision is just going to be what technology (X10, Insteon, Z-Wave, etc.) gives me what I want, and which one will minimize the amount of custom hardware I want to build.

If you've done ANY searching and reading whatsoever... than you already knew when you posted... that your own experiences with your one room test/trial were flawed. X10 has been around automating homes... for decades. And as a Linux user.... you should have already known that X10 is your best chance of automating (a home). At least when using off the shelf products.. with Linux as your chosen OS. 

Your reference to a desired level of control as "industrial control/automation"..... does seem excessive... for the use of X10 (or any Home Automation setup). I am pretty sure.... what you will find [with industrial level monitoring and control] is system failures within your automation setup will far outweigh any normal failures within the infrastructure it is built on.

I truly wish you the best of luck with you persuit.

[pconroy]

I'm going to be doing a HA project with X10-based components and a Linux/CM11A controller.

Jason,

What are your thoughts on implementation?  Going to do it in Java, C++, C?

Just curious.
Thru the years, I've written software to control my CP-290, Powerlinc and CM11A in Java and C, and am now doing it again in C++.
I don't get to program much professionally anymore and I do it simply to refresh my skills.

I need a *real* project to work on to keep my interest. 


I'm using NetBeans and Cygwin so I can move stuff from Windows to my Ubuntu boxes without any hassle.

[Brian H]

One point to think on.
Insteon is mostly power line only signals also. The same signal suckers and noise that can be a problem with X10. Can effect Insteon also. Maybe not quite as badly as the modules relay the messages.
Many Insteon users think each module has both RF and power line capabilities. They do not. Smarthomes/Smartlabs description of "Dual Mesh" is not too clear on that point. Only one or two recently released modules can do RF and Powerline.

Though the modules are two way and you could do status requests if needed.

[jantman]

What are your thoughts on implementation?  Going to do it in Java, C++, C?

This may be due to my relatively young age, or that I'm really a sysadmin by trade, but I'll most likely be implementing whatever I do primarily in a scripting/interpreted language (Python or PHP). I'm not *so* worried about performance (with the time it takes to send X10 or Insteon signals, I'm not worried about a 1/60 second delay) and I'm a lot more interested in being able to reuse code to implement a web interface as well as a CLI interface. Since what I do will primarily be for *me* (I'm not looking to rewrite Mister House) I'm fine with just makign exec() calls to control the devices.

I already have MythTV running for all of my entertainment stuff, using a high-end Universal remote and an IR receiver on the computer. So I figure a lot of it would be using scripts called by IR signals, or using an app for my Droid.

One point to think on.
Insteon is mostly power line only signals also. The same signal suckers and noise that can be a problem with X10. Can effect Insteon also. Maybe not quite as badly as the modules relay the messages.
Many Insteon users think each module has both RF and power line capabilities. They do not. Smarthomes/Smartlabs description of "Dual Mesh" is not too clear on that point. Only one or two recently released modules can do RF and Powerline.

Though the modules are two way and you could do status requests if needed.

My thought process is more or less this:
1) Whatever I do needs 2-way communication. I see it as useless to have control unless I also have status reporting so I can know what state something *actually* is in. This is thinking about both unreliable signals and the use of RF keypads/switches (i.e. control signals sent from something other than the computer). It's just an added bonus that this would allow me, in software, to confirm that a command was acted on properly and repeat it if not.
2) Insteon seems to have a more robust communication model, as the commands are sent 3 times, ACKed, and repeated. That, in my opinion, makes the extra cost over X10 worth it.
3) I don't want to buy into a system that either I will outgrow (when I automate more or move) or which has performance that doesn't satisfy my long-term plans (no matter how elaborate they are).
4) I'm already building a bunch of custom hardware for IR control stuff, control of window shades (lanyard shades with attached stepper motors), etc. If I have to do a bit more hardware work, it's no big deal.

As to the reliability of X10, keep in mind, I'm a sysadmin and a network guy. The idea that something has wires connecting it and could be expected to lose more than 1 in 1,000,000 packets in normal operating conditions just doesn't seem acceptable.

If there were any pre-made HA hardware that was inexpensive and just used 2 wires (*not* power wires - telephone wire or 2 conductors of Cat5) I'd buy it in a second.