Product idea

[CraigINPA]

While I was setting up my Christmas show for this year (144 LOR channels plus dozens of X10 devices) and thinking there has to be a better way...

The background is that I have a large stone house. In each window (23 of them), I have a wreath with the plug hanging inside the room and the wreath hanging outside on a thick ribbon. In most rooms, all of the wreaths in one room are connected to a single X10 lamp module in that room. In some rooms, each wreath has its own X10 lamp module because it would be unsightly or dangerous to run an extension cord from window to window. Outside, I have icicles, c7's, and snowflakes on the eaves and gutters, mini lights in trees, more wreaths on the garage and wooden fence, and a couple of mega trees to balance the whole "classic white light look". Over the years, I've doubled up most of the lights and now can flip between white and red, and in some areas tripled it up with green. I don't like the look of wires running down the front of the house, which is why the wreaths are wired inside. Each year I spend a lot of time hiding the wires going to the eaves and gutters to keep the house looking like a holiday postcard during the day.

I'm now at a crossroads. I'd like to extend the red and green to the wreaths, but X10 is limited to slow fades and on/off, plus it's got a 1-2 second response time, which makes it virtually impossible to use it like the other 38k lights connected to LOR controllers. LOR controllers in the rooms would be an option but for the fact that these wreaths are spread all over the house and I can't justify buying 16 8 channel controllers, having a lot of unused capacity in some rooms, and drilling holes room to room for the cat5 to run, or buying 16 ELL's in addition to the controllers to avoid running the cat5.

My product suggestion is a 4 channel controller with an integrated ELL, packaged like a wireless router is today, but instead of cat5 connections, it would have electrical sockets, and instead of a tiny power supply input, it would have a business machine cord.

I'm sure I'm not the only person who needs just a few channels in a number of geographically remote areas where running cat5 would be impractical and the cost of an 8 channel plus the ELL not justifiable.

[kzaas]

That sounds like a handy product but most likely with the capabilities that you are talking about, it could become expensive although cheaper that buying a bunch of eight channel controllers and ELL's.

Good Idea, make it happen Dan, I'm sure you would have some buyers out here:).

[jimswinder]

Not that I know...but I would imagine the cost in parts for 16 channels vs 4 channels would probably be less than $20.

So...probably not worth making for the difference in price...

[kzaas]

jimswinder wrote:

Not that I know...but I would imagine the cost in parts for 16 channels vs 4 channels would probably be less than $10.

So...probably not worth making for the difference in price...

He is not just talking about a 4 channel controller, he would like to see it with an integrated ELL for remote operations. I'm sure it would cost more than a sixteen channel controller but I would think it would be cheaper for them to build both items integrated together.

[jimswinder]

kzaas wrote:

but I would think it would be cheaper for them to build both items integrated together.

only if there was a HUGE demand for that kind of a product..

like anything manufactured, the price starts dropping only when you order huge quantities...

[CraigINPA]

Unfortunately, because the ELL's are linking rs485, they're expensive.

An alternative would be to do some code development on the pc to operate over a ip-based lan. Then we'd be able to use inexpensive 802.11g and 802.11n products to link the wireless mini controller (WMC) (my new name for this idea). The set up would be pc to a wireless router, wirelessly connecting to the WMC via ip protocol, which would then control 4 ports. The WMC would be about the size of a thick paperback book with 4 sockets and 1 business machine plug. A setup program, like the hardware utility, could detect WMC's on the network (by scanning every IP address for a particular open port) and then set a permanent IP address by telling the WMC what its permanent address is. The LOR software would have to do an abstraction layer to map controller numbers to ip addresses, but that's pretty trivial. Using a single chip controller, like the Realtek RTL8191SU, it wouldn't be that difficult to directly replace the SIO used for the RS485, albeit that there would be additional code required to handle the tcp/ip on the controller.

Yet another alternative, and one I was thinking of tinkering with myself, was to use the $69 US Converters BF430 product with a cheap wireless router, or the Amplicon product that does RS485 to 802.11g directly, to not require the expensive ELL. As an old-type soldering iron junkie who built his own modem back in the day, it's not inconceivable that I wouldn't even roll my own solution with a couple of chips and a microcontroller.

[k6ccc]

I understand the problem well. Just ordered an 8 port controller that will have 3 lights on it (and if I wanted to I could easily put two of those on the same circuit). Using WiFi as you suggested would likely work well. Also remember that they are coming out with control over power line that will first come out with the Cosmic Color Bulb. That might be a concept for your mini controller. Another possibility...

[beeiilll]

Well if you can get Winder interested in it, then the price will be low since he will be able to use 1000 or so - LOL.

What do you say Jim?

[Guest guest]

Beeiill, that sounds like an excellent idea to me too! I could use a few of those 4-channel wireless controllers myself in my display. It would definitely cut down on my extension cords!

Would be very interesting to see just how much they would cost if they could be produced!

[WilliamS]

I would like a 5 channel but could make a 4 work as well. There has to be a way, but like what was already said a 16 channel controller in terms of parts I would by best estimate, now this is just part is only about 15 bucks of hardware. That beind said, as it was already stated by Jim that the EEL is the expensive part. Getting it to work on a differnt network would be a way to make it cheaper but then you risk outside interference.

[CraigINPA]

The ELL is the expensive part because it's a low production, special need item. All the thing does on a technical level is take in RS485 (which is a serial data stream) and broadcast it serially, a really simple and basic task.

Compare that to a 802.11 router in terms of complexity and it's like looking at the difference between a Ford model T and a formula 1 race car. The router is only cheaper because they're making tens of millions of them, which drives the price down. But, because they're making so many, there are companies making components that make the process of building one easier. It's those components that would make this idea viable for the low production need we have.

In just an hour of googling, I found several single chip solutions that included the radio (802.11g and n, wep, wpa, snmp), tcp/ip stack, and even a management web browser. The difference between using this chip and the serial chip for rs485 comes down to implementation more than an exhausting bill of materials to add to the present controller.

A simple approach to make this work is to use a node and broadcast design that mimics rs485. At start up, each node will get a tcp/ip address either from a dhcp server or statically (that's a feature of the chip itself). Since the controlling pc isn't expecting back-channel communication from the nodes, the controlling pc can use udp broadcast to talk to the nodes. Udp broadcast works within a single network, so as long as there are no repeaters, every node will receive the message simultanteously. This limits the size of the network to 254 nodes per tcp/ip segment. Multiple tcp/ip subnets could be employed to bypass this limitation, with the only caveat being large spanning trees may present a small delay per hop that after some number of hops might amount to a show timing issue. Even in the physically largest LOR displays, it's unlikely that anyone would need more than a few wireless access points, so this possible delay is largely theoretical.

To implement the hardware utility's ability to see what nodes are on the network, you'd implement a response request to a particular type of broadcast message. e.g., When the controlling pc wants to know which nodes are out there, it will send a udp broadcast to the tcp/ip network asking the nodes to respond to it with their node numbers. The nodes respond via tcp/ip to the controlling pc, which allows the pc to know actual nodes ip addresses and their node numbers. This also the basis of how to speak to the controllers individually, such as when it's necessary to download new code to it or do an initial set up (such as supplying a wep key).

RS485 is a simple serial protocol that, when run at low speed, makes it possible to span long distances. It was great in 2000, before wiresless networks were ubiquitusly available. Wireless networks today span the same distances at speeds 5000 times faster with no latency. And, most importantly for the original concept of the product idea, is that a wireless node makes it easy to place a small node anywhere you need it without worrying about how to get it connected to the network or the expense and complication of a separate ELL.

Craig

[Paul Roberson]

k6ccc wrote:

control over power line that will first come out with the Cosmic Color Bulb. That might be a concept for your mini controller. Another possibility...

I would think this would be the best solution. I would love for ALL my controllers to operate this way.:cool:



BTW. I have no knowledge of such things and I did not stay at a holiday inn express last night, but being able to just plug a controller into a power line to power it AND receive lighting commands sound good to me.

[CraigINPA]

Control over power line is problemmatic if your display spans multiple residences or multiple drops from a transformer, or your neighbor has a display, or you have a power line speaker system, or you have less than ideal connections throughout your electrical system, or you have large power loads (A/C, heat pumps, etc...). I've been there with X10 for years, and there is not a day when I don't silently curse some part of it not functioning.

[Paul Roberson]

CraigINPA wrote:

Control over power line is problemmatic if your display spans multiple residences or multiple drops from a transformer, or your neighbor has a display, or you have a power line speaker system, or you have less than ideal connections throughout your electrical system, or you have large power loads (A/C, heat pumps, etc...). I've been there with X10 for years, and there is not a day when I don't silently curse some part of it not functioning.

Guess INSTEON has me spoiled. Works flawlessly.

Not trying to say there would/could not be problems, but what I have put in bold above would not be a problem for most LOR users.

[CraigINPA]

Living in a house that was built in the post-and-tube era of wiring, with every variety and type of "upgrade" over the years, and with period "vintage" pieces remaining (like ceramic-bodied switches and a 200 pound cast iron chandelier), one becomes quickly unenthused when it comes to any powerline communications option. My house, which admittedly is the size of an antibellum mansion, is on two different transformer legs (400A service) with two 200A panels, put in 50 years and 100 feet apart, each serving a different part of the house. It's not unusual for those who have staggering amounts of lights to have a dedicated 200A (or more) service for their display. And, while I'm a realist that knows most people don't have the unusual needs I enumerated, I also know that people who are in the situation I am in or have multiple service panels are more likely to need the wireless product and will run into problems with a powerline solution.