arduino on WIFI with vixen drivers for g35 lights

[roboengr]

I wish to discuss Using the arduino with vixen to control the g35 lights. Has anyone done this to date?

------------------------------

I hope KC (vixen) is designing a pixel control into Vixen 3.0. I have not had time to check on the beta. Drivers will have to be written for then new version as I had done that for 2.5 using vb2010 over Ethernet.

================
AH! imagine using the home router to send a sequence via IP address?

Yes, I have done it. Was my new project for 2011.

I also ran a Renard board from wifi no Xbee. This was cool.

I hope this will sper on a discussion not about design but options.

-----------------------------------------------------------------
Written from a licensed professional engineer

[roboengr]

The design ideas are going away from a good design and moving to code that is not simple or easy to read. I want some input as to how everyone is wanting it to be easy and cost effective.

This is WIFI g35 lights, Any one interested? Easy, programmable, and low cost.

[Skunberg]

The design ideas are going away from a good design and moving to code that is not simple or easy to read. I want some input as to how everyone is wanting it to be easy and cost effective.

This is WIFI g35 lights, Any one interested? Easy, programmable, and low cost.

Are you referring to GE Color Effects G35 Lights? G35 is a size designation like C9.

[roboengr]

Yes the GE Color Effect lights by Santa's Best Craft. some times referred to as g35 on the posts.

[roboengr]

We can do this for all Color effect lights as they are the same controller.

I have been following the others on there design originating from Darc's work. I have studies them and build some. There has not been an easy to use solution yet. Lets replace the original board. Synchronize the lights strings across the house or a board. This can be done using wireless technology. XBEE, WIFI and other pieces.

I would like to co develop using this forum a new design. The original controller is very lacking. They (GE) are releasing even another incompatible remote to previous lights.

First we must get the requirements:

* Low cost to add to the high cost of the lights
*Wifi
* keep the original case for use
* Vixen compatible with drivers either in protocols such as Renard, DMX, ect...

Perhaps it could use any RGB string and be flexible.

Some have done high level Propeller chips but that is not easy for most. The arduino is free and they are about to release a 32bit version. we can develop a pic version for those preferring that Micro.

[joej85]

A "wireless" replacement controller board for the GE Coloreffects lights was already created called the "Colornode". http://www.digitalmisery.com/projects/colornode/

This board can communicate wirelessly (Not WIFI) with a controller which is connected to a computer. It uses a custom protocol and is not compatible with DMX/Vixen, etc. It uses the HopeRF RFM12B module for wireless communications with the controller. The wireless communication of this board is not fast enough (115.2kbps) to handle real time individual control of multiple strings of lights via DMX, etc.

I picked up several sets of GE Coloreffects lights on clearance last year, and I am in the process of making a new hardware controller based on the Colornode concept, except my version is going to use a wireless communication module based on the Nordic NRFL2401+ IC. This wireless hardware can communicate at 2Mbps which should be adequate for real time individual control of many sets of GE Coloreffects Lights. I have performed some range tests on the wireless modules and they can communicate up to at least 100 feet away from the controller, which is all the range I need to cover all the lights on my house.

Initially, I am just going to be trying to get everything working and just using the controller to set up some standard sequences of lights in each of the strings. My ultimate goal is to be able to use Vixen to control each light of each string via DMX or some other protocol wirelessly. The total cost of the controller boards that I am making is going to be <$10 including the printed wiring board and all components. The NRF24L01+ wireless module is only about $2.50! (The RFM12B module of the Colornode costs about $7 at sparkfun).

WIFI would be ideal, but I was not able to find any way to easily interface an Arduino to WIFI that didn't require an expensive WIFI interface module. This is why I chose the NRF24L01+ modules that are fast and cheap, but unfortunately do require a dedicated master controller to send the wireless commands.

If I can just be able to control canned sequences or run non real time download sequences on all my lights with this setup I will be happy, but I'm hoping to take it all the way to being able to control individual pixels wirelessly with DMX.

[roboengr]

A "wireless" replacement controller board for the GE Coloreffects lights was already created called the "Colornode". http://www.digitalmisery.com/projects/colornode/

I have already seen this design and it has many flaws that rinder it basically usless to the point as you said the driver selection and reseting in the ability to keep things synchronized. Large inprovments are left to be done. The canned sequence has its place, but it needs the Pixel control and a coordinator that can real-time down load an new sequence remote.

This board can communicate wirelessly (Not WIFI) with a controller which is connected to a computer. It uses a custom protocol and is not compatible with DMX/Vixen, etc.

More flaws as you described.

I picked up several sets of GE Coloreffects lights on clearance last year, and I am in the process of making a new hardware controller based on the Colornode concept, except my version is going to use a wireless communication module based on the Nordic NRFL2401+ IC. This wireless hardware can communicate at 2Mbps which should be adequate for real time individual control of many sets of GE Coloreffects Lights. I have performed some range tests on the wireless modules and they can communicate up to at least 100 feet away from the controller, which is all the range I need to cover all the lights on my house.

As have a bunch of use. It is interesting on the communication Speed. Is that at the 100 feet? how did you arrive at this figure? Just curious. $2.50 is not bad. WIll you be releasing the design for others her to use?

WIFI would be ideal, but I was not able to find any way to easily interface an Arduino to WIFI that didn't require an expensive WIFI interface module. This is why I chose the NRF24L01+ modules that are fast and cheap, but unfortunately do require a dedicated master controller to send the wireless commands.

I did do the WIFI with individual controllers in each string. IT works however the flaw in using the yellow jacket is the Shield modified Stack using the existing socket code; as it is bottlenecking the sequence throughput. However it worked well for individual pexels. They information to repair the code is bleak at best. I did not have time to Troubleshoot the problems. I am too looking at a solution to use something like the NRFL2041+IC as you described but with the coordinator that is WIFI. In fact that is what the Arduino project is about on the other thread to incorporate a whole system. I am waiting for the 96Mhz 32-bit Arduino's to arrive for the coordinator.

If I can just be able to control canned sequences or run non real time download sequences on all my lights with this setup I will be happy, but I'm hoping to take it all the way to being able to control individual pixels wirelessly with DMX.

EXACTLY!, that is what we are looking for here. Are you in on contributing to a working system?

ROBOENGR

[joej85]

As have a bunch of use. It is interesting on the communication Speed. Is that at the 100 feet? how did you arrive at this figure? Just curious. $2.50 is not bad. WIll you be releasing the design for others her to use?

The communication speed of the NRF24L01+ is spec'd to be 2Mbps max. which is why I thought it would be a good choice. The $2.50 price for the modules (on ebay) is darn good too. I have been able to communicate at this rate between 2 devices 100 ft apart. I did use a special NRF24L01+ module on my "controller" node which had a LNA and preamp included. Once I get my first printed circuit boards assembled I am going to do a more detailed test. My printed circuit boards arrived in the mail today, so hopefully I will get one or two assembled soon. Once I'm sure its something thats worth releasing I will release it. It is really nothing more than the Colornode with the NRF24L01+ module in place of the RFM12B module.

I am definitely ready to help out in creating a working system.

[roboengr]

[COLOR=#0000ff]
I am definitely ready to help out in creating a working system.

I personally, From industrial side of things, am using the DIGI Xbee modules with a more robust and no need of a Library except to send out the protocol in what ever flavor. I got the devices in a trade for $8.00 each and have 15 units on hand in additional to prototyping devices.

Have you seen this library development? I am sure there is more.
http://maniacbug.wordpress.com/2011/...-started-rf24/

I am thinking we can develop both options. The cost of the the rf24 is the additional code. The similar code will be the protocol to each string circuit from the concentrator. The concentrator will handle the DMX or Vixen or other driver protocol.

We must borrow from the DMX group or employ their help on the DMX protocol code. I have not had time to study it or the other work done on the RGB part of vixen. I am sure that Vixen will have some RGB tools built in V3.0 soon.

In this software I am thinking that a combination, as you stated joe, of fixed routines for the lights. To add to that, pixel addressing directly and maybe in the form of a programmable animation on the fly from a vixen software driver. Meaning the paterns can be downloaded to the concentrator and it plays a sequence from the driver. This could be from the out put of the Nutcracker -RGB Effects Builder. Take a look at the nutcracker software. Nutcracker RGB Sequence Builder http://meighan.net/nutcracker

I have already developed by Xbee boards with through hole technology but have decided to change to Surface mount for more room.

There is on going designs happening this year and it is exciting but we are running behind for this years end. Now note the other arduino forum on the standard and the arduino system design this all has the design noted. Also Note the Forum for the horea design

What do you think Joe?

Roboengr

[joej85]

Have you seen this library development? I am sure there is more.
http://maniacbug.wordpress.com/2011/...-started-rf24/

I have been using the NRF24L01+ Arduino library. I started by purchasing 2 Arduino Uno's and 2 NRFL2401+ modules. I hooked everything up and ran the sample library programs. I played around with the data rages, CRC, Acknowledgments, etc. I ran some initial range tests using the "ping-pong" library examples. With the standard NRR204L01+ modules the range is somewhat poor, something less than 50 feet. I then purchased a special NRF24L01+ module that contains a low noise amplifier on the receiver and a power amp on the transmitter to use for my controller node. With this setup, I can easily get reliable communication over 100ft or more from the inside to the outside of my house. Once I was happy the NRF24L01+ was going to give reliable communictaions, I hacked up some simple code to let me control a string of GE lights wirelessly from the control node. So far so good...

I have designed and purchased PC boards and I'm about to assemble a few to test with. I am working on trying to establish the protocol that the controller node uses to communicate with the string nodes. I want it to be as fast and efficient as possible. The NRFL2401+ can transmit 32byte packets and I am trying to figure out how to best package address/intensity/red/green/blue value for each pixel into the smallest number of packets. It looks like at least 2 32 bit packets are required for 36 lights. If this is done right, the controller node can talk to many string nodes fast enough to be "real time".

The bottleneck is going to be in the controller node, I think. IF the computer is providing real time data, the link between the computer and the controller node must be more than a simple usb to serial connection. The controller node is going to have to have a "real" usb connection or a ethernet or wireless connection to the control computer to make real time data work, at least if we plan on controlling several strings of lights. The controller may need a microprocessor more powerful than an Atmega328 to pull all this off too.

We must borrow from the DMX group or employ their help on the DMX protocol code. I have not had time to study it or the other work done on the RGB part of vixen. I am sure that Vixen will have some RGB tools built in V3.0 soon.

Yes, this is the part that I am least familar with. It might be just as simple as making the controller node be equivalent to a USB DMX transmitter. The controller would receive the DMX data from Vixen (or other), convert it into appropriate packets for the nodes/pixels and send it out. The DMX protocol looks pretty straight forward. I did find an example of using an Arduino as a DMX receiver, so it should be possible to borrow some code for the DMX receiving from this.

It all seems possible. Will there be enough time?