Developing Pixelnet compatible alternatives to the SSC

[Steve Gase]

Along the line of "Kick back and talk about anything"...
 
Is it permissible to develop alternatives to the SSC (which support TM180x chips) in order to have a solution that supports other chips (28x1)?
 
We know that SSC does not support anything but TM180x, and RJ is not interested in expanding beyond that.  There are other suppliers, and other chip sets out there with lower prices, different form factors, etc.  What if other community members were to develop alternatives to the SSC that still are compatible with the rest of the infrastructure?  (smart hubs, pixelnet, EtDs, Conductors, ...)
 
Obviously, to play in this game a developer needs to know more about the pixelnet system.  There would be rules dealing with max power draw that need to be respected.  I'm wondering if RJ shares this sort of info.
 
I'm just not sure where the lines are drawn... if Bobcat, xLights, and Nutcracker development is encouraged... would pixelnet-compatible development also be allowed?
 

[n1ist]

The pixelnet spec is in the wiki.  One thing to watch out for is the pinout for pixelnet in the spec is wrong.
/mike

[dpitts]

The Pixelnet spec in the wiki is what software should send out to P-Dongle. The P-Dongle and Etherdongle add a small header to Pixelnet going to SSC.

The start sequence is...
0xaa
0x55
0x55
0xaa
0x15
0x5d

Then 4096 bytes of Pixelnet data.
   

The pixelnet spec is in the wiki.  One thing to watch out for is the pinout for pixelnet in the spec is wrong.
/mike

[sittinguphigh]

The idea is cutting the price down so more people can participate in DLA RGB smart strings.

There are other company's that have these same TM1804IC chip.
There cost is .25 a node. But you have to buy 20 to 50 at a time.

There is the shipping cost which plays a big part in it.
Try to find other ways to ship it to the US would help. Even if slower.

Getting coops together and sending them the cheapest way could help.

There are different chips to choice from also.
It would help if RJ would let us know why he pick the TM1804 chip.
There may be a good reason.
Quality may be a reason.

Get a other companies to compete with Ray may help bring the price down.

 

[Steve Gase]

The Pixelnet spec in the wiki is what software should send out to P-Dongle. The P-Dongle and Etherdongle add a small header to Pixelnet going to SSC.

Yes. 


What we now need is the interface between the smart hub and the SSC to be shared.   It should be formally defined so that future devices (DLA and non-DLA) would be expected to remain compatible with the system.  It shouldn't be based on reverse engineering which can be susceptible to future changes and improvements.

[RJ]

Just use other systems you like better. You know how I feel about people doing such stuff. But maybe the point is to aggravate me?

Seems the same few people are the ones always starting this stuff up.

RJ

[Steve Gase]

RJ,
 
The point is not to aggravate you or in any way put down what you've already done. 
 
Instead, it is to suggest ways that the community can fill out pieces of the overall solution that you cannot provide because you are stretched to cover so many other areas.
 
There are some bright people in this forum (I am not counting my self in that number).  If turned loose (with your right to review and reject) they could help you realize your vision. 
 
I don't think anyone wants you to sacrifice the new and cool stuff you are doing with EtD, Conductor, Slave, etc. etc.  There are no doubt many additional things in your head, and on your bench that we cannot guess.  These one-off requests discussed in the forum are valuable to us, but not at the sacrifice of the coolness that has yet to emerge.
 
But, in order to fill out the solution and realize the even greater value that can be seen, there are elements that others can help with...  a Pixelnet-to-8-Universe-DMX converter is one example...   a SSC corrolary with 28x1 chip support is another.
 
What you've put together is awesome...  the financial investment that people have made in your system is also incredible.  If my wife realized just how many DLA kits I've obtained and built she'd scream.   But if we can leverage that investment into even more areas, then its just a testement to how powerful your design had been from the start.
 
There is no doubt that you could do it all, we are just not asking you to do it... just to let the community help with that far-reaching vision.  You can protect your ideas with licensing restrictions.  (ie. you can prevent people from monetizing any designs that are based on your pixelnet specifications)
 

I appologize if my suggestions are seen as rabble-rousing.  I was just trying to help the community take a step further along the trail you've blazed.

[RJ]

To answer why the 180x chipset?

This is no secret an has been posted.

When I started looking at pixels long before SS were ever posted about. I found the 5 volt limits to be a problem. I found a number of 12 volt units were being used in china but no regular strings. After trying them I found they were preferable to me since they were 12v and could run longer strings, longer cables to them, could power them on cat5 and they seemed to dim a little better.

So I spent $1250 of my own money to have molds and pcbs made to build them as strings also. This gave me a complete set of styles for everyone.

As always I want to keep it as simple as possible for new users and since there were no other chipsets with 12v at the time I put them out as one supported protocol.

Now the ssc existed already and had been tested with all the original chipsets as they are all easier than the current firmware.

In fact simply look a one of your version 1 ssc pcbs. The surface mount resistor spot and pad for the 4th wire is still on them.

So you see there was no other choice for 12v at the time.

And with all he flurry of post from those using 5v systems How dumb using 12volt was and all the issues we would have I did not expect them to make their stuff 12v capable and have 12v light made on their chipsets. I expected we were the only ones that stupid.

RJ

[sittinguphigh]

RJ did invest a lot of time and money into the system. And it works very good.
I don't think RJ's choice is not the problem.
The smart string system is being corrected and improve as it is being used.
This is normal product manufacturing developing.
To go back and reconfigure and reconstruct can be costly and time consuming.
And will there be a big improvement and will you accomplish what you want.
If there are problems with multi chips connected to the system that DLA has no control over what then.
It seems hard enough with just one chip and problems that come up.

You have to ask yourself what are you trying to accomplish.
If it the cost factor?
I don't see a lot of cost difference with any other chips. comparing apples to apples.
Again if it's the cost factor.

Many people would like to use the RGB smart strings but can't afford the cost factor.
Others would like to use them on there whole house but can't afford the cost.
 

That being said, I don't think talking about is a bad thing. To think about the possibility is how this site has become what it is. 

If cost is a factor I think you should look to see if you can find smart strings with different chips that are more affordable and worth the cost difference first. If not it is a waste of time.
See if there are ways to ship less costly.
See if you can figure out a easy way to have coops and cut the cost down.
We have a lot of people using these smart strings and more wanting to.
That adds up to a lot of buying power.
Ray has brought down the cost when more people were competing for DLA business.


With out RJ and the time, effort and cost he put into DLA wouldn't be here.
I thank him for that. This is RJ's baby tread lightly.

 

[rm357]

Please take the time to go to the wiki and read what is there.

Specs are there
Pixelnet protocol is there
Why 12v nodes is there
Current draw is there

It's an open protocol...
It's all there.

The hard rules if you want to roll your own.
#1 - roll your own - new hardware, new firmware, new software
- writing new firmware for someone else's hardware is forbidden unless you get permission
- don't duplicate something that already exists

I may be out of line, but I think it is perfectly ok for you to design and code a pixelnet to 2811 controller, but it needs to be your own design from the ground up.

And this is just a personal pet peeve, but please don't imply that someone else should do it, and how great it would be if someone else would do it, and how the world would be a better place if someone else would do it, and why won't someone else design this thing that I want.... Would you like a little cheese with that whine.

This is DIY. Do it yourself, and offer the results of your ingenuity and hard work to the community.

Ok. Sorry that turned into a bit of a rant...
But please, go read the wiki.
Thanks,
Robert

[dpitts]

Please take the time to go to the wiki and read what is there.

It's an open protocol...
It's all there.

The hard rules if you want to roll your own.
#1 - roll your own - new hardware, new firmware, new software
- writing new firmware for someone else's hardware is forbidden unless you get permission
- don't duplicate something that already exists

Robert I think your intentions are good but your statements are inaccurate. Pixelnet is not an open protocol at least not on this forum. RJ just said he does not want anyone to create an alternative to his SSC to control other lights. He wants Pixelnet to control only TM18xx lights and all controllers will be created by him. That is his vision and his system. Sure you can create anything you want and do all your own design but without the blessing of RJ you will not be able to post or share your design with the people that would use it (DLA members). It is more than just creating your own design, it is creating your own design that is within the RJ's vision. I have my own firmware and software completed and board design drawn up for WS2811 and 4-wire pixels to run from Pixelnet but I have not released it because that is not what RJ wants. So although your rant was inspiring it does not pertain to the OP's original question which was definitively answered by RJ. No creation of alternatives on this site.
 

[jess_her]

From what I have taken away from this site is yes RJ develops this stuff. SO he has a very large say on how this is driven. He won’t stop any development of new features if it fits into his vision, bottom line it’s his.  BUT to have somebody develop something then expect RJ to support it or help support it isn’t fair to RJ especially since he is devoting his time and money.
 I would be very careful, this might become more trouble than it’s worth to him.
Jess

[rm357]

I haven't seen a post from RJ against development of a new pixelnet device. I may be wrong, but I have not seen that. I have seen that he has no intention of developing a solution for other pixel types.

I have seen very clear posts against repurposing of someone else's hardware and reverse engineering of existing items.

I would agree that if you have a new device in mind, you should pm RJ to get his feelings on it. But having said that, there are several items that have been made by others that work in that arena - specifically the splitter, the discussion/work on strobes and other items.

One of the things that people don't realize is that it takes many hours and a significant cash investment to create something new. Since the ssc works for the 1804 type pixels, the cost of duplicating the ssc for another pixel type is a waste of time and money - no "new capability" would be created, there is nothing that you can do with the 2811 nodes that you cannot do with the1804 nodes. The only reason to develop the 2811 controller would be to save some else a few pennies per node. People in their own woe is me I can't afford that pity party Forget the fact that they are being saved dollars per node over something like firefly or the LOR CCR...

[t.jo13]

Every time a topic along this line comes up on this forum,it kinda reminds me of back in the 70's & 80's when people had Fords and wanted to use/modify chevy parts cause they were cheaper.

Simple solution: Buy a Chevy 

[taybrynn]

I think once you chasing the chipsets with firmware, you know you'll never get out of that and it may actually make standards harder/slower to happen.

One great thing with the SS choice oof the 180x chipset was that RJ declared SOMETHING which the supplier (Ray) could then start building RGB devices for ... and which we all benefited from.
At some point, 10 new flavors doesn't really add anything useful to any of us.  The only reason seems to be price.  And again, if the 180x chipset becomes the next mini someday, the price goes down (and already HAS) and thats the goal, is options which are reasonably priced.

I don't know if this makes any sense at all, but I have sometimes wondered if a bridge type device would make sense ... kind of an xlights type hardware device ... where you tell it what the incoming data and power wires are (say output from a SSC) are and it can bridge that into various RGB light types for you and could potentially convert each incoming channel into a different RGB light type as configurable via. a web interface or something.  Again, I don't see myself wanting to use this ... because then I have to keep track of various types of lights and hardware and it gets complicated and confusing ... why?  because I wanted to save a few bucks?

Part of me also wonders why smart string would need really to support all these other things when we can already get a j1sys or sandevice item thats also E1.31 and then just run our etherdongle alongside those ... provided we run software capable of it or xlights.  I know the conductor doesn't support that, but I think you have to make choices and ever choice has its own limits.