New WS2813b

Am curious myself! single pixel controlling would mean a great definition against the mega tree. According to parts of Rays site the same protocol is used as the 2811 so the SanDevices E682 should control no problems.


So the main advantage of these are the no IC chip on the strip (integrated) and the redundancy agianst single pixel failure (which i've never had before)?
 
wondering how the single chip failure would work. If the BIN line is connected to the second last pixels DO then I guess it strip out the first 3 channels then uses the next three. Otherwise everything would end up out by 3 channels down the line.

Also in thinking about it, how do you know when you have a fault with the output side of a pixel. Most of the time when a pixel has failed to work it's the output of the one before it (that displays fine itself) that actually stopped working.

Cheers,
Rowan
 
It seems a great idea but the way I read the info if you power down and then back up, you effectively create a reset which would seem to shift the pixels by one.

I think as long as the power does not shut off causing a reset, it would just ignore the dead pixel and read from BIN.. again, I am with you, I don't see how it will know it's place in line, but perhaps it does..
 
From the description, it seems like any time it receives data on the DIN it compares it with the BIN and, if there is a difference it sets a flag to ignore the DIN and use the BIN instead. So if a pixel goes dead and doesn't forward data out, it is removed from the sequence and the sequence is shifted to cover the hole (since each pixel strips it's 24-bit frame from the serial stream).


What I wonder is how they do this:


"The BIN receives the data signal and, then compare it with the DIN side after phagocytosis of 24 bit data..."


I assume they're talking about stripping the 24-bit frame, and not about the ingestion of cellular material? ;) ;)


Dennis the Menace
 
Hi, I think the interesting bit is "untill restart after power off". The way I read this (and I might be totaly wrong) is that there are two data signals DO / DIN (which has a delay I'm not sure how that comes into play) and BIN which is just the whole of the data we send to the string (which begs the question how far is the backup good for ?). During normal conditions the IC could learn it's postion in the string by comparing the two sources of data and if the main source (DIN) failed it could use the backup, however with no memery onboard as soon as the power is removed the pixel no longer knows it's postion in the string and the string fails as any other does. This would be great for something like P10 screens where one pixel fail would not take out a section during an event but would need to be replaced before it could be used again after powering down. If that is correct I don't think they will be a game changer for seasonal lighting out side of panels.
 
There are many different failure modes to a chip and im wondering how effective this would be for many of the failure modes. It will be interesting to see how this works and how effective it really is or is it just marketing and only works with one particular type of failure mode.
 
I recently approached Ben Brown about some Strip lights and he recommended this too me. I am going to use it on my sleigh so going to be ordering a few rolls of this very soon. I guess I will be able to do some testing and anything else once they arrive which could be a month away.


Ben may have more info on the types of failures, but the cost is only $7.00 per meter so that's not too bad I don't think especially if they work the way we all hope they do.
 
Well I have ordered and received my 8 rolls of WS2813 5v strip lights, and so far I can't get them to work on my Sandevice e682.


Tried a few different protocols on the controller and managed to light the string up full white maximum brightness lol blew the fuse on my e682 after I pushed 7 amps through the 5amp fuse whoops, but still not able to get the strip to light up the way I want it to.


Tested the WS2811 5v bullet lights that I got at the same time and they are all working on the controller so no issues there from the controller side, all I can think is something is different with these WS2813 lights which the sandevice isn't able to control.


Tested 2 different roles neither work, tried connecting data to the DI pin nothing then tried the BI pin still nothing, at a loss now to work out what's going on, anyone got a suggestion? anyone want to try a roll on another controller and see what happens (Only SA people for obvious reasons)
 
You're connecting the data out from the E682 to both the DI and BI leads on the first pixel in the string, right?


Dennis the Menace
 
Have you tried checking whether they will accept the data out from WS2811 or WS2812B pixels? Try putting these between the controller and the WS2813B.
 
dennismc said:
You're connecting the data out from the E682 to both the DI and BI leads on the first pixel in the string, right?


Dennis the Menace
Click to expand...


Hey Dennis,


I was advised by Ben Brown not to connect both DI and BI to the data out on the control board. So I have only been connecting one or the other in my test. I can try connecting both to the output and see what happens, what's the worst thing I can release the magic smoke from the first pixel I can just cut that off and solder to the next one anyway lol.


i13 said:
Have you tried checking whether they will accept the data out from WS2811 or WS2812B pixels? Try putting these between the controller and the WS2813B.
Click to expand...


I will try that but I feel that won't work either as the WS2811's I have plugged into the same port on the control board and they work fine, but yes if the data coming from the control board is a little wrong this might help to clean it up and pass it on.


Will test both tomorrow and advise if anything worked.
 
Reading the specs I would think you have to have both DI and BI hooked up, otherwise it will see different data on each and switch to the BI because of the differing signal. This obviously would give you no output. Anyway just my take on the specs.

Regards,
Rowan
 
Yeah I was thinking the same thing that both would need to be connected so it can compare the signal coming in and work out which is correct.


I guess I am the guinea pig for these new lights so a testing I will go lol.
 
You must log in or register to reply here.
Back
Top