ws2811 what type/voltage are these

hello guys i am hoping someone can tell me what voltage these pixels are.

i am new to all of this and i purchased a secondhand setup from a guy at the start of the month, he had a fire in a matrix screen so i guess he was done with the hobby.
i got all of the pixels he used and i showed some to paul from lightitupleds and he said they looked like 5v pixels because they had no resistors or regulators on them and that could explain the fire.
i looked at the examples from some of the threads here that pointed to wiredwatts and indeed these pixels look like the 5v example but when i power a 10 node test string with 5v they barely work.
single colors red green blue will work but if i flick it to white i get nothing, if i up the voltage to 12v they work perfectly.

i have attached a picture of the 2 types i have.
the ones with the green pcb were in the matrix fire, i have built the merry christmas letters with these and they are running on my show at 12V with the occasional led having a faint red glow when powered but no data is being sent but they operate perfectly.

the ones with the blue pcb are in all of the props i purchased and they have been running on my show for 3 weeks at 12v.
both types have no resistors or regulators.

i purchased 4000 12v ray wu pixels off paul and they all have resistors and are also in my show in a matrix and icicles running at 12v along with everything else.

i am confused, the guy i purchased the kit from said everything is 12v but the pixels look like 5v, i even put the string of 10 on my shed floor and left them on solid at 100% white at 12v for about 6 hours and they were fine.

any insight would be greatly appreciated.

I think you have got some GC8208 12v pixels.

They take WS2811 type data, but the chip is completely different to run off 12v. Notice the photo in that post had no regulators or resistors (on the back) inside of it.

I'd be inclined to say they are 5V WS2811.
Contra to Mark above - they can't be GS8208 as those are a 4 wire pixel, with Backup Data as well as Primary Data, and these are 3 wire pixels.
There is a single 33R resistor on the data out line, which is relatively standard, and a decoupling capacitor.

I note that at the Sydney mini, we pushed some pixels pretty hard in terms of voltage and miswiring to see what could/would happen. 5V pixels would often hang, but not burn out, at 12V. Burning out only occurred at slightly higher voltages, and it was the actual LED that had its PN junction burn up on certain colours, the 2811 chip continued without hassle. 12V LEDs could suffer up to 20V before they started to have physical damage.
In most cases the 2811 chip locked up prior, but then dropping the voltage back down showed the pixel to continue normally.
So - it's entirely possible that running those guys at 12V despite being 5V might actually be OK, as long as you dont rise the voltage much higher.
I will 100% say though- - this would NOT be recommended in any way shape or form.

A simple test would be to run them from 5V. If they continue working, they are 5V 2811's. If they are series-connected 12V's like the 8208, they will stop working as there wont be enough voltage to reach the forward voltage drop of the individual colours.

thanks for the replies.
i did test a set of 100 of them at both 12v and 5v and when i use 5v they will not go white which i am pretty sure lights all 3 led diodes(R G B) to produce white and i figure this color mode would be the highest load and the 5v just does not have enough to power them.

single colors on 5v work fine.

when i put the same string on to 12v the white mode works.
10 pixels or 100 it does the same thing, on 5v no white and on 12v they work as expected.

i think i might get a 5v meanwell psu and attempt to power the whole merry christmas prop(650 pixels) from it and see what happens.

the guy i purchased the kit from almost burnt his house down with these pixels so i have a fire extinguisher next to the prop lol

Interesting. Sounds like they are 12V then.
Are you able to read ay print written on the actually black chip at all?

i will try and cut one open to read it.

i just did another test with 1 pixel and the same 5v no white and then i tried that one shady pixel then a ray wu 12v(resistor type) connected to it and the ray wu goes white but the shady one has nothing on 5v.
i plug in 12v and both are white.

on the green pcb pixels the ic has TX5803A CD2272
on the blue pcb pixels the ic has Power Mos P9828M 2049

i think the blue pcb ones are 12v as i found this page for the P9828 which lists it as
"LED lighting constant current step-down constant voltage P9828 12V SMD single-point single-control driver integrated circuit"

maybe this is another way they are smashing them out of the factory? cheaper?

i cant find anything on the green pcb ones.

EDIT: i will add this here just for reference...
the ray wu pixels ic has WORLDSEMI WS2811 20220605

The datasheet for the P9828 provides a hint I think. While the link above goes to a page in a language I can't read, there is a schematic of I presume the IC's suggested connections. The LEDs are stacked/series-ed instead of a parallel arrangement as is the case for a WS2811. So that explains why white does not work on 5 Volt. The blue PCB pixels if the datasheet is reasonably accurate likely operate similar to a GS8208; 12 Volt and without a backup data input.

No success looking for a TX5803 datasheet.

I wonder if it is gamma corrected like GS8208. I'm as interested as anyone can be in chips that work that way, or as interested as anyone can be that already has too many of the original style of 12V pixels.

Hard to say. Possibly buried in the link's Mandarin/Chinese text that I don't have the energy to translate. I've never yet been too concerned about gamma correction anyway as I doubt that for what I put out anyone would notice or care about off-colors. My perspective with the gamma correction ICs has always been their primary purpose is for use in commercial advertising displays where regardless of illumination levels accurate or near accurate color rendition is desirable. For visual product photographic accuracy of course.

My interest lies with power efficiency; ie: GS8208 and WS2815. 12 Volt and regardless of color less than 20 mA.

I have orange, gold, purple, etc., in my sequences and some images so I care about gamma. If you set it for WS-2811, it will also save a little tiny bit of electricity because it does substitute darker colors. Better colors + less juice = win.

i usued to have 12v 2811 pixels and 5v that look the same infact i conected 5v nodes to 12v psu without realizing and the rest was history so i decided from that day to only use/ buy 5v nodes plus there cheaper and all my 12v is strips but next year changing strips to all to 5v

Does anyone have an update of these pixels? I had a conversation with a guy that experienced a pixel fire and was trying to dig up an actual datasheet or website for the P9828M controller.

steves said:

Does anyone have an update of these pixels? I had a conversation with a guy that experienced a pixel fire and was trying to dig up an actual datasheet or website for the P9828M controller.

Click to expand...

yeah i ended up having a fire with them, it can be seen in the Y of merry in this video


View: https://www.youtube.com/watch?v=Qf73AUz8Zqo


these were the different types i had


the guy i purchased them from had a fire with the same pixels which caused him to sell all of his show.

i could not find decent data sheets on them and i ended up throwing them all in the bin.

stooged said:

yeah i ended up having a fire with them, it can be seen in the Y of merry in this video

the guy i purchased them from had a fire with the same pixels which caused him to sell all of his show.

i could not find decent data sheets on them and i ended up throwing them all in the bin.

Click to expand...

Thank you for the feedback.

I am having difficulty finding a datasheet as well. I have been chatting with a fellow that has some of the P9828M pixels. He experienced a pixel fire and I am attempting to help him determine the failure mode. On a related note, I spoke to @merryoncherry recently about some of his GS8208's that failed. The GS8208 drives the LED's in series similar to the P9828M. The others he caught before combustion. One failure was catastrophic and resulted in combustion. He noticed that during the progession the failing pixel would begin with one colour constanly on dimly. As the failure progressed the light intensity increased and eventually the pixel began to heat and crack at which point he removed them. Do you recall or can you see in your video if you have a similar progression?