You would think pushing data for 6 meters through an led strip wouldnt need a logic level shifter right? For some reason it appears i do it locks up and does funny stuff when i adjust brightness after wled does its measurements i guess. Or i use certain effects. It runs fine on the original controler. Reason is i think its limiting the power to work via brightness level limits wled can disregard. Its probly data. I just dont know why. Its got power at the beginning and end of the 6 meters its 12v what gives. Its not a long run.
I wanted to add im using a 12v 3 amp power supply. Same size wires that come with the strips connecters. I assumed if they were the same thats good. Its also extremely close to the first led like 6 inches max. So help please lol.
So, to correct an initial misconception - youâre not âpushing data for 6 metersâ You only have to âpushâ your data the distance from your ESP board to the 1st LED (I think 20cm?) and then from the 1st LED to the 2nd and then the 2nd-> 3rd, etc.
The reason you donât have to worry about adding the whole distance (for the data stream) is because every LED in the strip regenerates the data stream before it transmits to the next LED. As far as data is concerned you have a new âdata boardâ at the output of every LED. Normally the distance between successive LEDs is <15cm, so itâs not an issue.
So if you think it canât be a data issue, that leaves power.
One way to test is try a small set of LEDs at the middle of the strip (you want the test as far from the injection points as you can get). Set up a segment of 20 LEDs across the middle and set it to white at 20% brightness. Try running some effects, like Wipe and/or Sweep on that segment.
Watch and see what happens as you turn up the master brightness.
If it works OK, start adding more LEDs to both ends. You should hit a number of LEDs and/or brightness level where it misbehaves.
If you can, measure the strip voltage in the middle when it starts to cause problems.
BTW, you donât say how many LEDs you have in total over 6m?
Note, all of this depends on that 1st LED getting clean data, whether it lights up or not.
Without clean data, nothing is guaranteed, thatâs why we use levelshifters - even when it âshould work over such a short distanceâ
1 Like
If its leds its 108 or im sorry ic count. If that helps. It could be data. To bad its made into the final product. I would need to break it . uhg i checked as i went i got lights each time. I needed to stress it each time with a effect i guess to see if it was all ok. Now some work some dont then suddenly it freezes and nothing works just frozen flicker. All when i adjust brightness or the effect draws more power. So ill try mt meter see if i find the issue. Ty . Do you beive trying a wled board like quins dig uno would be a good way to troubleshoot data ? It has level shifters for sure so in theory? It should be better if its data right?
Certainly worth a try with a levelshifter.
The 74AHCTxxx devices are really very simple to use, just make sure youâve got a 100nF (0.1uF) ceramic bypass capacitor near pin 14 and to ground.
Everything else is just wiring, if you want the simplest diagram possible: Ground all the inputs (4x~OE, 3xA) youâre not using, Ground the GND pin 7 (duh), connect Vcc pin 14 to +5V and add the bypass cap from 14 to ground.
GPIO data goes to the A input youâre using, corresponding Y output goes to the strip.
Ty when understand all that im sure ill get it lol. I had a thought after your information. I have the 3 wires between each strip in the visualizer over 6 inches long and connected with butt connecters. Im thinking maybe the length between each strip. That and possible the connecters causing me this trouble?
That sounds like a power issue. I would test like divsys said with a small segment in the middle and see if they work correctly. If they do it further confirms a power issue.
1 Like
So im stuck thinking its the length between strip sections i have 12 pieces at 18 inches each . Thing is i chained them all. Between each individual strip is about 6 inches to maybe even a bit more of wire . Could this be causing issues are the led at somepoint to far from the next to transfer the data ? Thinking so. So now what a logic shifter is great between the first led . I understand that but how do you boost this stuff when you got multiple small strips? A data booster between all 12 connected segments?
Fixed i guess. I had a bad connector. I injected power then the connector went bad so it still appeared to have some power. It was from the end . So replaced led that had the bad conection and reinjected power. All works. As intended. Power from both ends and not just thevend yay. Lol you guys probly think what a newb lol.
Hey, in the end you learn to be methodical about these problems.
I had one of my long term (6 year) installs start giving me all sorts of bizarre display issues.
Since itâs outdoors, I naturally thought Iâd be tracking a bad connection in the rain.
Turns out I had a basic power distribution issue back at the PS end.
Much easier to fix, but back in the day when I was experimenting with the original layout, I obviously had a âthatâs good enoughâ moment w/ the power lines.
I was as wrong as everyone else trying to make that assumption.
IME almost all of these power issues are self-induced and avoidable.
Glad you got it working and have fun! 
1 Like
Ty guys for the information and guidence. Im learning to fish guys.
1 Like
Awesome! Glad you got it fixed.
I had one of those pull your hair out moments also when I setup my mega tree this year. I took a strip down 3x and was convinced each time it was one of the solder seal connectors that went from the LED string to an xConnect pigtail. It turned out to be a bad pigtail. Sometimes it would work and sometimes it would not. TG I took the time to diagnose the problem and didnât just go âoh must be a bad LED at the end of the previous string or beg of that stringâ and start cutting out LEDs.
1 Like
There is almost no limit to the number of LED strips that you can put in series.
Limit 1: Your controller needs 3-bytes of RAM per LED.
Limit 2: Your power to the LED strips needs to be sufficient to keep the voltage within 5-10%.
Limit 3: 65,536 LEDs, which is the limit of the addresses in the protocol.
If the first LED in the string responds to data properly, then the rest of the data path is good because each LED chip regenerates the data for the next chip.
There is if you want effects to work correctly and with a decent FPS. If using solid color only then yea maybe the # does not matter (though it may take a while for the whole strip to change to the same solid color). But who is buying addressable strips to use only solid color?
According to the WS2811 datasheet, the transmission delay time, Din to Dout is 300ns. So, it would take almost 2 seconds to propagate a data signal to the LED # 65534.
Put 4,000 pixels on a port and tell me how you make out⌠
I donât care if itâs an esp based controller or a fancy kulp, falcon, genius etc. The fx are going to have trash FPS. If it worked people would not need multi port controllers lol. Will your lights light up and do things? Maybe. Will they be in sync? Not so much.
I commend you for trying to help the OP, but advising someone that itâs ok to do something that literally no one does and that no one would recommend is just going to confuse them more then they already are.
Here is more info on the subject: Max amount of Addressable LEDs? - quinled.info
Think you lost a decimal place or two.
Based on that math (propagation delay) 300nS =>300x10e-9s/pixel x 65536 pixels = 19,660,880 x 10e-9s
= .01966088s
That doesnât mean you can actually update that fast as there are other timing factors involved, but transmission delay through a pixel is not the limiting factor here.
BTW, I agree on all the other comments about real world experience with ESP32âs.
65K LEDs on one ESP32 is ambitious at bestâŚ
1 Like
I was referring to the theoretical limits. Some displays can tolerate a slow FPS. The link you point to is specific to that controller. Overall, RAM will be limiting factor for most of us.