💡QuinLED💡 Ever had flickering ws2812b LEDs? Watch this video!
HTML-код
- Опубликовано: 20 окт 2024
- Have you ever had flickering or other issues with your ws2812b LEDs running over longer cables? Make sure to watch this video!
#QuinLED #ws2812b #dataissues
📝 Discord 📝
If you still have any questions, let me know in the comments or join the Intermittent Technology Discord server here: / discord
░▒▓[Shopping links]▓▒░
❕ All below links can be affiliate links which provide me with a small kickback in case you purchase something. If you use them, thank you very much! :D ❕
👉 100pcs 0603 33R 1% resistors: s.click.aliexp...
✨ QuinLED-Data-Booster(-Maxi): quinled.info/q...
✨ QuinLED-Dig-Uno Pre-Assembled: quinled.info/p...
✨ QuinLED-Dig-Quad Pre-Assembled: quinled.info/p...
✨ LED strip buying guide: quinled.info/2...
░▒▓[Associated blog post & Mentioned in the video]▓▒░
📚 Written version of this video: quinled.info/d...
📺 QuinLED-Dig-Uno and WLED tutorial: • WLED Tutorial - The ul...
📺 + 📚 WLED v0.12 QuinLED Update guide: quinled.info/2...
📺 WLED and Home Assistant: • Home Assistant: Contro...
░▒▓[Music]▓▒░
Intro&Outro
TheFatRat - Xenogenesis
( • TheFatRat - Xenogenesi... )
My God, you are a hero. 4 years of multiple ways of troubleshooting, and the fix is found because of a ground wire run with the data wire. I am extremely happy and somewhat upset because of how simple the fix was and would have been in the past. I did upgrade my lights this year finding a work around, however I ran the lights the way I wanted thanks to you and the info you provided. I happened to run across this video yesterday morning right before I plan on putting the lights up on the j channel I installed on my house. I tested it on the ground where I ran a f amp with nothing 4 days ago. Seperated the data line and wala. Installed on the house and forgot the f amp. Decided to test, and it's a miracle. It is so simple, yet for some reason, I never would have run a test to find the answer on my own because I would have never imagined that being the issue. HUGE THANK YOU!!!!
Awesome to hear it helped you out!
I have power and ground in one cable and 3 data together in another cable. The cables are ran about 5 meters and I'm having flickering. Would just adding the 249 resistor to each data line fix the issue, or is there also an interference issue with them being in the same cable. Thank you for the video, very informative..
@justintucker2172 no, running multiple signals next to each other is the problem there. Best way currently known would be to run 3x 2-wire cable with a data+GND on there in 33R resistor mode. That will very likely fix it!
@IntermitTech I was under the impression that the ground was causing drain to the data, but you're saying I need a ground with data in 3 separate wires with the 33r resistor, correct? Thank for the quick response, had a little bit of a what went wrong moment when I finished the wiring the other night..
Yes, this type of data will have lots of crosstalk in that sense. So running multiple together over long distances is a bad idea. Running 3x 2-wire cable with GND on one of the pairs of each cable will provide a bit of shielding and probably make it all work. This does require the 33R resistor though.
Thank you for your products. Appreciate the thought that went into the dig-quad and the videos and articles explaining how to use it and the concepts behind the board. Getting ready to start a permanent install of about 900 LEDs via 3 home runs. Was testing today and ran 5m ws2815 over outdoor 100' 3 conductor 16 AWG copper cable. (Soldered the backup data and ground wire together.) 249 gave some flicker. Was going to cut the wire at 3 foot increments to find the limits. However, found your guidance on the 33 ohm resistor. To my surprise it worked straight away over 100'. Super pumped about my project and showing off the lights. Thank you so much!
I started just with WLED and was frustrated that the LED behaved bad. Eventually I saw some article saying put a 240Ohm in DL; had none, but used 220Ohm. That worked. But you explained why. Thanks for that. Now I need to find my little digital oscilloscope and try to record the data myself.
Anyway: subscribed …
Thanks for the clear explanation. You're right, it is a topic that's not talked about much, probably because it's considered "too technical". I'm glad you hold us in higher regard 😁
😊So I had a problem with flashing lights, I thought I was going crazy because I couldn't solve the problem, I run an xlights show with two Dig Octa cards, I switched the switches to 33 and the problem is gone. If I had known where you live I would have brought you a smoked salmon from Urk, I can now sleep peacefully again. I did this after seeing this tutorial. I have everything with 3 core cable. Thank you for the explanation.
I know this is an old video but thank you, I’ve watched so many tutorials where they say “You need to add this resistor” but they’ve never explained WHY you need to add it. I always want to learn more but I couldn’t until now.
Thank you Sr.
Man was loosing my mind over this.
I had the resistors.
Hooked up BAM FLICKERING GONE.
Again thank you
I came here after looking over my new dig-octa. Makes sense. I’m using it to run my Christmas lights and looks like I’ll be switching to 33ohm. Thanks for your knowledge.
Thank you so much for this advice, i finally fixed my ws2811 leds that have been flickering forever. Thought it was the power supply but nope i was using a 220ohm resistor. Changed it to a 20ohm and its working 100% fine now.
Awesome, that's great to hear, exactly why I made this video and article! :D
tq for the video . finally found the answer for my led problem . but now i have a new problem . my neck turn to left and right automatically .😅
The distance between my D1 mini and my LED strip is roughly 20m. I am using a sacrificial pixel and also added a 249 Ohms resistor but I still had this flickering and could not control the strip (465 pixels, 12V). I ordered several QuinLED-Data-Booster boards. They arrived yesterday and I installed one and it did solve my issue!!!
Excellent and glad to hear!
Perfect timing for this video. I just connected my WLED digquad tonight and the colors and animations were all over the place. I have about 25 ft or 7.6 meters of 3 wire cable for the hook up. I ordered 3 of the data boosters and when they arrive I will post an update. Thank you for your channel and all the great information.
I think that will perfectly resolve it, let us know! :)
Update: Fixed! By adding the booster and set to 33ohms it corrected the flicker and random colors.
One note to add. When I first hooked it up it didn’t not work. I took it apart and verified my connections but still not working. Well… to my embarrassment I forgot about the slider switch to select from the 249 to 33 ohm. Once I selected the 33 everything worked perfect. Maybe if someone reads this they will not make the same mistake.
Once again thank you for your channel and all the great technical information!
Thanks a bunch for your video and explanation of this. I had this issue and every other video suggested to try another power supply. 33Ohm resistor fixed my problem straight away. Greeting from Australia!
Awesome, glad it helped!
Thanks, I'm having this exact issue with a swimming pool light run, which is between 10-15 meters before the first LED. I'll give the board a shot!
Wow thank you Quindor!!! This fixed my problem I’ve been struggling with. Night and Day difference once I changed the resistance!!!
Love watching your videos you explain the topic perfectly, thank you!
Great video, very informative and a great product promo, as I will be buying a couple of your digi quad boards.
Good video. Something to think about for my future projects.
I'm thinking that a tiny, adjustable trim pot would be a nice solution. That would allow for better impedance matching with a variety of wires, and, at the same time, be easy enough to test with someone simply seeing where things were most stable on their string. Cheap, easy, and flexible, and, avoids resoldering different values, or fixed values that may be compermises.
For me it was so frustrating to get all 4 of my data outputs to stop flickering that I ended up simply using a sacrificial LED for each line lol. End result is ama-zing if I say so myself
Getting into signal integrity :)
I think your video will save me for tomorrow, i just put 4 120 Leds strips separated betwrenn them for like 4-6 meters, so what i will do is after the end of each strip i will add the 249 Ohm resistor to check if the signal goes good to the second and thris strip! ill let you know!
Interesting video! I've had success with WS2811 and 12b using a single electrolytic capacitor in the data line, ~100-200 microfarads. This is what I use for a 1300 LED bulb array (similar to those xmas lights) with data running adjacent to pwr+gnd, as well as for a 91 LED strip series (+a bunch of parallels, 1091 in total). The cable length to the first pixel is ~2m max for both.
Only the distance to the first led should matter. The signal is buffered from led to the next.
Would this degrade over time? I have that second scenario ~8-10 m of 3 strand wire. I set it up on the 249 ohm and it worked perfectly initially. However, after running for about a week, i now get major flickering and strobe effects. It seems to be exactly what you're describing, but I dont know why it would work perfectly initially and degrade over time. Any explanation? Or should I look for a different solution?
I have two jumps on my roof, I would fully be able to place the controller in the attic and start in the middle, eliminating the biggest jump and making it work. I am hopeful to make both jumps with one continuous run. However, I am having data drop. I am running in a 3 wire cable with 2811 and 248 pixels. When I get home, I am going to run a dedicated data line and cross my fingers. Good luck, everyone. Thanks for the video. I do have a f amp I put in the line not helping a thing.
Ты крутой Чувак , Я рад что смотрю твои видео.
Good explanation, with the ground running next to the data signal, its acting as a capacitor, so you have an RC circuit. That ramp is the charging of the RC circuit.
Not just the ground, but _all_ of the other conductors in the cable. The more conductors in a cable, the more its capacitance per foot changes. (think of larger plates in a capacitor) The rounding off of the waveform corners is an example of _integration._ The small peaks on the corners of the signal are _differentiation._ Picking the value of the resistor is actually easy. Its value should be as close as possible to the cable impedance. What's _not_ easy for the typical user is determining the impedance of the cable.
Excellent information, thank you!
Unfortunately, this video is 4 to 5 years too late. Had a nice project and came to this problem with an esp8266 with a mega it worked off unfortunately my wife's swaying came in between and the project was never finished, and my friend has now decided on something else. But thank you for letting me know where to look for future projects.
amazing explanation! Thank you
My experience is that putting a resistor at the start of the led strip, so at the end of the data wire fixes everything. So no resistor on the side of the board. I made an infinity mirror kitchen tabel light with 2 power wires and 6 data wires in one shrink sleave, so all on top of each other. Running 2040 leds and the best resistor value was between 0.4K and 1K for a crystal clear data signal. Verified on the scope. I ended up using 1K. So the value of the resistor might not even matter that much if you put it on the end of the signal wire. (See my YT for that led project)
Yeah this makes the most sense.
Oh, interesting stuff! But everything I see on your channel only has very short data wires, I wonder if the same is true if we're talking 5m or 10m wires which are the specific thing we are taking about here. For much shorter wires it's all a lot less crucial and lots of values can work. Also not having a resistor on the output but only at the end, I wonder if that influences EMI at all, although I do get that it would at least dampen ringing, for that purpose it doesn't matter where the resistor is. I'll play around with it a little bit when I have a bit of time!
@@IntermitTech It may look very short, but those dodecahedra use arround 5 meters of signal wire (28awg 6 core sillicone wire) And i also tested this before making it to size and that was 10m. The signal is boosted right after the esp with a 74HCT245.
Ik ben benieuwd naar het experiment met die kleine boosters. In theorie als je leds aan staan zou het prima moeten werken. Maar als er weinig effecten actief zijn kan de ene een bron zijn voor de andere zelfs bij een klein voltage verschil. Overal 10A diodes tussen doen is ook niet echt de bedoeling… Die boostertjes doen vast niet aan load balancing….
If you drive DIN without a resistor and WITHOUT the main supply, you will burn at least the first WS2812B LED.
Having a 230+ ohms resistor will protect somehow the input by limiting the current. That's a good reason in itself. It should also protect a little against electrostatic discharges.
Originally I was searching for best practice regarding ethernet and router configuration, but now I wonder if (in case one has an oscilloscope on site) the best practice for varying cable lengths simply a poti is.
Great video and thanks for sharing!
would the booster be better nearest the strip or halfway? what about the resistor? i do have a very long data line bundled with others for half of the way but it is dedicated. im thinking a 'differential' signal might be useful, if you have a video or guide somewher eon that? or a chip i could buy? or would it be simpler to just have another controller mounted nearer that strip?
Thats what am looking for! Great Video! is that 33 ohm for one data wire... what if I want to use 3 data from one ESP32 board to three WS2812B strips. does that mean 33 ohm resister on each data output? 4 cores cables (data1, data2, data3, Ground)
I enjoyed watching your channels :-)
what resistor should I use if the strip starts almost right at the controller?
Does a sacrificial LED not work for 10m wires?
Alternatively maybe a 110R resistor at the LED end, without the drive resistor, might balance the line better?
Sure but since the LED strips are bought as general items it's better to optimize for the real world situations where we can't control all the variables in that sense.
I’m at a loss even after this video. My cables are all separated, but I tested with a 33ohm, 67ohm, and 270ohm resistor. 67 seems the most stable, 270 made the LEDs difficult to see (as though they barely worked anymore) and 33 made them flicker ALOT.
My guess is that I need to find a sweet spot resistor for my setup, just despise all the testing that’ll have to be done.
Any ideas?
Yeah it very much depends on the setup, from all my testing 249 and 33 was best for specific situations and made a lot work, but it's possible your setup reacts a bit different.
Okay, but WHY does the resistor help? You explained the scenarios and electrical values and what happens on a scope, but not why a resistor (and not a capacitor, inductor, etc) is the right fix here.
Ah right, that wasn't really the intent of the video. In this case the resistor better matches the impedance of the cable and for scenarios where GND is close it allows a much higher and quicker pulse of energy over coming the effect of having ground so close to it for the long distance. It's just a different scenario to tune for basically. The resistor also provides in a bit op dampening for the single ended signal to dampen down reflections and such issues otherwise.
@@IntermitTech Ahh impedance matching! I know just enough electrical engineering to be dangerous. Thanks for the explanation!
Huh. I never considered the transmission line issue, as my data lines are all under 1m. Makes sense though. I do have one set of addressable LEDs around the base of my bar (in Al channel) that get triggered every time my bar fridge turns off though. Got a solution for this? I've tried separating the LED and fridge power supplies with a high quality mains filter (this prevents the fridge glitching my HDMI over Cat5 but not the LED issue). I've tried shielded data and power cable. I've tried ferrite chokes on the data and power cables. Nothing prevents the first 6 to 8 LEDs randomly switching on when the fridge switches off. At the moment I have a Home Assistant automation that checks if the bar is occupied and if not switches the LEDs on briefly then off every time the power from the fridge compressor goes below a set value.
Oh, that's an interesting issue. You could maybe add a fairly low value pull-down resistor behind the level-shifter to the data line, that would try to make sure it's grounded. Not sure how low you can go really before it would influence the signal too much, you could try 1k and see if that does something.
Other then that, actually using triple wire so GND close to data and then using the appropriate resistor to compensate for that can actually add some noise immunity because of the transmission line effect. So that might be worth a try too.
If it's actually triggering the LEDs "directly" so not because of garbage on the feeding data wire, poh, that's going to be tough, are the LEDs in an Alu profile? You could maybe do that and ground it since the LEDs will be insulated from it anyway, but it could maybe act like a shield a little bit.
Hope any of it helps! :)
Great Video, I'm doing a project, 17 stairs, 59 led per stairs, installed in a diffuser, (60led/meters WS2812B). I have 2 types of Boards I can use (ESP32 CH340C, ESP8266). I've tested with both boards and it works A1 up to the 14th stairs and then the 3 top stairs either blink or stay lit, I wonder if adding a resistor would fix my issue or if I've reached the limit of my 2 boards. I've injected voltage but I can get the 3 last steps to work properly. I'd really like some help.
Poh very hard to diagnose like this! Maybe join the Discord sever? Easier to dicuss things.
The shipping for those booster boards is $32 canadian from the worldwide store. Almost 10x the cost of the board itself. Great information in this video for sure, but I can't pay that ridiculous shipping charge.
Oh, euh, that doesn't sound right. I do expect most people to add them to an order when ordering some other Dig boards, but I'll look into it for sure!
Yea, I'm in the same position, I already paid $32 shipping for the Dig boards, only to find out later I need Boosters. Will try the solder route but not sure my resistors are 1%...
The % of the resistor won't matter too much. In that regard, the new version being released this week has a new resistor switcher feature built in to exactly prevent these issue from causing problem for people! I try to improve where I can, sorry you ran into this!
Thank you Quinn, the resistor swap worked great, and my tolerance seems right at 1% as the Manufacturer claimed. BTW, I thought you should know that my resistance problem occurred going from 1m to 2m strands of entry-level 2812b 144 L/m (from the Asia region) with data bundled next to ground, so it's definitely not just a long strand/string problem.
Oh, that is interesting indeed! But very happy to hear it worked out. As said, boards coming out now have new switches built in solving this issue once and for all, just wish I could have done it sooner, but it required a bit of a redesign of some parts and lots of validation and testing, that always takes longer then you want it to.
Great info, I do have a question regarding my setup. and hopefully i can get someone to help me out. What i use: DC5V10A power supply, 5V BTF-LIGHTING WS2812B RGB 5050SMD, ESP WROOM 32 ESP32 Development Board 2.4GHz WiFi Dual Cores. I have them installed and they were working ok but i saw power hungry by the end of the strip (this is installed on the top of a room) so i provided 5V at the end of the strip and that fixed the power issue. 3 days later, the controler fried. I changed it for a new one. It worked fine for about 10 mins then i got that crazines color changing and flikering. I already separated the data cable from the power supply wires and still the same. the controler and power supply is about 1.5 meter away from the 1st LED and the full LED run is about 50 feet and 900 pixels. What DO I Need to fix the flickering and random color changing?? TIA
That's a bit much to go into on a comment. If you are using multiple PSUs make sure to have a strong GND connection between them.
Data wise without a level-shifter and resistor it's more luck then science if it works at all. So that could be a problem.
900 LEDs on a single pin will also cause big framerate lowering, I recommend no more then 600.
Wire diameter for the injections are important too, especially with 5v stuff.
Not sure what is causing the ESP32 modules to fry though.
@@IntermitTech thanks so much… I am going to change the wiring for a smaller gauge and get a higher power supply and go from there.
UPDATE: I change the power supply (the previous one was rated 5V 10A but was pushing out 4.3V and about 3A) I installed the new one that it is a 5V 60A. this one is pushing out 5V and about 40A. I also changed the wiring that goes from PS to Strip for a smaller gage and worked perfect. I have my auto brightness set at 5000 in the settings and it is more than enough.
Keep up the great content!!
What is the appropriate wattage rating for a resistor on the data line?
Doesn't matter much, we're talking less then 1mA generally.
Great timing for this video. I run multiple d1 minis in my Christmas light display with the d1 within 1m of the start of the strips. Last night I added 200 pixels {in one prop - 4 x 50 pixel stars) to the display. When connected by themselves, they run perfect. When I hook up the other controllers (all running off the same power supply) the stars flicker. It's either a v- issue as the lights are not on the same v- as the controller OR its this issue. I have 3wire running 4m from the controller to the first pixel. Do you think is the missing resister causing this issue?
Hmm, if it doesn't happen when the others are off it's more likely a power issues (either supply or certain cables dropping too much voltage).
@@IntermitTech Thanks for your reply. It was a power issue, OR more specifically the grounds. I hooked up the D1 mini ground to the same one as the lights and the flickering stopped. Feel free to check out my display videos from last year.
The flicker problem came back even when confirming all the grounds etc. I put a 120 ohm resister in there (thats the smallest I have) and the flicker has stopped. Yay
I have a question regarding the total length data can travel on strings via a Dig Uno. I know power should be injected every 5 meters but at some point does the data signal become too week to drive the LEDs?
No in theory the data signal gets repeated by each LED so can run for longer lengths (not long gaps though, the LEDs need to be fairly close. But there are limits, see here: quinled.info/2021/03/23/max-amount-of-addressable-leds/
Busy writing up my BOM for whole house LEDs on the outside, love these boards and their pricing. Plan to use them at power injection points because hey, might as well clean up and boost the data signal at the same time.
Took me a while to realise though what it was doing. You are really working on the shift register being better at reading a weak signal compared to the LEDs and spitting the signal straight back out at full high/low levels again if I have got this right?
Yes, it is generally more sensitive then an LED itself but it's mainly meant so that the output cables be longer after the board then straight from an LED.
Shouldn’t the resistor be as close as possible to the strip? You’ve integrated it directly into the dig lol, I’m confused
If I wanted to replace the resistors on my quads and uno’s with a 33 ohm, do you have a link for preferred ones? Maybe something from Amazon.
Just generic, the size is listed in the article (0603 from head)
Hello sir, I hope you see this - I know the video was posted quite some time ago.
I have a btf 5v ws2812b strip powered by a btf 5v 60a power supply, with a esp8266 board flashed with WLED.
The ground, power, and data cables are ~2m (~6 to 8 ft) at most.
When powered, the color and brightness are perfect when untouched. But when I pick them up to try and install them, they randomly flicker different colors, blow up to max brightness, etc. When I place them back down and leave them alone, the correct color and brightness resumes.
The strip has 40 LEDs on it. Do you feel I need to add a data booster? I have no resistor of any kind along the 6-foot data run between the ESP board and the first LED
More info:
Vin and Gnd on ESP are running to power supply via 6-inch 18awg wires, and twisted together with 20awg wires running to the first LED for 6-8 feet.
The D4 pin is running about 6 feet uninterrupted to the data contact on the first LED via 20awg wire.
I'm missing the level-shifter in this story?
@@IntermitTech ahhh - I actually am not using a level shifter right now. I think you just solved my puzzle.
Thank you!!
Hey, Thanks for the great information. I will definitively buy a few of those boards. I was also looking to buy the QUINLED DIG QUAD esp32 with LAN but the worldwide side (I am from Canada) says it's out of stock.... : (. Any other way or place I could purchase one of those board? Thanks in advance.
If Worldwide is out of stock too we've really run out. A new batch is on its way, check back at the end of the week or begin of next please!
@@IntermitTech Managed to order 1 with 3 Data booster!! : ) Can't wait to get them and start working on my project. Thank you.
@@IntermitTech I ordered Two V3 DIG-Quad last week do I still need the data boosting board ?
Which Pre-assembled Quad Boards have the resistor installed on them and which don't have any? Is there batch/version numbers or a date you added the 249 resistors to the board design?
There are revision history pages for both boards on the website. ;)
quinled.info/addressable-digital-leds/
@@IntermitTech Thanks looks like I have a pair of version 1. I ran 2x twin pair 2.5mm2 wires around 7meters as +5v and 0v for my LED Strips and then branched them out in a parallel star config wiring through xmas pigtails. for the data line I used a 7 core .44mm2 for each core to each of my LED strips. The flickering seems to occur if I try and run 2 strips off 1 of the Quad's LED channel outputs regardless of whether I have a resistor of any value or no resistor. But solved the flickering by just having 1 stip per output and syncing the quads together in wled.
Interesting, I have done tests with parallel outputs from a single data port and generally 2 to maybe 3 worked ok. That was with short cables though, so maybe that changes thing a little bit.
Glad you where able to find a solution though! If you have a recent version, there is actually a 5th channel as the Q1R port (in the 5vEXT block) is also level-shifted to 5.12v!
I got a question... I have the latest WLED v12, installed on a NodeMCU, but some effects are sluggish... not smooth, what causes that?
Maybe you are running too many for an ESP8266 to handle? Read this: quinled.info/2021/03/23/max-amount-of-addressable-leds/
Interesting video, this only applies to long data cables leading to the first pixel? I currently have two strings of 2815 running off a new dig quad, that have flickering and corruption after roughly 200 leds, but my data line from the quad is maybe 1meter. Brightness didn't change it and power injection didn't help either.
Yeah, 1 meter should be fine regardless data wise, have you hooked up the backup data input to GND?
@@IntermitTech Thought i responded to this last night but apparently forgot to post it. Anyway, so your saying to connect the backup data line on the led strip to a 2nd common ground instead of a second data line? Currently I just have 12v+ Common Ground and 1 data line. Overall length is 255 leds, and I have injection from the end at #255.
Well to any GND, even just that of the strip itself, but yes
@@IntermitTech Thanks for taking the time to look at my problem. I'll give that a try. And Report Back. Thanks again.
I'd be lying to myself if I said I understood what's going on.
Does this also work as a level shift?
I have changed resistor to 33 Ohms, it got better. But still some effects is flickering and some don't. Why can it be? Do I need power injection or data booster?
How long is your data cable and what wiring are you using exactly? It's really best to troubleshoot on the Discord server
The length of the first data cable is about 10 m on the first channel. There are 106 LED's in it, powered on both sides. The length of the second data cable is about 5-6 m, there are 91 LED's on the second channel, the power is connected on one side. The effects start from the first channel and move on to the second one. The second channel works fine, but in the first channel half of the LED's, which are close to the end work randomly, but not in all effects. For example, Candy Cane, or Chase 2 works fine, but Bouncing balls does not. In addition, when the power is turned off on the WLED app, some light bulbs continue to light
@fevgg yeah that sounds like data issues. Are you running 3wire bundled cable?
Does it still work if you place the Resistor on the board? I was advised that the Resistor on the Data line must be physically placed close to the first LED or it's useless..?..
Also none of these online guides ever mentions what WATTAGE should the Resistor be?
Any wattage is OK -> as small as you want. To me the resistor is more useful for the protection it brings against electrostatic discharges (and over current into the led DIN if the led strip is not powered). Where to place R ? Indeed, by placing it close to the Led, the driver (cpu) will be able to inject more current into the line. But then it is the pin of the CPU that is also now unprotected against ESD. I won't talk about the characteristic impedance of the line, because that depend on your specific setup.
@lcdvasrm Thank you for your reply. I wish I understand it as well as you.
What if I have 10m+ of separate data wire going from a controller hub to the wall of my house, then another 10m+ of bundled wire running from the house to the led strips?
Put a Data-booster board at the transition point and switch it to 33R for the combined section, then it should in theory work fine. Especially with long lengths like that, putting a booster board somewhere in the middle isn't a bad idea anyway, since it sends the signal out freshly again at 5v, even if it sagged on the way there.
I have ordered these. When I get them do I install them close to the Dig Quad or closer to the led strip? Led strip is about 30 feet from DIg Quad?
Depends on what you use it for, if resistor change, close to the Dig-Quad
@@IntermitTech I think it would be for that as my led are 30 feet away from the DIG Quad and I'm using a 3 pair cable
Yup, that is the exact case, should instantly fix that.
Hello. My configuration has 15 meters led ws2815 I did not give a resistor everything works or do I have to provide a resistor? regards :)
Well the length of your LED strip doesn't matter, it's the length of the wire going to the first LED from your controller.
Great stuff. Yet again. Thanks for that. Question: when you refer to cable lengths (10m or 20m in this video), do you mean the length of the cable run from the controller to the first LED on the strip, or do you mean the entire length of the run from the controller (through the wire that get you to the strip mounted over there and then through, say, two or three 5m led strips-- another 5 or 15 meters)? More specifically, if I've got 10m of cable before three 5m led strips, do I have a 10m cable run, or a 25m cable run? (in lengths you refer to in the video)?
Followup: in the 10m cable + 3x 5m strips (15m of LEDs), do I do use your data booster once right outside of the diguno controller, or do I (also?) put one in the run itself? Like between the 1st and 2nd strips?
We're only talking to the first LED. Depending on what you use the booster for but if it's for resistor change, right behind the controller.
@@IntermitTech Excellent. For 3x 5m led runs then (900 LEDs) would an additional booster board be needed to get signal to the last few hundred LEDs? And if so, placed somewhere between the last 2 strips? Or is the booster literally for only right behind the controller?
No, as long as all the pixels are close togeher, if you are bridging a larger gap between pixels then mahbe. You might also want to keep the amount of pixels to about 600 per data channel to keep everything running ay 60FPS.
hey you gonna do your oled update vid soon?
I've been thinking about it yes, especially after the LTT video this week which was quite good! :D
Would a separate screened cable for data transfer work better?
Maybe in some scenarios but you'd still have to tune the resistor and the fact that such a signal isn't designed for these distances also plays remains, if you really want to do long distance, differential really is the way to go.
@@IntermitTech Hi will you be talking about differential signalling soon i have looked into it but seems you need twin twisted pair for each signal is that correct? would love to see a live stream on that subject.
I do have some plans for it yes and it's correct you need twisted pair for it. You can go quite the distance with it though!
Is it ok to disconnect the ground coming from the quad v3 to the leds and fix the flickering problem? Anyways i have the ground connected from the power injection in a separated cable, right now i have a 3wire cable connected to the lights and that’s why i have the data problem , so I just want to know if is ok to disconnect the ground from the quad?
Well the one running to the LEDs of its running over a different cable, yes. But the Dig boards does need GND to it, can't work without it.
@@IntermitTech i have two different elevations, two different set of 18/3 cables connected to the quad v3 and i have two power supplies at each end of the leds and power injection in every end or union of the strips , i have the dig uno , and when I connected the dig uno to one elevation and the quad to the other elevation works good, but I thought the quad was going to take care of me having two different controllers,
@@IntermitTech in this video you said by disconnecting the ground that goes together with the data cable to the leds will fix the problem, I did that and got worse, so I will try the separate data cable from the quad to the leds , but we have snow on the roof now, 😁 if that doesn’t work maybe try to put the 33ohm on the data line going to the leds? But the quad already have one on it ,
By default it is not configured for ground next to the data cable so it comes with the 249R and not the 33R. So if that saves you from going up on the roof, maybe try that first? Even just bridging the resistor spot could already do that as a try out.
Well sure it can, depending on total amount of LEDs and power draw, but still. Just need to make sure the resistors match the wiring setup then.
What about twisted pair wires?
Won't do much, wrong type of signal. quinled.info/why-ethernet-cable-cat5e-cat6-etc-isnt-suitable-for-leds/
What if a Balun is added?
Would this work with the WS2813?
Yes, same thing. :)
@@IntermitTech As the ws2813 are a 4 wire config what should I do with the "backup data cable"?
Generally the back wire is tied to GND, the first pixel wil generate the double data.
Very nice explanation, what about some jumper pads for the resistors so the user can bridge them with a little bit of solder?
Or maybe that's a bit too much towards the diy way.
Well I'm all for the DIY way, for sure! But if you have the soldering iron in hand, whipping them off the board is 2 seconds of work and then soldering on new ones a few seconds more. But that's why I've included the locations and pictures and all. :)
@@IntermitTech I know, what I mean is that the user then has a choice instead of fiddling with a resistor and have to order them.
I don't mind swapping them either, just a thought after watching the video.
I suspect the reason the added resistor to the data line or, for that matter, the decoupling capacitor across positive and negative lines are not talked about much or when they are talked about there is no consensus is because it's part hogwash and part voodoo superstition. The very same reason you'll be told you MUST use a resistor any time you use an LED yet billions, literally billions, of LEDs fairy lights are sold without a single resistor. Myself, I have never once used any resistor on my data line, not on any of my projects and have no evidence to suggest it has caused any problems. I generally don't use a decoupling capacitor either unless the distance from controller to LEDs is greater than four or five feet, though this is done more because of my own superstitions rather than any evidence it solves any issues. I am currently looking at one of my earlier projects that has neither a resistor or capacitor on the twelve foot span of lead wires and has been running error free for years. Years, and not a glitch. Your mileage may vary so take my voodoo advice at your own risk.