You are most welcome! I'm glad what I shared helped out. Thanks for watching and taking time to comment.

Do note that voltage drop along a lengthier run of LEDs can also result in fading, flickering or other misbehaving LEDs. The data signal is boosted by each pixel, so the shifter is generally used for the longer run between the controller and start of the LED strip. But if the voltage along the strip drops too low, you can see similar issues (although it is normally preceded by fading color/dimming LEDs first). If you've been doing LED installs for a while, I'm sure you are familiar with doing power injection. But just in case, I do cover power injection in my video on common LED questions ( ukposts.info/have/v-deo/f4qnaZCroWyal30.html ). Jump to the chapter on power injection. But including a logic level shifter is never a bad idea... I include one on all my LED controllers, regardless of length of the wiring to the start of the LEDs.

Never mind I am 6 months behind and I now see the video you already made on using the I2c chip 😂. Keep up the good work. Appreciate your videos.

You are welcome! I include it on all my LED projects... even if I could get by without it. I actually have another video on logic level shifters coming out tomorrow (May 6), addressing some of the feedback I received from this initial video. Thanks for watching and taking time to drop a comment. I greatly appreciate it!

You're welcome! Many will say that a shifter isn't necessary... and in certain cases you can get away without out it. But as I mentioned, for the $1 cost, I'd prefer to use one and avoid any potential signal problems like I demonstrated in the video. Thanks for watching and taking time to comment!

Good luck! And of course just let me know if you run into any problems or question along the way.

You are more than welcome. There just seemed to be a lot of comments and debates about the need for the logic level shifter. So, I thought I'd provide what information I had, why I opt to use one, and let folks decide for themselves whether they wanted to include it in their project or not.

You are very welcome! I'm glad you found it helpful. While not everyone agrees about the need for a logic level shifter, I still say it is a cheap insurance policy to assure a good signal for your LED installs.

Three feet is by no means a guarantee. It somewhat depends upon the gauge of wiring used, the tolerances of your controller and LED strips, etc. In some cases, you might get longer distances without issue.. and in others you might have problems with distances shorter than three feet. That's why I recommend, and always use, a logic level shifter on all my LED projects. Then I don't have to worry about having a problem with the signal voltage.. whether the distance is 6 inches or 6 feet. But like all DIY projects, the decision is yours whether to use one or not. Thanks for watching and the kind words!

Well, a lot of folks will contend it's not needed... and as I show, in some cases you can get away without it. But for the $1 cost, I always include one to assure a good data signal. YMMV. Others claim that this I2C shifter is "too slow". I have yet to see it, or any negative effects with this shifter. Maybe if I had an install with 1,000s of pixels or was looking for super smooth animation on a large matrix, then I suppose that speed/framerate might be impacted. I do plan on doing a comparison of this shifter to the much more difficult to use "recommended" shifter to see if I can detect any differences (at some point). But with well over 2 dozen LED installs (both WLED and using NeoPixel/FastLED), I not seen this simple, low cost shifter cause any problems. Thanks for watching and taking time to leave a comment!

@@ResinChemTech Hi. I've also heard that I2C is too slow and that that is what's causing data issues with my 480 pixels. I've just bought a 74AHCT125 shifter in the hope that this will work. I'm not doing fast changes to my LED's as it's for a scoreboard where I've made big digits out of 32 LEDs per digit, so they only change every few minutes, but after a while (and sometimes immediately) the data gets corrupted along the length of the strip. Any advice for a newbie like me is very welcome.

@ianwoodford9725 Yeah... I've had others say that about the I2C shifter... but I just haven't seen it. I also have a matrix that is a clock, scoreboard, etc (I have a couple of different videos on it) with 400 pixels (ws2812b). I've seen zero issues with the signal, regardless of the mode. I also have an install with over 800 LEDs (linear, not matrix) and again, no issues with the signal using the I2C shifter. I'm not saying others haven't experienced an issue, but I just haven't seen it. I've purchased some 74AHCT125 shifters and am planning on doing a video comparing it side-by-side to the I2C shifter. But please let me know, one way or the other, if switching shifters solves the problem you are seeing. If you are still having issues, then the signal problem is elsewhere and not with the shifter. But I'd love to know your results either way.

Let me know what you find! As I've mentioned to others, there may be no problems with one install but another nearly identical install could see issues. That's why I always include the shifter in all my builds... whether I could "technically" get by without it or not. It is a cheap $1 insurance policy to assure I have a good 5V signal leaving the controller and I don't have to be overly concerned with the length of wiring run between the controller and LEDs (well, at least up to a point). If I have an issue, then it is likely related to something else and not the data signal voltage. So, as I say to all the others that say a shifter isn't needed... go for it.. it's your build. But the first time you have a data signal problem, you'll wish you had included it... especially if you discover it after everything is installed. It's a lot harder to add the shifter (or even the sacrificial pixel) after everything is built and installed, as opposed to just including it up front. So, let me know if this does solve the issue for you. Thanks for watching, taking time to comment... and the sub! I appreciate it!

You're welcome. While you can sometimes get away without it, using the logic level shifter assures a strong signal and can often avoid any problems when using longer wire runs. Thanks for watching!

Glad the shifter solved your problems! The default 3.3V from the ESP is already right at the bare minimum signal for LED strips. Any further voltage drop (or slight interference, as you found) will lead to all sorts of strange behavior. People still want to state that it isn't necessary and they've done x number of installs with no problem. Fine... I still believe that taking a few extra minutes and spending $1 to assure a strong 5V signal to start is worth it. And it is a lot easier to install that shifter up front, than to wait for problems to occur and then have to try to add it later. Thanks for sharing your experience... and confirming that a logic level shifter can often solve issues related to the LED signal. I appreciate it!

Thanks! I know this video will probably generate a lot of different "opinions" as I've received many comments on my other LED videos telling me "you don't need that logic level shifter". So I thought I'd try to show 'real' stats and test results, why I always include the shifter, and let folks decide for themselves. Thanks for watching and taking time to comment.

I guess that was the point I was trying to make to all those that say you don't need it. Maybe you don't... but for me, if I'm going to spend +$50 on LEDs, power supply, etc. for a project that I'm going to permanently install in my house, I'll spend the extra $1 and a few minutes soldering to include the logic level shifter and know that I won't run into signal problems (well, at least due to voltage drop). Thanks for taking time to comment.

I used to add a 1000 µF capacitor at the power supply, but haven't added one in the last dozen or so projects.... and haven't had any issues, although it sure doesn't hurt to add one to help protect from power fluctuations. A resistor is sometimes needed on the data line if you have a "dirty" signal or interference. This mostly happens when you have a long run where the data signal and ground are in the same cable (e.g using a 3-cable cord to run both power and data lines together). This is also rare, but I have had to add one for a given project. Fuses, on the other hand, are probably a good idea... especially with larger install/high amp situations. I've been meaning to cover the use of an inline fuse in one of my videos... then I always forget! To be honest, the only place I've 'fused' my install is with my Christmas tree. All of my other installs are 5V runs with generally only a couple of hundred pixels max. And I rarely run those at full brightness/full pixels. I also always oversize my power supply so that it is not taxed even at max pixel load. I also continuously monitor any new installs for heat (power supply, controller, wiring and pixels) until I am comfortable with it. So, to answer your question.. are they "necessary"? Probably not. Are they a good idea? Probably yes on the fuse. The cap if you are interested in 'protecting' your controller/LEDs, especially if you have frequent power issues. The resistor would only be needed, in my opinion, in rare cases to solve a particular issue. But adding all three won't hurt anything and, again, would provide 'peace of mind'.

Great info. The 125 is how I’m going to go. Thanks.

Yes.. and to be fair, I was using very thin gauge wire (Dupont cables). Heavier gauge wire would result in less voltage drop and likely a longer run before problems occur. What I had hoped to get across was that while a shifter might not be required in every case (and might even vary in the same case, as you mentioned), for me I'd prefer to spend the $1 and not have to worry about voltage drop or signal issues. I learned this the hard way on one of my early projects... and it was a real PIA to have to tear everything apart to upgrade the controller and add the shifter. Now, I just include it for all my LED projects, whether it is truly "needed" or not. But thanks for sharing your experience without using the shifter.

@@ResinChemTech For digital signals very little current flows on the wire, so wire gauge will not matter unless the frequency is very high or the cable extremely long and that thin wire you used is fine. One thing that isn't ideal, and probably makes you more sensitive to low voltage is that (correct me if I'm wrong) you ran out the positive data cable to 10ft, but left the return signal routed through the (much shorter) power ground. You really want to keep those two lines close together and close to the same length, like with twisted pair cable. Either way a bigger problem is impedance matching. Since you aren't using a controlled impedance cable, you're generating lots of reflections at the ends and possibly along the length too. These attenuate your signal and also generate spurious edges which confuse the receiver at the far end. Boosting up your signal level makes you more resilient, especially for shorter cables a few feet long. In addition the switching speed on those level shifters is pretty slow, much slower than an ESP, so they won't generate spurious edges unless the cable is extremely long. I think you're really doing two things with the shifter (although I haven't used one so I could be wrong), giving yourself more signal to work with and stretching out your clock edges so far that they don't generate those spurious edges that confuse the receiver. You could go controlled impedance with twisted pair or coax cable and put a source resistor in series with the ESP to impedance match to the line, but that slower level translator is probably a more cost effective option and likely just as good as long as your lines aren't hundreds of feet long.

See this for a number of my standard wiring diagrams, including the shifter: resinchemtech.blogspot.com/2023/01/wiring-diagrams.html For future reference, _most_ of my project videos have related written guides that will have wiring diagrams, parts lists and code samples for those things shown video. Please check the video description where you will find links for this sort of additional information. The type/size of ESP board used really doesn't matter. All ESP boards, regardless of size or whether they are 8266 or 32, output 3.3V on their GPIO pins. This means that you still need the shifter to "boost" that to 5V for the LED strips. The wiring diagrams in the link above shows the shifter wiring for a few different boards... and even shows the wiring for a different type of shifter (see the Logic Level Shifter section). Hope that info helps.

Thanks for the tip and info. It is entirely possible that they changed something (likely to save a fraction of cent on manufacturing). I haven't actually ordered or used those shifters in a while, as this video is nearly two years old. Unfortunately between videos, blog articles, Github repos, etc., it is impossible for me to keep all older links up to date. More recently, I've been ordering these: amzn.to/3SZTGe3 I've ordered as recently as this past July and they still seem to be working fine. But since you have confirmed that there are issues with the Hiletgo, I will go ahead and update the link in this video description. Thanks again.

Well, if changing power supplies to your bench power supply fixes the problem, it sounds like the issue is power related and not related to the data signal/level shifter. You did not mention how many LEDs are in your strip nor the maximum wattage of your other power supply. As a _rough estimate_ of the current needed, take your total number of LEDs and multiple by 0.06A. Your power supply should be capable of supplying a minimum of this many amps. While it isn't required as frequently as with 5V LED strips, I'd also try power injection at the end of the LEDs. Run separate wires from the power supply to the +V/GND connections at the end of the strip or string. Otherwise, it will likely be just common troubleshooting steps... swapping out the strips, changing controllers, etc. But flickering or flashing LEDs are almost always the result of one of two issues.... the data signal or insufficient power/voltage drop. One key hint as to which one: If nearly all the LEDs are misbehaving, from start to end of the strip, it is likely a signal problem. If it is the LEDs towards the middle/end of the strip that are flashing/flickering, it is probably power/voltage drop. Hope that helps or gives some additional ideas as to where to look for the problem.

@@ResinChemTech Thank you again for intense analysis and response of my case :). Herewith some additional information. Brand of strip BTF, 24v, 5 meter, 720 leds/meter. Max power drawn on bench supply (visible) is 7.5 amps. The power supplies giving the issue are 24v 300w / 12.5amps max, the small, ultra thin ones with aluminum case. Same goes for that "laptop power supply with plug" type. I didn’t do the power injection at the end of those 2 strips. The behavior is for the whole strip, not from somewhere in the middle. In previous projects I tried the both ends powering to see if that helped. Same brand strip, split in 4 parts of 2 meter, all powered from both ends. All segments connected to each other for the signal. Also with the “ultra thin 300w supply” I couldn’t get them stable, while the bench supply did the trick easier (but you don't want that powering something permanently). Was finally able to solve this by having the supply extremely close by the esp8266 and the led strip, and wiring between strip and esp very short (max 10 centimeters). But this limits me to have 1 strip on 1 esp, because not able to run a signal cable for over more than +/- 1 meter (or even less). And having several strips with several esp's (several wled instances) makes it very hard control instead of via 1 wled instance. If you have any idea to try, much appreciated 😊, still trying to understand how to fix, to create more safe setup. Cannot imagine I’m the only one with those struggles, so hope others are also helped with this case and possible causes + how to solve. Thank you! John

The GPIO pins on most Arduino boards already put out the 5V required by the LEDs on their GPIO pins. ESP boards, however, only output 3.3V. Are you saying you ran a 20 meter cable using an ESP board without a shifter? If so, I guess you are pretty lucky to not experience any voltage drop over that distance that would result in the signal dropping below a usable voltage for the LEDs. As I said in my video (and repeatedly in other responses), the choice to use a logic level shifter is totally up to each individual. For me, when using an ESP board, I'll spend the extra $1 to assure a good 5V signal leaving the controller... no matter how long my wiring run might be. That's just my choice. But you are right that you don't need a shifter with most Arduino boards as the outbound signal is already at 5V, so the shifter wouldn't provide any additional value.

@@ResinChemTech of course I was talking about 20 meters in the context of a 3.3v signal from esp32

Thanks! In the background is my "hexagonal lights" (aka NanoLeaf) project. You can see a video of that project here: ukposts.info/have/v-deo/pH6Up4pse5CbooE.html Thanks for watching and taking time to leave a comment!

I haven't experienced anything like you are describing and it is a bit strange that just unplugging and plugging back in results in the LEDs lighting up. I know you said you've tried multiple shifters, but have you tried a different D1 Mini? Most of them are pretty reliable (depending on where you bought them), but I have definitely had a few bad ESP boards over the years. Maybe there is something strange or not quite right with the D1 Mini.

@@ResinChemTech Thanks for the reply. Yes, I tried two other Minis with the same results. These were 3 of a batch of 5 I had bought some time ago. Just tried one from another batch and works fine. Thanks again. Enjoy your videos BTW.

Unfortunately, it happens. I had a batch of 5 ESP32s and 3 of them wouldn't even flash. Luckily for me, that's been pretty rare. But glad you figured out the cause. Thanks for the kind words... and good luck with your own DIY projects.

Glad it helped! Like the ESP32, the RPi GPIO pins output 3.3V. So they are susceptible to the same issue, especially with longer data lines between the Pi and the start of the LED strips. Thanks for watching!

THANK YOU for saying that! That is the point I've tried to make over and over to those that say you don't need the level shifter. Maybe you don't... and maybe you'll be fine without it. But I'd a lot rather spend the extra $1 and a few minutes up front to assure a good data signal... as opposed to having everything work great on the benchtop... only to have the problem show up after final installation... just as happened to you. It's up to the individual to make their own decision, but I always use the shifter even if my data run is only a couple of inches, and will continue to recommend it to others as well. Again, thanks for sharing a real life situation that shows why... although I'm sorry you had to find it out in this way. Thanks for taking time to post a comment!

You might check out my latest video on common LED questions and watch the section on the logic level shifter, where I show and describe exactly how to wire it: ukposts.info/have/v-deo/f4qnaZCroWyal30.html I also have a written guide with various types of wiring diagrams for LED controllers. This might give you a better view of how to wire up the shifter: resinchemtech.blogspot.com/2023/01/wiring-diagrams.html Hope that info helps.

Just to be clear... you need to connect 3.3V, ground and your signal IN on the LV side of the shifter (3.3V to LV, GND to GND... then the signal to any of the LV1 - LV4 channels). On the high side/HV of the shifter, you connect 5V and ground (5V to HV, GND to GND.. then your outbound signal to the same HV1-HV4 as the signal on the LV side.) Think of it this way... you need to supply voltage on both the low (LV side) and high (HV side) corresponding to the voltage shift you wish to have in your signal. So, since the signal from the controller is 3.3V, you connect 3.3V as well as the signal on the LV side. Then you want the outgoing signal to be 5V, so you connect 5V on the high (HV) side. The outbound/shifted signal just needs to be connected to the same 'channel' number as the low side... if you use LV2 for the incoming signal, then you use HV2 for the outgoing signal. I hope that clears things up a bit! Let me know if you have any other questions.

Many thank @@ResinChemTech

@@ResinChemTech That's a really helpful explanation. Thanks.

You can find the design files for both the d1 mini and the full size board in my Thingiverse project: www.thingiverse.com/thing:4969733 As far as a video, not much to show! I load up the .gcode file into my 3D printer, hit the "Print" button and wait a couple of hours for it to finish! If you are looking for how to design 3D printed objects, or how to use a 3D printer, there are much better UKposts channels out there regarding 3D printing than I could ever cover. My designs are pretty basic... just a square box with a couple of holes in the end!

I'm not 100% sure I understand what you are asking, but I'll take a stab at what I think it is. Let me know if you are asking something different. This video was really targeted towards using ESP boards. Since you are using a Pi (and I'm guessing powering via 5V USB), the hookup might look a little different. When you say "...connect 'it' to directly to the LEDs", what do you mean by 'it'? Maybe the +5V line? If so, the logic level shifter requires +5V and gnd on one side and +3.3V and gnd on the other to shift the signal up (or down). I guess it really doesn't matter where you get the 5V and 3.3V supplies for the shifter. You could, which I believe is what you are suggesting, run the 5V side from the 5V supply at the LED strip. But I think that would just mean an extra longer wire run (assuming you are locating the Pi/controller a distance from the start of the LEDs). Alternatively, you could power the shifter using the 5V and 3.3V pins from the Pi. Ground can come from anywhere, but it does have to be connected to both sides of the shifter. I hope that is helpful. For my ESP boards, I am not powering the board via USB, but from the power supply and to the 5V pin on the board. At that point, I have both 5V and 3.3V available on the board and I use these to power the two sides of the logic level shifter (plus, obviously running the data line through the shifter as well). Because the ESP boards are not designed to handle high current, my 5V power is run in parallel to both the controller and the LED strip (you can't pass the 5V power through the controller with anything more than a few LEDs... well, you can, but you will probably fry the board at some point). It's a little different with the Pi (I believe) because the 5V pin can tolerate up to the max of the USB power supply... at least up to a couple of amps... but there may be other limitations just from the traces and other components on the board... I'm not an expert at the Pi or its specifications as I haven't used one with LED lighting. But I always recommend running your 5V power separately to the LED strips and not "through" the controller itself. If that's not what you were asking, just let me know and I'll try to answer your question.

Nope... the LED strip itself will boost the signal from pixel to pixel after it reaches the beginning of the strip, using the 5V from the power. So, while you will need to do 5V power injection along the strips, you won't have to worry about boosting the data signal along the way. Good luck with your project!

Hi! Glad you found me and hope I've been at least a little help. First, you are right in the fact that the code that I am running is Arduino based, but I didn't develop it. It is firmware called WLED by a developer called Aircookie. I don't know 100% which library it uses... FastLED I think? I am by no means an expert, but I do know that there can be issues with speed with using the I2C logic level shifter (like the one I show), but that usually results in some visible lag when doing larger displays or thinks like animation where you need a smooth 30 frames per second. I've used it with displays of up to 600 WS2812b pixels without issue... but again, I'm just running patterns and not trying to do real animation. I've not experienced anything like you describe, but all of my LED projects are ESP8266/ESP32 based. I haven't used either an Adafruit board or MicroPython in any of my LED projects. However, in situations where I have seen what you describe (the first pixel... or sometimes it's the first 2 or 3 behaving oddly), it has always been signal related for me... or interference with the signal. I had one install where the first 1-3 pixels would just light up by themselves at random. If I powered the strip on and back off, those pixels would turn off only to find them back on again the next day. In my situation, I was using the shifter, but I had a long (~15ft) run where all three wires (VIN, Signal and GND) were within a single cable. I guess interference can occur with long parallel runs of the ground and signal. In my case, adding a 33 Ohm resistor to the data signal cleared up the problem. I don't think this is your case as your data runs are short, but maybe it gives you another idea? I definitely appears to be some sort of signal problem/interference (which is pretty obvious I guess). Any chance you could test a different LED library to see if that makes any difference? I'm guessing you are maybe using the NeoPixel library? I think there is something similar to FastLED (maybe called FancyLED). Might be worth a shot. Let me know what you find. I'm still learning myself and I'm always happy to learn any troubleshooting tips I can share with others if they run into a similar problem!

The particular shifter shown will handle up to four different data signals, connected to the corresponding LV1-4 and HV1-4 pins and on each side you connect the low voltage (e.g. 3.3V) and ground to the LV and GND and high voltage (5V) to the HV and ground on the other side. Of course, you don't have to use all four channels, but can just use one if that's what you need. But it might be easier to look a a wiring diagram than trying to describe it here. I have a blog article that shows various LED controller wiring, including a section on wiring the shifter that you might find more helpful: resinchemtech.blogspot.com/2023/01/wiring-diagrams.html

There are many different ways to sync separate LED strips... too many for me to cover in detail here. I have not explicitly tried the way you are describing, but I _think_ it would work. Other options (that I have used or tried): - Split the single outgoing 5V data line from the shifter to two different strips (always synced) - Use multiple output pins (WLED), shift each and run to different strips. Create segments in WLED to control separately or treat all as one strip. - Use different controllers for each LED strip and use the WLED sync function to have the controllers sync to each other, effectively having all the strips running the same effect, color, etc. The last two options do require that you are using WLED and while they are versions of 'synching', the LEDs may not be precisely timed together on different strips. The only way for that to happen is to send the exact same signal to each LED strip. That is what I did for my stair lights, with a separate LED strip on each side. The difference is that I split the single data line AFTER it was shifted to 5V. You are talking about splitting the data line before shifting, multiple channels on the shifter, and using separate runs from the high side of the shifter. I do think that would work as well... I just haven't tried that method. Let me know if you do try and find that it works.

Thanks a lot for your input. This is the kind of information that helps me. Some of the ways you mentioned I thought about as well. This idea just popped in my head this morning and I thought I would reach out.Thanks for the examples. Hope to build my own project and post it soon. Much respect for all of the time and effort that goes into filming and editing not to mention the actual knowledge you share. Thank you.@@ResinChemTech

While you should always check the spec sheet of your LEDs to be sure, most 12V LEDs still only expect a 5V data signal. Of course if you are going to use a single power supply, you'll need to step the 12V down to 5V to power the controller and shifter, as most ESP dev boards operate on 5V.

I work in lightning protection. Part of my job is testing older systems to make sure they are still in working order. You would be surprised how much resistance and how quickly resistance can change due to things like a few extra feet of wire, a few degrees change in temperature or the wire being coiled up vs stretched out. If the voltage for controlling the led strip is already on the boarder line of not being high enough, I can see how a couple feet of extra wire could cause it to have some issues.

hey, what exactly is a sacrificial pixel? i'm new to this kind of stuff. thanks!

@@AlNexus Cut off one of your pixels and put it in the box and wire it next to your ESP (make sure the Data in/out order is correct). When data hits the ingress of the LED at 3.3v it then will output it at 5v, and you can extend to 10ft or so as suggested in the video.

@@ReganMcCullough thank you for replying!! this helped a lot in my project

You can probably get away without it... but as I mentioned, for the extra $1 or so and a few extra minutes of soldering, I always include it as a precaution that I won't have any data signal problems due to low voltage on the signal line. But you most likely won't have a problem with only a few inches. As far as recommendations, I keep a list of my favorite and recommended items here: resinchemtech.blogspot.com/2022/04/my-favorite-items.html You can find links for both the ESP32 and ESP8266, along with the logic level shifter that I use on that page. For an install of 268 lights, you can use either the ESP8266 or the ESP32. If you plan on using a sound-reactive version of WLED, then I'd recommend the ESP32. Otherwise, you can use an ESP8266. Personally, I like the D1 Mini or the ESP32 mini just for their smaller size and therefore the overall controller size is smaller. Just let me know if you have any additional questions!

This device is meant to shift a logic signal... not true voltage like you are referring to. If you need to step up or step down voltage for powering other devices, sensors, etc. you need to use a step-up or step down voltage converter. These are often referred to as buck converters.

I cannot post images or attach files in UKposts comments. I have wiring diagrams available in my WLED controller-related blog here: resinchemtech.blogspot.com/2021/01/building-your-own-led-light-strip.html It has multiple versions of wring for both the full size NodeMCU and the D1 Mini. The wiring for the ESP32 is basically the same... the pins are just in different locations. You can also find a wiring diagram for a WLED controller on the official WLED web site: kno.wled.ge/basics/getting-started/ That's about all the help I can offer in terms of a wiring diagram.

I'll admit to not being as familiar with the performance range of WS2811, but according to the spec sheet, the data signal input range is 3.5-5.5V. But again, like with WS2812, you may get OK operation at less than that. Or it might work with one board/LED lights, but have issues with the next. Inclusion of a logic level shifter isn't always required in every situation, but I will still continue to recommend it so that you are operating within the actual specifications: www.tme.eu/Document/26d574b43ad9ddaffa4d5bcd140ec145/WS2811.pdf

@@ResinChemTech I started thinking more into this and realized that the difference between your experiment and my real life setup of a long WS2811 strip is that I’m powering my mini D1 with a separate 5v 1A power supply through the micro USB port and I’m powering my LED strip with 2 24v power supplies. The ground on the mini D1 is tied in directly to the ground on 1 of the 24v power supplies, about 1.5 ft run of ground wire, and all 3 power supplies are on the same outlet. The data out of the mini D1 is about a 10 ft run to one end of the strip, same power supply as the ground. Can you humor me and try a separate 5v power supply in the micro USB port and power your strip with a separate 5v power supply and see if it makes any difference in your test? I’ve got the exact same mini D1 board so this is really bugging the heck out of me!

I agree 100%! But I repeatedly have comments telling me that they aren't needed (and all the reasons why), which is why I threw this video together. While in some cases, you can probably get away without it, the first time you complete an install and end up moving the controller farther away, you'll really wish you had spent that extra dollar and a few minutes soldering. Don't want to use one... that's completely your choice. I opt to always include one and not take the risk of potential issues. Thanks for the comment (and answering the other comment about the diagram... I appreciate the help!).

It depends on numerous factors... the gauge of wire, type of wire, temperature, etc. I have run at least 15 feet between the controller/shifter and the start of the LED strip without issue, but I can't really give you a definite maximum distance. I always recommend bench testing first with the actual controller, LEDs and wire you plan on using. If you have a really long run and need to "boost" the signal again along the way, you can use something like this data booster by Quindor: quinled.info/quinled-data-booster-for-sale/

While boosting the signal from 3.3V to 5V will definitely help with the longer run, 30-40 meters is still a VERY long run between the controller and the start of the LED strip. My longest run was about 8-10 meters, so I've not tested anything longer than that. I would also use the largest gauge wire that you can because it will result in lower resistance and therefore less voltage drop. It might be necessary to 'boost' the data signal again along the way. Quindor over at QuinLED sells a booster meant just for this purpose: quinled.info/quinled-data-booster-for-sale/ This also has a resistor onboard, which you may also need on your data line... especially if you have a long run where the data line and ground line are running together in parallel. This can introduce noise into the data line and cause all sorts of problems with your LEDs. Good luck with your project. I'd recommend bench testing the signal at the planned distance prior to completing the actual install. That way you'll know if you need to boost the signal further and/or include a resistor on the data line.

@@ResinChemTech Thank you for sharing your experience.

If I understand you correctly, you are using 8 different GPIO pins connected to your LEDs? I'm not sure I fully understand why you would use multiple GPIO pins in this case... if you want different segments of the LEDs to be different colors or have a different effect, you could use one or two GPIO pins and use WLED segments (I am assuming you are using WLED). Maybe two GPIO pins if you wanted a different signal to each side of the stairs, but 8 sure seems like a lot for a run of only about 16 feet. But in any case, if you are using multiple data signal outputs from different GPIO pins, they are all outputting at 3.3V while the LED strip expects a 5V signal. So, yes, for a more reliable signal, each data line should be shifted up to 5V with the logic level shifter. Sorry if I misunderstood your particular install. But I think the general rule (for me) is that any and all data signals should run through the shifter and be boosted to 5V before going to the LED strip.

The VIN is the same as 5V. Some boards label it as 5V and others as VIN. Check the pinout of your particular board to be sure. However, if for some odd reason you truly don't have a 5V pin (if you are powering the board off a 5V connected to VIN, then this is the 5V pin), you can use split off the 5V that is running to power the board and connect that to the HV side of the logic level shifter. Hope that helps. It can be a bit confusion when first starting out since different manufacturers label pins differently. Let me know if you have any additional questions.

@@ResinChemTech Thank you very much. that is what I thought but I wanted to double check before I blow something up.

@@ResinChemTech Thank you very much

Well, even 12V LEDs expect a 5V signal just like 5V LEDs. So this same shifter would work for 12V.... but you do still need to feed it 5V and 3.3V. If you are using a 5V controller (e.g. ESP8266 or ESP32), then you should already be feeding in the 5V and you can get the 3.3V from the 3V3 pin on the ESP board. I get that a lot... thinking that 12V LEDs require a 12V data signal. But that's not the case... most LED strips, regardless of whether they are powered by 5V or 12V, still expect a 5V data signal. But you can check the spec sheet of your particular LEDs to be absolutely sure.

There is not a set distance where you might experience problems or might be fine. You could have one install at 3 feet with no issues at all, but another identical install with the same components that might have issues. There are variabilities in the manufacturing and tolerances of the components that just simply make it impossible to say that you will or won't have signal issues as a particular distance. Other factors like the size of wire used will also have an impact. That's why my personal recommendation is to always include a shifter, then you simply don't have to worry about it. But you are free to make your own decision on your project. You might be fine at 3 feet, or even longer, with one install, but experience issues with another install at even less than 3 feet.

Most 12V and 24V LEDs still expect a 5V data signal. Check your strips specifications to be sure, but if does expect a 5V signal then this same process is used regardless of the voltage used to actually power the LEDs.

You can actually use the 5V pin from the ESP8266/32 to power one side of the shifter and the 3.3V pin to power the other. The shifter that I use has 4 channels available. You run 5V and ground to one side (HV) and 3.3V and ground to the other side (LV). You then have 4 different channels, or data signals, that you can shift up from the outbound data GPIO pin and then out to your LED strips (effectively boosting the signal from 3.3V to 5V). I can't post any diagrams here in the comments, but you can see a diagram showing what I am describing in my related blog article: resinchemtech.blogspot.com/2021/08/sound-reactive-rgb-floor-lamps.html And yes, you will want to power the two LED strips in parallel to your controller and not through either the board or controller since you will potentially be drawing more amps then the board was designed to handle. There's a diagram of this at the related blog as well.

@@ResinChemTech you mean powering both directions, 5v and 3,3v, or only one side? Thanks for many answers

If you look at the logic level shifter (at least the one I am using), on one side you will see pins for HV1, HV2, HV, GND, HV3 and HV4. The other side has the same thing, except LV. You feed 5.0V to the HV pin, ground to GND on one side, then 3.3V to the LV pin and ground to GND on the other... yes, both 5V to one side and 3.3V to the other. Then you can feed your data signals from the ESP8266/32 pins to any of the LV1 - LV4 pins. You then feed the signal out to the LEDs from the same corresponding HV1 - HV4 pins on the other side of the shifter. Just be sure to use the same pin "number" on each side. For example, if you data line comes in and is connected to LV3, then your data out line to the LEDs need to be connected to HV3. Each channel is separate... so you can't, for example, connect the data in to LV1 and the data out to HV3. Again, this is shown in the wire diagram I referenced above.

This is actually pretty easy. Once you have the desired settings (e.g. color, effect, brightness, etc.), save it as a preset. Note the number of that preset. Then go to Config -> LED Preferences. Scroll down to the defaults. You will see an option to apply a preset at boot. Enter the number of your preset. You can also opt to have the LEDs turn on automatically when booting. Once set, any time power is applied or the board reboots, your desired settings will be applied. Now, note that this only applies at boot. If the board maintains power and doesn't reboot, WLED should always turn back on in the same state that it was when last turned off, which I think is what you were wanting. If for example, you have the LEDs set to a green and blue chase effect and turn off with the Power button in the WLED interface, it will be green/blue chase the next time you turn them on via this power button... unless the board loses power or reboots. Then defaults are from the LED preferences are applied. If you find that you when you turn the "Power" off via the interface and then turn it back on later, the LED strip is a different color (especially if it is solid orange), then this is a sign that your board is rebooting at random. I actually had this with a board once... but by setting a default boot preset, it always turns back on in the preferred color and effect.... even after a reboot. I hope that helps.

@@ResinChemTech yes this helps a lot . Thanks again

Well, if you are using a 12V power supply and are not stepping that down to 5V, you will definitely fry the ESP32 as most ESP dev boards are designed to run on 5V. I'm assuming you are using 12V LEDs? If you are trying to connect 12V to LEDs designed for 5V, then the same thing will happen...they will burn out. So, first you need to assure you are using the right voltage for your components. When it come to 'size' in terms of amps, when using 5V strips, you should allow for a maximum of 0.06A per LED. So for 300 pixels, you need a power supply capable of providing at least (300 x 0.06A) = 18A. I'd recommend 20A if you want to run full bright white. You might be able to get by with 15A, but you'd need to limit the brightness using the brightness limiter setting in WLED.

@@ResinChemTech i got ya that makes a lot more sense. Thank you should save me some boards

@@ResinChemTech so, i probably fried my lights too, most most light up and the ones that do only have 1 or 2 barley lit up one in green one in red.

You can use one or the other, both or none... whatever works in your situation and installation. That's why I recommend that you do a bench test with the total number of LEDs and the expected wire length run before you actually install the lights. That way you can see if you have any issues with dimming/flickering of the LEDs and can easily add/remove things like the resistor, power injection, etc. until the lights are behaving as expected. This is much easier to do on the bench than it is after the install. And note that what works for one install may not work for the next, even if they are nearly identical. I bench test every LED project prior to installation, but ultimately that is your decision. Good luck with your project!

Thank you so much for your response sir..

I have once... to solve a signal issue due to about a 15ft. run with a 3-wire insulated cable that had both power and signal running in parallel within the cable and was causing some intermittent problems. I used a 33 ohm resistor in that case on the data line. But that is the only situation in nearly two dozen WLED installs where I had a problem that needed a resistor.

I haven't but just to confirm your wiring... 3.3V and GND connect to the LV and GND on the shifter. 5V and GND to HV and GND on the other side. Your LED signal from the ESP32 GPIO pin should go to any one of the 4 channels on the LV side and the wiring out to the LEDs to the same corresponding channel on the HV side. I don't know of any reason why the shifter would impact the wt32, as it is a serial device and the only thing this shifter is doing is boosting the data line from 3.3V to 5V. If connecting to the shifter is causing a shutdown, then there is either a short somewhere (assure you haven't inadvertently created a solder bridge somewhere) or it is possible that you just have a faulty shifter. Personally, I've never had a shifter to do anything like that... but I have also not used a wt32 either.

@@ResinChemTech Thank you for the reply. Yes I have 5volt and ground from android charger to power the wt32 and also the HV and GND of the shifter. I then grab 3.3 and GND from the wt32 to the LV and LV GND side of the shifter. When I did this with the ESP32 it worked great. When I plug both 5v and 3.3v into the WT32 it don't like it. I see the the WT32 is either 5 or 3.3 volt so I don't think its and output on the 3.3v side. Just curious if you ever ran into this.

As I mentioned, I have not used a wt32 before so I can't say for sure, but from just glancing at it, I would think that as long as you only feeding incoming voltage TO the board on 5V, should be able to provide 3.3V OUT to another device. Again assure that you aren't somehow providing both 5V and 3.3V to the board... it probably won't like that! You might check out the wiring diagrams for my WLED controller in this blog article: ( resinchemtech.blogspot.com/2021/01/building-your-own-led-light-strip.html ). This won't be exactly the same as you wt32, but it does show how the logic level shifter should be wired to the microcontroller. Otherwise, you might need to check the spec sheet on the wt32 and what it says about connecting the 5V and 3.3V pins.

@ResinChem Tech

There are two potential things that come to mind here. First, assure you have a common ground for everything. Both power supplies, the controller and all segments of the LED strips need to be connected to a common ground. You should not connect the +V from multiple power supplies together, but connecting all the grounds gives a common ground reference. The LEDs can misbehave due to different ground potentials... which may be what is happening in your case. From what you describe, it sounds like you might just be able to connect the grounds from the two power supplies together. The other thing to double check is that you have the correct number of LEDs defined in WLED for the total number of LEDs in both strips. It could be a number of other things, but those are two things I'd verify. Alternatively, instead of using two separate power supplies, you could just run separate power injection wires and inject +V/gnd from the first power supply to the end and/or midstrip of the total run. You didn't say now many LEDs you have, nor what voltage strips they are, but also note that you should only have so many LEDs per data pin before you need to consider multiple data pins. You can find the recommended number on the official WLED web site: kno.wled.ge/features/multi-strip/

Either option will work equally well. It somewhat comes down to personal preference, soldering abilities and how you want to build the controller. For me personally, I prefer to spend $1 for the level shifter that I can securely solder onto my controller board as opposed to trying to have the extra wiring and a 'dangling' pixel (or hot gluing the pixel to the board). To me, it makes a neater and more secure controller with less likelihood of a soldered wire coming loose... but that's just my personal preference. From a purely functional standpoint, either option will work the same as long as the sacrificial pixel is kept close to the ESP controller.

@@ResinChemTech thanks for reply... I have seen poeple complain about lag while using level shifters.... May be some ms delay compared to Sacrifical led... Not sure is that true.... Have you seen any difference between those..... If not... Please can you try and update on this.. Comparing sacrifical led and level shifter

That might be a valid point, although I suspect it only becomes apparent with a high number of pixels. I do know that the developer of WLED states that about 800 pixels is the limit that an ESP8266 can manage on a single pin without starting to experience lag. I suppose that any additional lag introduced by the logic level shifter might impact this... it is entirely possible. But I have not personally experienced it, although my biggest single install from a single controller pin is 600 pixels. While he doesn't explicitly address the logic level shifter, Quindor over at QuinLED has a great article that describes that lag is actually caused by the LED technology and not truly by the controller (and by extension the logic level shifter, as he also includes one in all his pre-built boards): quinled.info/2021/03/23/max-amount-of-addressable-leds/ And while I'm sure manufacturers vary widely, the spec sheet on the shifter I use reports a max response time of only 42 nanoseconds... much, much faster than the LEDs can respond. So, my guess (although I can't prove it) is that lag is caused by trying to run too many pixels in a single string/connection and nothing to do with either the controller nor the logic level shifter. I can certainly consider a video showing use of both the logic level shifter vs. the pixel, but I'm guessing I'd have to use a very large number of pixels (which I would have to order and buy, along with a big enough power supply to run them), and even then it might be difficult to show that lag on video.

@@ResinChemTech that's really pain to make contents show in video... But I really appreciate your response... It means a lot to me... You have always been soo motivate.... I will try myself and update you If possible....thanks again for these type of videos and contents... These have always been soo helpful.... Also as a suggestion... Can you make videos on diy smart curtain open and close integrates with HA... Most of the devices you use.. Cant be purchased in my place or either it's very very expensive. Diy works for eveyone I guess.

You can find recommendations for the max number of LEDs and best performance here: kno.wled.ge/features/multi-strip/

All I can tell you is what I am showing is the official spec sheet. I'm guessing that they are showing a percentage because Vdd, listed right below, is a range and not a fixed value either (Vdd=4.5~5.5V)? So, the data line-in minimum is 70% of whatever Vdd happens to be. But I am far from being an electrical engineer, so maybe someone else can address that question better than I can.

@@ResinChemTech So if I follow that thinking, a lower VDD would allow 3.3v to work better at longer distance..... That doesn't quite follow either does it? I'm all aboard the 5v train, but that datasheet . . . :)

Primarily because the signal being received by the first pixel is too weak (voltage too low to begin with). So even though the first pixel "boosts" the signal to the next one, the signal received by the first pixel was already "bad". So the first pixel is boosting a 'bad' signal and therefore it impacts all the LEDs in the strip. It like receiving a poor AM radio reception... you can boost the volume or try to process that signal in various ways, but if the signal is missing information from the source, no increase in gain is going to clear up that signal. By shifting/boosting that signal, you are assuring the first pixel in the strip receives a good, complete data signal. Each frame of the data signal includes the 'instructions' for each LED in the the entire strip. So any missing or corrupt portions of that signal has the ability to impact any LED.... not just the first one. Hope that info helps.

@@ResinChemTech That's a very helpful explanation. Thank you very much!

There could be a number of causes. First, assure you have a common ground between the controller and LEDs. But it sounds like you are using the same power supply for both, so you should be good on that issue. Next, have you tried power injection for the LEDs? 300 would be a lot for 5V without power injection, but it might not be necessary for 12V. It is worth a try to add a 12V connection to the end of the strip as well as the beginning to see if that helps. Finally, you didn't mention that you were using a logic level shifter. If you aren't then this is likely your problem. It sounds like the LEDs are behaving similar to the ones I showed in this video. Remember that the ESP32 only outputs 3.3V on it's GPIO pins... but the LED strip expects a 5V data signal. Occasionally, you can squeak by without one as 3.3V is at the very minimum voltage. But any voltage drop and the LEDs will start to behave just like you describe... because they are not receiving a good data signal. If you are not using a logic level shifter, I strongly suggest this is your problem... and adding the shifter will likely fix it as well.

@@ResinChemTech sure will try this with logic level shifter. There were 2 reasons I did not use Logic level shifter as per your video ... 1: In your video lights started behaving odd once distance between ESP and strip was more than 6 feet. In my case it was 2 feet. 2nd reason from your video i could not understand wiring for logic level shifter and hence thought due to distance between ESP and strip I can do away with it.... Requesting if you can plz share a Lil detail video of using logic level shifter wiring.... Also would a 5v data signal for ws2815 be enough? Also is it because that ws2815 is 5Amp rating and my power supply is 12v 10amp?

While I haven't used WS2815 strips myself, according to the data sheet, acceptable data signal voltages are 4.5V - 5.5V, which is even higher than the 3.5V for the WS2182b. This almost certainly means that your problem is low voltage on your data line... something the logic level shifter would fix. It also explains why you have the problem with even a short wiring run. I cannot share a diagram here in the comments, but there are plenty of other online resources for how to wire the logic level shifter. But the short of it is that you feed 3.3V and ground to the LV and GND pins of the shifter. You can use the 3V3 pin of the ESP32 to supply the shifter with the 3.3V. Your data signal from the ESP32 feeds any one of the four channels on the LV side (LV1 - LV4). On the opposite side, you feed 5V and ground to the HV and GND pins. The outgoing data line to your LEDs is then connected to the same HV channel as the LV signal on the opposite side (if you data line from the ESP32 is connected to LV2 for instance, you connect the line to your LEDs to HV2 on the other side. That's really all there is to it.

@@ResinChemTech sure thank you so much for checking the datasheet for me and confirming the problem area. I will try this for sure and share the photo or video with you....

@@ResinChemTech It worked. Would love to show this on a video. Is there a way i can share my video with you? Like a WhatsApp number? Also how do i now add MAX9814 for Sound Reactive to this setup itself? I know u have a video on Sound Reactive ukposts.info/have/v-deo/l6h2eoCEp2pp22g.html but thats with WeMos D1 Mini. What would be the wiring for ESP32?

It should be linked in his blog in the video discription, if not, Check out some of his other videos i know its in his "how to build a controller for $6" video

Thanks John! You're exactly right. The diagram/build for the controller can be found in my Build your own LED Controller: ukposts.info/have/v-deo/oZanaq2Naoiel3U.html and written info/diagrams can be found in the related blog: resinchemtech.blogspot.com/2021/01/building-your-own-led-light-strip.html

Thanks Sir

Yeah, I think I mention a similar thing in a number of my videos when it comes to using a shifter. What might work for one build might have issues in another exact same build... due to variances in tolerances of the strips, ESPs, etc. For me, it is just easier to always install a shifter during the controller build and not have to worry about the data signal strength. But thanks for watching... and taking the time to leave a comment.

Basically, you can use the same 5V, 3.3V, ground pins like you would with the ESP8266. You can use a number of different pins for the data signal, but GPIO16 is the default. I have a number of standard wiring diagrams for the LED controller that I use, including some of the ESP32. You can find them here: resinchemtech.blogspot.com/2023/01/wiring-diagrams.html

@@ResinChemTech can esp32 able to control 2 visual effects with 2 separate wled?, ty

One ESP32 can control more than one LED strip by using different data pins and creating segments within WLED. See for information here at the official WLED web site: kno.wled.ge/features/multi-strip/

@@ResinChemTech thank you teacher...😆👍