Need help with extremely complicated ws2812b strips project!

Projects
61

How you wire them and number of controllers etc will all depend on the software, controller and firmware you are using.

In the diagrams you show variable spacing between LEDs, this is not really a thing outside of the main #leds/meter.
If you want fine tuned distance between leds then this will be more custom.

That said, Im not really sure why this is a puzzle, so I can only assume you dont understand how they work; so lets get back to WS2812b basics.
The LEDs (WS2812b) will run off 5 Volts.
Modern microcontrollers tends to be 3.3 V (e.g. esp32, pi pico).
the led strips can be cut and joined as needed; as already shown there will be “cut points” and pads to to solder the extension strip or wires for power/data etc.

So as already covered we supply the 5V to the power pin on the LED stripe (DONT provide 5V to your controller if you have USB connected).
Connect the controller ground pin to the LED ground pin (or some common ground point).
So now we are down to how the LEDs actual know what to do.
If you have a strip of 5 leds.
the controller data wire goes to the first LED data in.
that LED will get the the first clocked out RGB data; keep it for itself and not pass it on (via its Data out pin).
The 2nd set of RGB data will go into the first led, but it already has its data, so it will blindly send that out its data out pin, which goes into the 2nd LEDs data in. Since this is the first bit of data the 2nd LEDs sees, it will work just like the first LED, keep the “now” first RGB data set for itself, then when it sees more data, it will pass it on, and repeat until not more data is sent to the stripe. when no data is seen for a fixed period of time (see the data sheets), they all will then display the color the collected.

As such, you can take the data out of the end of one stripe (regardless of length and # leds (for the most part), and connect that to the data in of the next collection of leds.

Very rough image
IMPORTANT NOTE: This is an example only, you will need to check the actual pins on the strips for what wire goes to what wire… e.g. +V to +V, Gnd to Gnd, Data Out to Data In.
(I did not check this for this image, as I don’t have time to do detailed design and write guides)

image
image988×676 219 KB

Assuming (for this example) that there are 12 Leds per finger, 6 in the front and 6 in the back.
Then …
Finger 1: LEDs 1 - 12
Finger 2: LEDs 13 - 24
Finger 3: LEDs 25 - 36
Finger 4: LEDs 37 - 48
Finger 5: LEDs 49 - 60

So map that into your sequence software (how ever that needs it to happen) and you now have “ONE” long strip of 60 LEDs spread over 5 fingers.

Given challenges with wiring, you may need to reverse the direction of every 2nd strip so that the data link can happen on the same side as where the other one ends up (rather then wrap around the wrist). That will mean every second finger will run on opposite direction, so that would then need to be feed into your sequencer.

I want to stress here… that their may be limits in the software you choose to use and your physical design will need to work around that. If you are going to write all your own software then you can then make the software work around the physical challenges.

Also for the choice in controller, again that will need to be support by what ever software you want to use; but if your going to write your own code, then use what you know.

e.g. Im sure any 32 bit controller should have no issues, like the ESP32 and RMT (they provide the LED example code for this), but if using something like WLED then you need to check what controllers that supports.

2 Likes
62

According to your diagram for the glove’s wires and LED strips, it seems the black 5v ground wires are interconnected, as is the case for the red 5v positive wires. Does that mean that they all are soldered together in-between each 2nd strip? I doubt that is the case, but having the wires connect in the middle doesn’t make any sense to me. Could you explain those in-between connections?

Also, just to be on the less-risky, safer side; I’m aware of splitter cables only supporting a max. of 4 connections to 4 LED strips. If there are 5 LED strips, would it overload the microcontroller, or affect the performance of the power supply, or somewhere along those lines?

1 Like
63

You can loop the led power as needed, just keep in mind the number of leds in one run.
if you have 1 strip of 60 Leds or 10 Strips of 6 leds, its the same thing.

edit:
Note: splitters may not be doing what you think the are.
You need to look into the data sheets for what the do and ensure you don’t exceed there max ratings. But that does not mean there are not better ways to do it.
e.g. If you have 5 fingers, each of 12 LEDs and every led on each finger does the some thing as the one on the next finger… so all fingers look the same; this might imply that you can split/fan out the one data pin to the 5 fingers…
And this is the key point:

  • that MAY work, but MAY not work.
    when you have the LEDs as designed to work, its one data out to one data in and this is a point to point signal that will travel over that one wire. The moment you fork that out (just by joining wires) you now have create a “bus” which will behave differently; but may not be different enough and still work; but can just as easy create random data errors; or at the worst not work at all.

To ensure it works, you would treat all the fingers as one big string and send the data to each led as is meant to.

Note: that the only way you are going to know it spitting will work (on the data line) is to build fully, and test it in the final working settings. e.g. It MAY work when on the bench at home, then NOT work when someone is wearing the suit.

So, split if you want to, but that will be a design risk on your side, that may end up with you needing to re-wire it later, then re-program the sequence to adjust for the rewire.

2 Likes
64

There is no point is trying to design the glove until you have got the prototype up and running and confirmed that you can do what’s needed in terms of battery power, wiring, effects and so on. What you discover from that exercise will provide the detail needed for the body sections, and will also advise what is different about the glove and where you need more information or additional decisions. Get something simple working and move on from there. It’s even possible that the prototype will eventually morf into the glove - if not it might become a headband.

2 Likes
65

That’s the most complicated hypothesis I have ever tried to understand. Based on your suggestion: I’m confused as to which type of wires out there match your insulation AND thickness/diameter suggestions, where/which ones to find, if it has a specific name/type thats easy to identify, etc. Based on my searches, I may be wrong or I may be right, but this is what I found:

https://www.tradezone.com.au/product/electra-3-25mm-2-core-low-voltage-double-insulated-automotive-cable-100-metre-drum-22938.html

Also there is an article that I read up that mentioned AWG wires with a range of 18-22. Perhaps those are compatible with the project’s requirements?

And are the wires’ colour coding (white, green, and red) any important to ensure the right data and energy input & output?

1 Like
66

I also wanted to compare ESP32-C3 microcontrollers with other ESP32 microcontrollers. What’s the difference, which one is recommended on the easier side/results expectations and any advantages/disadvantages compared to other ESP32 modules?

1 Like
67

the wire math… that is an example. There are plenty of online calculators that will give you the gauge of the wire (in US or metric). The actual insulation was just an educated guess as that can vary based on minimum thickness for that size wire as well as deal with mechanical protection and material type. As such you work with your MAX +buffer current needs and length of the wire (and connectors etc). Then feed that into a DC (as this will be DC current) calculator and that will spit out the size wire (some will round to the nearest common size, some will give you an exact value. Once you know the gauge of the wire, you then can use any wire that is that big or bigger.
Side note: AWG (Americian Wire Guage) the smaller the number the bigger the wire. e.g. AWG12 can carry more current then AWG14.

As to the ESP32 module… this can be tricky as there are ESP32 chips (that will support things; e.g. some will have USB on board some wont). Then Those chips can be made into boards with other things added, such as Power regulators, USB to UART (for programing), exposed Pins etc.
So, you could have 2 ESP32 S2 chips but with very different boards you would buy pre-built.

As such, if you want to know the Chip Differences, read the data sheets.

Normally you would have a need.
e.g.
I need

  • X number of GPIO Pins exposed.
  • Need Y amount of flash storage
  • Need hardware level SPI (and support minimum speed of Z bps)
    and so on…
    You then read the data sheets for each.
    Keeping in mind, that, for example, if you are not going to write your own code, then you need to use a board supported by the firmware/product you plan to use.

If you are going to write your own code, then make sure you chip of choose is already supported; sometimes there can be a lag between chip release and all the needed libraries being supported.

I write my own code using a text editor, then copile and flash via the IDF provided by espressif for the ESP32 series. Since the make the chips they will update their IDF to support them and I can then do what I needed without relying on someone else to update their libraries.
BUT: This is not something people do if they don’t know enough about coding for micro-controllers. I expect most people would use Arduino, as such if you use that, you need to ensure it supports the “board” that has the chip.

But you need to define what you need then find the best match for your needs.

1 Like
68

Hey @Darkmaster288884,

While not a complete comparison of all the ESP32 variants available this page is a decent starting point for comparing these chips.

https://docs.espressif.com/projects/esp-idf/en/v5.0.4/esp32s3/hw-reference/chip-series-comparison.html

Hopefully, that can give you some idea of the differences between some of the chips available.

I am personally a big fan of the Seeed Studio XIAO ESP32 boards as the small form factor means they fit nicely into nearly every project regardless of size. My go-to for general microcontroller projects is the ESP32-C3 as it is normally priced somewhere in the middle of the range, has all the wireless capabilities that I generally rely on like WiFi and Bluetooth and has a processessor capable of running most small scale projects.

If I find that boards processing power is limiting the project I will step up to the ESP32-S3 for the faster processor.

Hope this helps! :slight_smile:

69

thanks @Samuel just my thoughts and a small value add… My personal thoughts tend to be around flash size v code size and OTA Updates.
e.g. If your firmware is < 1 Meg. then a 4 Meg Flash is fine. 1 Meg for original flashed firmware
1 Meg for OTA1
1 Meg for OTA2
then about 800-900 Kbytes for Spiffs/File Store. (leaving 100-200 K for boot etc)
The moment your firmware grows over 1Meg then you need to repartition. so that’s where 8 or 16 Meg of flash have value.
But, all of this is around the OTA and on-chip file storage needs.

The amount of flash can vary between vendor boards, but the main chip features will be the same.

i.e. back to, what’s needed. I was using the ESP32 Wroom as my go to, then have the S3 when I want native USB, and have lots of those cheap and nasty for general play time.

I select form factor based on project needs. If space/physical size is not important, then I tend to go with the cheapest that provides my project needs (GPIO, Flash Size).

1 Like
70

I’m assuming the thicker the wire, the better the DC flow (5v for this case), so the most thickest (size 0?) is more than suitable for ws2812b LED strips and say; Seeed Studio XIAO ESP32-S3 and up to a total of 43.2 (or a little less?) watts powered by 2x 18650 battery packs (unless 20K mAh powerbanks is a considerable substitute since each strip’s overall total power is 43.2W= 8640 mAh) and a voltage booster if necessary.

The included connector wires seem to be color-coded (white, green, red) so I guess we need 3-set of wires that correspond to the same colors and thickness in order to get them to function and test them, plus to match the 3 contact tabs on the ends of the LED strips? I’ve seen someone else mention 18 AWG is the most ideal size for ws2812b LEDs aling with 1mm insulation. Again, it feels like I need a relevant example of which type of wires, if the 3-pin layout is crucial for the wires to be a set of 3, color-coding seems to not be the focus or necessary tip in order to connect them to the corrrect points, AWG or not, etc. at least a link to an example that matches the power info I provided and how many LEDs, the merchant/store, as this is confusing to compare between advices and suggestions from different sources and having to do more maths and theorising rather than be directed to a type of existing wire(s) like how others can recommend microcontrollers such as far as mentioning ESP32s or Pi Picos by name and functions without making it complicated for one to guess what microcontroller is ideally fit for my project constraints, without name type, etc.

I tried to use the link I posted in my last reply to query if that type of wire can do the job but it seems I didn’t get a simple answer as “yes it is compatible” or “no it is not suitable”. Rather it’s all "depending on parameters such as buffer needs (what’s relevant about that), mechanical protection (overheating & insulation- related?), material type (color-coding/universally all wires are copper no longer an important factor?) and thus more frustrating brainstorming and researching that goes nowhere or redirects me non-stop endlessly between different “experts” and sources. Just all I need is:
(Most suitable) AWG size
Is the White-Green-Red color-coding an important factor in choosing wire layout? The 3-pins on the ends of the strip, are they important in any way to determine how to solder between modules and wire ends?
The total power of each 1m strip containing 144LEDS as confirmed by me is 43.2 Watts, tell me at least one wire you have in mind that is sufficient to cater to 43.2 Watts or a little less.
Input voltage: DC 5V.
Assuming the LED strips consume more power than the microcontroller. 18650 battery pack powers the microcontroller while 20k mAh powerbank powers the LED strips?

Anything this straightforward and simple would save so much time, I’ve done as much research and calculating so instead of going in circles of “depends of this and that” or “AWG 12 carries more current than AWG 14 so I would take that as 'the thickest there is say: size 0 the better as the current it carries matters” and something like that and more obvious answers as “18 AWG will meet those needs given what components you indicated you will be using” or “I gave you the info: 0.3 per LEDs, 144 LEDs per strip, 43.2 Watts total power per strip, now tell me, powerbank does the job or 18650 battery pack?” And you tell me which of the 2 is most ideal. Something along those lines.

71

Neither. 43.2W @ 5V + 8.64A.

72

So I suppose the only power supply is thru a wall power socket and nothing portable and compact?

73

The challenge with 18650 is there are many different versions of them.
e.g.

so this claims to be able to supply upto 20 A with a life of 2.5A/hr.
Rough math: 10A (buffer on specs) with 2.5Ahr = 0.25Hrs (15mins)
Its output voltage is 3.6V so you will then need something to convert that to 5V for the stripe and to supply the controller 5V input (which then goes via the controllers 3.3V regulator.

My only comment here would be to test it. I will assume at higher current draws it will warm up, but if the specs are correct it should be ok.

I want to stress, double check the specs on the 18650 to ensure the one your looking at can supply “more” then the needed current.

e.g. : Discharge Current: 20A Continue, 20A Pulse

Revisit, from above for some ideas.

74

I would like to make a side comment:
We are not trying to make it hard for you, rather give you the tools so you can work it out yourself.

As the designer of the project, its up to you to ensure everything will work as expected. and within specs.

Its not our job to do the design for you. If you want someone to design it for you, I am sure there will be people who will be happy as a paid job.

We can only comment on exact things and while you may think you are supplying exact information, your not.
So when you said you need portable power for 43.2W @ 5V thats now an exact thing. So from that advice can be given on what could supply that; But you need to keep in mind that its up to you to ensure that is what you need. (Im not saying its not, but this is why we go threw the design phase).

Electronics (at a design phase) is really founded on (in my opinion) 3 Main things.

  1. Education: this can be formal or hobby learning; of which mistakes are part of the learning. Im sure everyone here has made there share of mistakes.

  2. Reading datasheets (and yes it takes time). This is part of the job of the designer. “I need some that that will…” then search for the best component to do that by reading the data sheets. While watching a video may give some ideas, most of them are more a demo then how things work… So I repeat, read the datasheets.

  3. Experiment/Prototype. While some very experienced people may get things right on the first go, at the hobby level that does not happen very often and several versions of a product would be common. So any expectation that it will work the first time is not realistic at a hobby level (for custom projects).

One of my project had a very specific need for a “fixed color” led. The color was set. But we also needed viewing angles and distance requirements to be meet. So I spent weeks reading datasheets on all the different LEDs I could find and filtered doom my list of options. Then I would order 20 for a test. (The final project needed a bit over 7000 of them). Most, while looking good on paper, did not pass out field test. But after about 3 months, I finally found the one that worked for that project. Even recommended LEDs from others did not work as we needed them to for our project.

My point here, is its hard to give correct advice on the correct bits without doing the design; Not knowing the exact needs its just a guess that may or may not work. Advice given on guesses may or may not work, then people get upset that what they where told to buy did not work.

As to the cable size, you can go all AWG 0, Im sure from a current perspective it will work; But being 8.25mm thick, it will be heavy and bulky. So the trick is to select the gauge that will support the max/peek current + a bit more; but don’t go to big that it cost too much, its too bulky etc.

Before anyone can recommend wire to buy, the will need to know the wire layout and currents that wire will supply over what distance. It would not be uncommon to have more then one wire size. e.g. Power Supply will normally be bigger as it passed more current. but signal wires normally don’t send much current at all, so thinner wires will work (just do the math and use the calculators)

As to your comments about wire colors… that will be a design choice from whats available. But try to define a standard for the entire project.
e.g.
Black = Ground everywhere.
Red = +5V
Orange = 3.3V
Green = signal/data for LEDs
White = uC sync inter links.

the key idea for colors is to allow easy visual inspection of your hook-ups/wiring. If all wires had the some color, then when you need to solder/connect a wire to the LED strip you would need to physically triple check you got the correct wire.

1 Like
75

You can’t rely on the current consumption figure that you have calculated. The numbers might be correct, but you do not know if it really applies to your project. As the current consumption affects so many other things it’s not a bad place to start, but you need a lot more detail before you can rely on the numbers. But the decisions are based on things only you can know, and many of those things can’t be known until you start building.

Take your power requirement as an example: it is very high and appears to be a problem. But are you really going to run each LED at full power? Then, for the effects you select, how many LEDs will typically be on at the same time? Will it be possible to power the strips from multiple connections (that will enable a thinner wire, at the cost of more wires)? Do you really need that density of LEDs in the strip (would 90, 60 or even 30 per metre be adequate)? All those things have an effect on the power you need, the battery packs that might be suitable, how many battery packs would be a good arrangement, how they will be wired, and so on.

Only you can answer those questions, and you can only do it by setting up an example: creating a strip, programming it with a suitable effect, sticking the strip onto an old hoody, slipping the battery and MCU into a pocket, setting up some lighting and checking the result in a mirror. When you have the result you are looking for you will know what is needed to build it into a full suit and you can start looking for components. Or, possibly, adjusting the design to make it possible.

76

Hi Michael.
I have been staying away from this thread for pretty obvious reasons. One is I don’t like wasting my time producing information which is obviously not read, or if read is not absorbed.

All you have said is well put, particularly about the designing.

I will throw in on last bit of info. There are is vast amount of information available on wire sizing and detailed information on the good old web. If Darkmaster typed “wire sizing chart” in his search bar he will find more info than he thought possible. Not rocket science.

As for “all” wire being copper another bit of trivia. Copper (which we all know about), Iron wire, aluminium wire and silver (or a silver alloy). These are just the ones I know about.
Cheers Bob

77

This is starting to go back in the direction of “its impossible to pull off”. I think even with all the video evidence something like this is possible, you’re pointing towards the notion that its unlikely to power so many LEDs and the fact that I tried proposing to reduce LED brightness down to 75-80%, it’s seems the problem isn’t just how bright they will be when running the animation sequences, but how many LEDs would be running at any given timeframe. Then the issues with voltage drop as a reason not to use powerbanks unless a KeepAlive module is used is as redundant as to 18650 batteries also not enough to supply the necessary power for the strips.

I guess if the videos don’t convince you at all that all LEDs have to be powered at some intervals plus they can’t have obvious/many gaps in between each diode then I think you’re basing all this hypothesis on something you think applies to simple yet more commonly done applications and couldn’t care less about the most crucial details. This is not a smart home project nor a decor related installation. I think despite watching the video(s) provided along with proof others have pulled it off you don’t seem convinced that it’s genuine and leave out the important details/in-depth observations as a means to keep marking this project as “too technically unrealistic”. I realised that despite all the evidence there is, no matter what proposals I give out to match the technical specs to ensure little to no issues in delivering more than the adequate amount of Watts (or mAh) it seems that every LED strip’s parameters only can be applied within a home decor project and not a lesser-practiced project that also happens to be a reality.

So I guess if others say 18650 batteries plus voltage booster isn’t enough and powerbanks at 20k mAh isn’t going to be adequate and Lipo batteries are uncertain as a substitute, then nothing else except power outlets would be on the safer side to power those LED strips.

I also take it as I’m the only one noticing even the smallest details while others are barely even watching the videos in-depth at all. There’s a difference if you are even paying attention.

78

@Michael99645 if Robert wants to avoid this thread while coming back then he is clearly contradicting himself. He should have stopped commenting despite his resentment and I never tried to directly seek his help if he believes he can’t give any useful info or workarounds.

Its not rocket science, but then again, for someone as a BEGINNER, that’s straight down dismissal and rhetorically dwelling on practicality rather than possibilities shows a lack of use for ideas but more of the problems of “too much/not enough power, too many LEDs, portable components not able to deliver said parameters/results, etc” and asking EXACTLY how to match the parameters he goes back to his statements to reiterate “it’s not enough” approach.

Please do me the favour if he decides to chirp in despite making it clear he wants nothing to do with me then he needs to stay somewhere else, and not trying to complain about someone who’s also not interested in disputing with him, yet here we are.

79

OK… lets try a different approach step by step.
Step 1. Let assume (for the moment) they LEDs to be used are WS2812b (as we have been talking about).
Using the following image, can you correct as needed (ignoring hands/feet and head for the moment), and update to include how many leds you see on each segment.

Note: I have assumed a separated segment at pivot points. If you wish to have the stipe continue around shoulders, knees and elbows than update the image.
What I would like to see come back is the number of continuous strips then for each strip
For each strip, add…

  • the density of each strip.
  • the length of each strip/segment
  • how many leds you think that strip density will give you

the common density’s I have see are
30 per meter, 60 per meter and 144 per meter. If you have already found a different density you want to use, please add and update.

At this stage, I dont care about any other details, I just want to see the exact number of leds per segment, how many segments and locations of those segments (with the stated exclusions and inclusions)

So, we are NOT talking about power, controllers, software, patterns or anything else.

image
image320×678 66.1 KB

81

On the contrary - I am trying to describe the process by which you will be able to pull it off, albeit possibly not in exactly the form that you envisage now. If you don’t start on it until you have finalised all the details for the perfect result then you will never start. And if you start with less than the perfect concept you will likely find ways to improve it as you go.

That’s two parts of the puzzle - there are others.

Voltage drop is not the reason not to use power banks. The reason not to use them is that they can turn off unexpectedly unless you add a ‘keep alive’ module. A simple battery bank that doesn’t turn off unexpectedly doesn’t need the extra module, and there are versions available that provide both 5V and 3.3V. Also a battery bank that allows replacement of the batteries may be preferable to trying to recharge between performances.

No idea where you got that from. 18650s are very widely used for powering LED strips.

I don’t look at videos - sorry. I’m not basing ‘all this hypothesis’ on anything other than the commonly used procedure for getting an innovative and complex project up and running. I have used some examples that might or might not be relevant, but they are examples.

Enough for what? That’s the problem in the discussion (but note it likely wouldn’t be one 18650 plus a booster - more likely to be two plus a regulator). It is absolutely enough to power one strip - but the question is how long is the strip, how many LEDs per metre, what sort of effect and what brightness? If you say all LEDs lit at maximum brightness and 144 chips per metre then the strip will be short. If you relax any of those conditions the strip will be longer. Try it and see.

See above.

Not sure what that is about. LiPo batteries (like an 18650) give you the best power density currently available.

Yes. Because the smallest details only become relevant after you have determined the scope of the project. When you know exactly what you are going to do then noticing the smallest details might become important and others will be determined for you by the design decisions.