Hi all, I’m trying to figure out how to do an extremely complicated project that has left me baffled, especially as a new user to WS2812B strips and Arduino/Raspberry Pi programming. Attached below is the link to the project, so hopefully someone here has ideas or clues as to how to choreograph complex light-up animation sequences, how many portable power sources and wire types to use, how to use little to no bulky components as controllers, and if splitter cables can help in any way to equally distribute adequate voltage and manage animation timing.
Is this a paid project ?
The very first thing to do would be to work out how many leds will be used. Next will be to define what the maximum number of leds will be on at the same time and what “color/brightness”. Once those two things are known, the power supply needs can be calculated.
e.g. lets say each R G B led when on generates white by having each at full power, that will be about 36mA for each so 108mA per led. So if you then had say 200 Leds on white, that’s 200 * 108mA = 21,600mA or 21 Amps.
So, as you can see the power needs can get very high.
Once you know that, you can then move onto both how to power it and how to control it.
Power needs will come down to how long it needs to run for at power power levels, but you may not know that. e.g. worst case for the 200 white as above for 5 mins will be about 21 Amps for 5mins (1/12hrs) => 21/12 = approx 2 Amp Hours (but must be able to dump 21 Amps as needed.)
Ok, so what we should be seeing from above, is that we need to clearly define what will happen. As this can be very hard, one way to work the problem would be to bench build the idea and power with a high power bench top supply that can support max amps, and to run it an measure the power consumption and peeks.
How to control it will come down to needs as well. this will be partly limited to the controller user (uC), how many leds in a string and how fast you need them to update.
from memory they led strip works by clocking out the RGB values one after the other, then pause. the pause needs to be long enough to activate the leds to switch out the new color. (I would need to look up the data sheet for actual times and do the math)
As to syncing to music, this can be as simple as getting the timing right. e.g. at the start “press go” and led it run its sequence. To track the music will be harder as you will need to find a way to ensure the trigger is correct. e.g. FTT sound input and if profile match’s trigger, change…
But how does it get the source sound? from a mic ? if so how to you ensure no false positives from other noise.
Or you can do what they where talking about in the clip, ie. someone sends the commands to it. Keep in mind that simple triggers will just be next “sequence”, but every wireless protocol will have pros/cons. e.g. wifi is built into some small devices, but creates overhead and latency (sometimes out of your control)
So the above is some ideas to get you started… Im sure others will have ideas as well.
BTW: Please be safe with this project, things can get hot and large amounts of power needs to be made safe as well.
1 Like
The thing is, it would be much easier to create a preset sequence of animations that matches the video clip. Reacting to music would meant the animation sequence could be random instead of going according to one shown in the video. Also, since its not just changing colours, its about having a combination of other complex effects such as fading/glowing/shimmering, movement-like animations, wipe-effects, etc. And it would be extremely difficult to arrange and set up with code alone, especially for beginners. A GUI-based software could enable a beginner like me to visualise how the animation sequence can play out, how intense the brightness is an different intervals, transistion effects, etc. and accurately program it to deliver the expected results.
Another issue would be the flexibility of the LED strips and the need to use splitter cables and other ways to connect to powerbanks, power converters, etc. controllers should be necessary provided it can run animation sequence presets, but without using larger, bulkier or heavier equipment and small enough to fit in pockets/compartments. For the glove, 5 smaller LED strips would be too much to connect and run given a splitter cable can only connect up to a max of 4 which would probably resort to using more than 2 splitter cables and more power source, cables and controllers to hook up to.
And finally, to answer your question: no, it isn’t a paid project. But it is paid in a way become of the budget spending to get all those components. It’s just a personal project that I could find uses for it in the future, since what I’m building is normally beyond expectations and is not for sale/for clients.
Hi
As Michael pointed out your power requirements could be substantial and by the very nature of such equipment could be bulky and heavy. I am afraid I don’t think there is any way around this, If you need Amps they are heavy unless someone comes up with a lightweight small battery in the interim.
Regarding all the other problems and brick walls you mention it sounds like you are talking yourself out of this project before you start.
Don’t do that.
I wish you every success
Cheers Bob
I came here to seek advice and guidance on achieving similar results, not to offend others or as you said: talking myself out of the project, which is not the case.
I merely brought up suggestions to see if Michael would agree/disagree based on what he knows, but if people are going to dismiss me instead of assisting me (mind you, I already mentioned I’m a beginner so that would clear things up) and have nothing good to say then I’ll keep that in mind that this is how you lose a member of this community forum and this is the best help I could get from you. Thank you for your time.
Try to give others a chance to comment. Alot of people here are not paid members from.core, but happy to try to help when we have time.
The project will seem big and complicated just becase there are lots of parts, but once tou break it down, it will turn into smaller managable sections.
The reason i asked about it being a paid job is very often there will be more specifc requirments and expectations that need to be meet, as well as a short deadlines, as such requires action.
I would try to firm up the requirments and drill down from there.eg what are the segments, what effects do you want on those segments, then workout the best way to do that.
On the build side, i would tend to start with one big string of leds, cut and wire links to create pivot points, but still on string.
Im code that is then viewed as on big array and clocked out as such.
Then for the segment effects, you could create pointers to that array where the segment starts, or store the offsets… at this point you can now mosly treat each segment as its own mini string…
Eg
Update segment 1
Update segment 2
…
Update segment x
Clock out.
I would only branch out if you are 100% sure everything after the branch is 100% the same.
Eg. Left and right side, each finger
All this then comes.back to the physical design phase to allow you to do what you want… sketchs and diagrams should help.
If the number of leds is withing the clock out limits, then you can get away with one controller like the esp32 or pi pico. For power there will be guidelines as to who many leds should be powered in one segment. This is not a hard rule as such, but if too many leds are on full current drain, they will tend to dim towards the end of that power run.
I would start with each segemnt, with its own power run, then trim back. Eg lower leg left, upper leg left lower leg right upper leg right and so on… but the number of leds could be small enough such that it will be left leg, right leg.
I hope that makes sense.
The way i think about these leds is simple… we are.just telling each led what rgb values to have at this moment in time. If these are in an array, we can work on sections of that array, so its not one controller per segment.
… but if this then takes to.long for the effect your looking for, then you can add a 2nd controller. For some guide, i built a 300 led display all running as a single.string. the display then had that broken into bars with each bar doing its own thing.
I thing wled supports 4 stings of 1000 leds each for a good result. Just may need to add some power taps.
Hope that makes sense. As you firm up what you want and need, the “what” do you need to do etc will just fall into place. Past that point it just becomes.code that can be worked on until working as needed.
As a project i would keep some $$ for things that pop up. Eg my controller cant keep up, so i need to split and get a 2nd controller.
Esp32 has some nice hardware level dma features where it will just clock out the array for you withing rtos getting in yhe way.
1 Like
This sounds like an interesting project! I did a quick count on the sample suit and get about 400 leds… (100 each sock, 20 in the glove, 50 each side @60leds/m), I get about 24A max, or 120W for fullpower power white. But I doubt that’s ever going to be the case. The song is 5 mins, so lets say 10 mins max or ~20 W/hrs? Using flat cells as used in a laptop battery which can possibly deliver the necessary current without getting too hot, I think the power can be supplied fairly easily.
I’d like to see each “section” controlled by their own device and all synchronised with the music or lighting controller, over a wireless connection (not sure which… BLE?).
If I get some spare time I’ll throw some bits together to see what’s possible.
Hi Darkmaster
My sincere apologies. I really did not intend to offend you and believe me you are not offending any others. I, Michael and many others will provide all the help we can.
I don’t think for one moment that is the case. That is just the way I read it. Probably me in old age. I was just a bit concerned about how you viewed power as being able to fit into pockets etc. It is an unfortunate fact that as power requirements go up size and weight usually goes up with it and thus portability goes down.
Your LEDs are going to be the most power hungry item. You will have to document just how many you need and where you need them. Work out the power required then I would add 50% for unforeseen items and allow a bit of overhead. THEN worry about how you are going to provide and distribute this power. Don’t start in the middle of the project and work both ways. I don’t believe that ever works in a satisfactory manner. And after all satisfaction in a job well done is most of the aim here. The data lines to drive these LEDs can be a completely different exercise and do not need to have anything to do with the power distribution.
Please read and take note of Michael’s input. He is talking a lot of sense. Particularly re breaking the whole up into smaller manageable bits then putting the complete thing together at the end of the day.
Once again I wish you every success
Cheers Bob
Thanks for the input. I was wondering with the case of power supply, would it be possible for powerbanks of high capacity of voltage input and output, small enough to fit into the clothing, and for controllers those that run on bluetooth/app control to act as data input/output?.. I’m not sure if on the side of variety of animation sequences would raspberry Pi or Arduino be sufficient and easy enough to do the job. It’s easy to visualise/picture the animation sequences but to do it all in code would make it all the more complicated. Not sure if there are graphics/visuals based controls/software that can make it all the more easier? Coding each diode could actually be more time consuming and more confusing, just needed to see if there are other alternatives.
I remembered mentioning before that instead of making it reactive to the music, that creating a preset animation sequence that matches the music’s timing would be much easier, because if it was reactive to the music, any random animation sequence would occur and may not match that of the video.
Perhaps not having the LEDs at 100% brightness, but more around 75% would reduce the excessive power consumption at least to some extent? As long as it can shine bright enough in a dark setting, that’s more than enough.
Lastly, for the LED strips on the pants, I was concerned that it might not be flexible around the level of the kneecaps, perhaps creating a small separation inbetween like a small strip disjointed in-between two longer lengthed strips would allow more flexibility?
Thanks for the enthusiasm and support! In regards to the socks, would LED panels suffice given that it’s more of a wraparound structure? And for the glove, I would be looking at more of a structure of 5 LED strips for each finger (covering both sides of the hand), possibly connected to splitter cables? Also instead of having the LEDs at 100% brightness, would 75-80% brightness reduce the heat and strain the LEDs would take, provided they can shine bright enough in a dark setting? I’m also quite curious as for setting up the animation/transition sequences, would Raspberry Pi or Arduino make the job more easier? I’m terrible at coding and would presume it to be more time-consuming and complicated, is there any phone apps or PC software that can provide graphics-based visual display of what the animations and transitions effects could look when setup and to be able to edit by directional movements, brightness intensity, frame-by-frame, etc.? And would you agree that there are powerbanks out there that can provide enough power to run the LEDs at least a min. of 10 mins uninterrupted use?
I tend to agree with most of that… I was just picturing a strip/segment of leds on the bottom part of the leg, then some link wires to get past the knee, then onto the next segment.
Note: Anywhere you solder a wire can become a weak point. Wires soldered to things tend to snap off over time and the solder can wick up the wire a bit; Since you need it to flex you “may” be able to lightly fold it such that it moves (… thinking… maybe) then have those leds “off” all the time… OR if you cut and join with link wires then ideally you will need to strain relief the wire and led strip.
As to control. I was thinking more the master controller would say “segment 1 preform action fade” sort of thing, but the local controller would know how to do the fade. A lot of this will just be patterns with some nice math.
As to software to do things, not sure on that. there are projects like wled with predefined things that may or may not meet your needs. But I suspect this project may be a bit to custom. I have seen sold kits that have an “grid” of leds, or you set them up and “scan in the grid” over which you display patterns or images; but again this seems not to be that sort of project.
Since this is your project and you are the one doing the work, I would tend to stick to what you know… eg. a Raspberry Pi will have bigger power needs then say an ESP32 or Pi Pico; both of which have Arduino frame works.
What might be worth doing is having a look at wled, even if just to get some ideas.
I would then just buy a small group of LEDs and get them working off your controller of choice, then play with those.
e.g.
Project 1: Get all LEDs to turn on blue
Project 2: Get one LED at a time to turn on e.g. led 1, then 2 … X
Project 3: Play with a color change from say blue to green
i.e. work with some effect primitives. I think once you get the basics working, you will find doing more complex things start to fall into place; and don’t forget there is lots of git hub projects where you can see the code they used to do something.
If we look at this the other way around… Im sure “someone” can do exactly what you have in your mind… Like any good hobby projects this will come with lots of learning; but also will come with rewards as you get each bit working.
One more thing while Im here… Check out “christmas led displays” I cant remember the name, but their is a sequencer and kits that allow you to plot led changes etc and to music.
Welcome to the forum!
I would recommend looking at LEDFx paired with WLED which will give you a GUI for controlling addressable LEDs with little to no coding required.
In the past for wearable LED projects I’ve used Battery Holder - 2x18650 (wire leads) and a voltage buck set to 5V as a means of portable power. The batteries with the voltage regulator end up being around 75mm x 40mm x 30mm give or take depending on where you put the regulator and weighs about 100g, you would also need an ESP device to run it, something like our FireBeetle Board ESP32-E (Arduino Compatible) will work, although any ESP32-WROOM will work, ESP32-C3, ESP32-S2 and ESP32-S3 may also work but certain features may not be available.
1 Like
Hi Darkmaster
Me again.
The ma/hr rating might seem enough for the 10 min or so that you mention but the limit would the Amperage they can supply at any one time. Some might be 2A but others will be in the 500mA range. Also be aware that if all the leds are off fr a short time and the current drops below 100mA or so these power banks will usually shut down.
I would remind you that these power “banks” are not power “supplies”. They are designed to be a back up charging system for other devices like phones etc. The lower current is how they detect “eb\nd of charge” and switch off. I know some people use them as “supplies” but I am not a great fan of this.
Cheers Bob
Hi @Dan, thanks for the welcome!
I was just wondering if any powerbanks with USB input/output and those with 5V/12V input/output could be used in place of 2x18650 (wire leads) Battery Holder, and whether or not bluetooth/wifi controllers the size of a USB stick (AliExpress has some examples) can be used in place of FireBeetle Board ESP32-E (Arduino Compatible) unless I am not using any apps with preset animation sequences?
You could look at using a battery bank, you would need to break out the power that is supplied by the battery bank as controllers generally won’t be able to handle any significant amount of current draw. You would also possibly run into issues with the battery bank shutting down if enough current isn’t being drawn, we do though sell USB-C Power Bank KeepAlive - Assembled. If you’re wanting small form factor we stock Seeed Studio XIAO ESP32-C3 (photo attached) which is about as small as you can go while still having access to the needed GPIO pins for running the data lines of the LED strips.
2 Likes
Hi @Robert93820
Thanks for the advice. I take it that battery packs running on AA or AAA batteries can provide an uninterrupted current flow if the strips connected to it are switched on, especially if the LEDs themselves aren’t for a brief moment. Would you also consider that measuring by voltage DC as compared to Amperes/mAh be more easier in determining the necessary input/output of the power source and even if power boost converters are still needed regardless? Also, would it come down to other factors such as wire/cable types (thickness, coating, etc.), silicon protective sleeves for the LED strips to prevent internal/external damage, anti-strain measures on the strips and wirings, certain environmental temperatures that is tolerable for the LED strips and power source to not be susceptible to overheat, overfreeze, etc.
Sorry to add another question, but would the LED strip controllers listed here in the links below work with WLED and LEDFx? Thought there’s normally a module needed to relay data communication input/feedback between the software/app and the LED strips themselves…
Thanks for the assistance and help anyways!
Hi Darkmaster
Not really useable. These cells have a very limited capacity and would probably run 5 or 6 LEDs for a short time. I think you are talking hundreds of LEDs here.
Voltage = Volts, Current =Amps/ milliAmps, Power = Watts, resistance = Ohms.
They are all related and each has an influence on the rest. For instance you need Voltage (as a pump) to push current (as in water) through a resistance (as a pipe, the resistance represented by a partial obstruction). The whole process is represented by Power (Watts) this is dissipated by the Resistance as heat and is lost. A short Circuit could be likened to the pipe with no obstruction allowing full water flow (usually a disaster electrically).
For comparison or say voltage conversion (every time a voltage conversion takes place there is a percentage loss) it is easier to work in watts then convert back at the end.
This is about as simple an explanation as I can make. If you liken all this to water as I have done I think understanding is a bit easier.
A head of water developing pressure as in your water mains or even a tank regard as a battery
A pipe could be likened to a length of wire. The water flow is current.
A tap is like a resistor by restricting water flow. A turbine as in a power house to drive a generator is another restriction but this time the water is doing work, like lighting up your LEDs. Thus the energy dissipated by the water turning the generator has been converted into useful work. This is usually expressed in Watts.
I hope all this helps you understand.
Cheers Bob
PS I have an idea that you might be able to work with and adapt regarding powering and driving your LED strings. Will try to produce something and post tomorrow.
Just a quick note:
Perhaps not having the LEDs at 100% brightness, but more around 75% would reduce the excessive power consumption at least to some extent? As long as it can shine bright enough in a dark setting, that’s more than enough.
These leds dont have a brightness as such. They are RGB (Red Green Blue) where each value is 0-255.
So 0,0,0 would be all off = black = 0 current needed
255,255,255 would be all on = white max current 3 * 36mA
255,0,0 would be pure Red = 1 led on so 36mA
So you can kinda reduce the brightness, at the expense of exact colors.
So brightness is more a function of dropping all 3 values by the same amount, but if already too low on one color, the actual color may change a little as the ratio of RGB will change.
If you read up on the correct way to adjust the 3 colors for brightness it can get complex as RGB does not really support it; but there are also hacks that mostly work.
Also I have seen different data sheets stating different current needs. So I went with the higher end to be safe, but the actual current could be lower; so some testing may be in order.
e.g. This datasheet
https://www.mouser.com/pdfDocs/WS2812B-2020_V10_EN_181106150240761.pdf?srsltid=AfmBOoqS42WoFeJUQ8l5TcsGtV8WAsmRCSOS_dpcRik9tkSAw5jMbdl6
states “16mA operating current per channel.”
So thats half what I have seen elsewhere.
This states 1.1 A per 30 leds (all on) so about 36mA for all on per led.
Those controllers are for a 4 wire non-addressable LED strip. Using WLED the ESP32 would control the LEDs and also the can act as a wifi access point (or connect to an established network), from there you can use a web interface to control the LEDs without the need for any sort of external apps.
