Looking for the smallest RGB LED Flexible Strip

A few thoughts.

I believe when people want a sparkly effect they use ground glass or similar materials (embedded, not where people can touch them obviously). Never tried it myself.

With high density the currents involved get to be a problem. If your LEDs each need 82mA and you place them on a 7mm pitch, that's 11.7A per metre. The tracks don't have to be able to handle that all the way down the strip, but at the start they do. And the connector has to be good for it too. You can help the situation by feeding power into both ends, but you might struggle to stop people using it by just driving one end. You'll need to see what effect the voltage drops have (if the addressable LEDs have an internal regulator for the logic and constant-current drive for the LEDs, perhaps it might not be too bad, but it will still compromise the control signals to some extent). For a long strip with hundreds of LEDs, you'll want to see what happens to the control signals as they pass from device to device.

If it were me, I think I'd prototype the whole strip with conventional PCBs and look at it electrically before spending money on the flexible part, though you're going to need to know the design rules for flexible boards to get it equivalent (I can't help with that - it's a long, long time since I last did a flexible board.)

Here's a simple simulation to illustrate the voltage drop problem. [Cut and paste was my friend here!] Each LED is crudely modelled by an 83mA current sink. It's a bit approximate but it gives an idea of how things pan out.

This is 100 devices running at 83mA (total current 8.3A), which is 0.7m of strip if you have a 7mm pitch. I've assumed a track width here of 2mm. This is exaggerated just for illustration, I'm not suggesting you use that as your track width. The resistance comes from 7mm length divided by cross sectional area of 35u thickness times 2mm width times conductivity of copper, which gives 1.68mOhms. As you can see, the last device gets just 3.6V when all the devices are full on and close to 5V when all the preceding ones are off, so it's going to be bouncing up and down like a yo-yo. Obviously the situation gets much better as you widen the tracks.

The other factor with pcb tracks and the one a layout person would normally focus on is the temperature rise from the heating effect of the current. For 11A and a 20C temperature rise you need about 6mm width for an outside track on FR4 - you might find it's very different for a flexible board (I don't know - it could be better or worse), so you need to find that out.

Hope that helps. Don't let it put you off - this is what engineering is all about (a problem is just a solution waiting to happen) - but you've set yourself quite a challenge here.

A few thoughts.

I believe when people want a sparkly effect they use ground glass or similar materials (embedded, not where people can touch them obviously). Never tried it myself.

With high density the currents involved get to be a problem. If your LEDs each need 82mA and you place them on a 7mm pitch, that's 11.7A per metre. The tracks don't have to be able to handle that all the way down the strip, but at the start they do. And the connector has to be good for it too. You can help the situation by feeding power into both ends, but you might struggle to stop people using it by just driving one end. You'll need to see what effect the voltage drops have (if the addressable LEDs have an internal regulator for the logic and constant-current drive for the LEDs, perhaps it might not be too bad, but it will still compromise the control signals to some extent). For a long strip with hundreds of LEDs, you'll want to see what happens to the control signals as they pass from device to device.

If it were me, I think I'd prototype the whole strip with conventional PCBs and look at it electrically before spending money on the flexible part, though you're going to need to know the design rules for flexible boards to get it equivalent (I can't help with that - it's a long, long time since I last did a flexible board.)

Here's a simple simulation to illustrate the voltage drop problem. [Cut and paste was my friend here!] Each LED is crudely modelled by an 83mA current sink. It's a bit approximate but it gives an idea of how things pan out.

This is 100 devices running at 83mA (total current 8.3A), which is 0.7m of strip if you have a 7mm pitch. I've assumed a track width here of 2mm. This is exaggerated just for illustration, I'm not suggesting you use that as your track width. The resistance comes from 7mm length divided by cross sectional area of 35u thickness times 2mm width times conductivity of copper, which gives 1.68mOhms. As you can see, the last device gets just 3.6V when all the devices are full on and close to 5V when all the preceding ones are off, so it's going to be bouncing up and down like a yo-yo. Obviously the situation gets much better as you widen the tracks.

The other factor with pcb tracks and the one a layout person would normally focus on is the temperature rise from the heating effect of the current. For 11A and a 20C temperature rise you need about 6mm width for an outside track on FR4 - you might find it's very different for a flexible board (I don't know - it could be better or worse), so you need to find that out.

Hope that helps. Don't let it put you off - this is what engineering is all about (a problem is just a solution waiting to happen) - but you've set yourself quite a challenge here.

Hello Jon!

Thanks for your thoughts on the matter, more knowledge is definitely more power.

The reason for having so many is because I want a high resolution animation, not just various glowing zones, I'd just like to make that distinction because I don't think my first statement was clear on that on my part. I think I could stand to make the strip less dense than the NeoPixel 144 now that I've seen it first hand (ordered one and played around with it), this has changed my perspective.

It must appear quite naive of me to just state my ideal solution as more LEDs instead of looking into material research but I am in no way a materials engineer beyond steels, fluids, carbons, and stoichiometry; the thought of creating my own fiber to display say 20 LEDs for example to be equally spaced apart and carry it to the brim of the hat and have it look decent is somewhat daunting, though I am open to new ideas. As someone that's an RGB maniac and that recently acquired the Corsair K95 RGB keyboard, RGB LED clarity in animation is everything to me, have you got any ideas as to making this more viable?

The prime example is say, the brim of clothing items, so shoes for example aren't a difficult task to accomplish, normal NeoPixel would otherwise work, perhaps even the lower density ones, maybe 60/m, but that works because of the size of the shoe's base. The rubber sole (or leather if you're classy) is thick enough to have a normal LED strip grip it, the standard 10mm width of the NeoPixel strip would work fine, however when you get to items like hats, it becomes more complex with the peak of a baseball/snap-back hat being about 2-3mm, the obvious choice without looking silly and having a 7mm overhang of electronics, is some form of optic, or a really really small LED strip that goes inside a tube that is stitched onto said hat (I figured a hat with a tube of LEDs is easier to wash instead of one that has a prism that must be aligned at both ends).

In regards to design rules, I don't know if you saw but in my first reply to dougw's I asked to be pointed in the direction of those design rules so I could go and learn them for when I get to the stage of making them flexible, it would indeed be foolish to spend a huge amount of money on professionally making them, though I appreciate and understand your statement on that with not having done it in a while.

What are your thoughts on a power supply? I don't mind carrying a 5V USB lipo battery pack on my belt (the kind used for charging tablets), say I'm working with about 13000mAh (as that's average size of the large ones) just for convenience. I think I will have to make acceptable losses for the density of the LEDs if I want them to be able to be on for extended amounts of time, though perhaps no longer than 3 hours, meaning we have shy of 4A to work with, and the power supply has a maximum draw current of about 2 or maybe (I doubt) 3 amps, I will mess around with some LED numbers to see what looks good with the NeoPixel (Disabling X out of X+1 LEDs), I will report back with what still looks good for my intentions. I will wait until the next replies before I start messing with materials for fibers (I really would like to avoid using them)

I don't think I'd be doing wristbands or anything else mounted to the body, perhaps some T-shirt decoration but I'd have to go away and design that first so if I do I'll try to keep it the same or less than the hat.

Thanks for the advice and diagrams you've given me, I can see what you mean about the resistance and voltage now, I appreciate everyone's help I've received so far, though I wont rest until I have the solution for this, I must have my RGB LEDs and I'm sure others feel the same way.

Best Regards

-John Webb

SK6812 SMD3535 RGB strips are available at 4mm width if that helps.

Doug