I've been meaning to test this out for ages so I'm actually getting on with building a testboard,

Not sure if their is a better place to ask my questions, I'm quite uneducated on a lot of electrical engineering so excuse my nievity.

I've got an array of individually programmable LEDs 8×32, auduino comparable controller, a little black Chinese wonderbox for pumping static electricity and I've spun up 256 little 8mm×3mm coreless coils.. the idea is to mount them on an acrylic sheet, connect them to some transistors an onto their appropriate colour LED connections, charge the surface with static electricity an then program polarity checkered rippling magnetic flux over the acrylic, so as to move the static electricity over it in waves (re-cycled from one end to the other via a shielded wire) at high frequency to attempt to move air flow more efficiently than silly ion thrusters.. naturally I'll be testing out different surfaces on top of the acrylic after that.

My question is this.. the LEDs seem to have 4 connections, one is positive and the other three negative for each colour I presume (I'll test them soon), to get my coils to magnetize at either polarity according to the relevant red or blue colour, I've done a little sketch that makes me believe 3 transistors per coil should work, but if one of you can suggest a way with less components I'd be very grateful, also, I suspect each coil should probably have a resistor so as each coil doesn't get power greedy.. I'm not sure what type would be appropriate.

The LED array is 5V but I've not checked what's going into the LEDs themselves yet.

 

If they are RGB LEDs they will have one common anode, and individual cathodes for red, green and blue. Or they may have one common cathode and individual anodes for red, green and blue. The datasheet should say “common anode” or “common cathode”.

Good luck with your project!

 

Plus each diode will need a current limiting resistor. 20mA is commonly used as MAX but can be less. Diodes are current driven devices, not voltage. Sounds like you may be in excess of the available number of digital Arduino pins... For a single board that is, you can master/slave them to increase the pin count. Plus use their analog output pins to drive a transistor switch as well. Something like a 2N7000 logic input FET to cut down on the current draw from the Arduino. And welcome to AAC!

 

Plus each diode will need a current limiting resistor. 20mA is commonly used as MAX but can be less. Diodes are current driven devices, not voltage. Sounds like you may be in excess of the available number of digital Arduino pins... For a single board that is, you can master/slave them to increase the pin count. Plus use their analog output pins to drive a transistor switch as well. Something like a 2N7000 logic input FET to cut down on the current draw from the Arduino. And welcome to AAC!

Can't just one resistor on the consistent polor terminal work?
That's a lot of soldering.

My LED array is a WS 2812B an only has 3 pins, so running out of data pins on the arduino isn't a problem, though it does potentially throttle the speed of it, the LED array is only good for about 800khz, an the auduino thing I've got only runs upto 160khz but upgrading that isn't a problem, i was also wondering if a better LED array type might save me a lot of hassle upgrading now,

Any good suggestions welcome, and thanks.

 

Can't just one resistor on the consistent polor terminal work?

Yes, and running the LEDs in parallel. Sounds like you have a good handle on it without seeing your schematic to check.

 

If your LED array is a WS 2812B you shouldn't need any resistors, are you sure you know how to drive such arrays?

That type of addressable LEDs have no common cathode or common anodes. (not externally available)

Just positive, negative and data.

I suggest you take the tutorials available at sites like Adafruit.

 

Yes, and running the LEDs in parallel. Sounds like you have a good handle on it without seeing your schematic to check.

Since you’re using programmable LEDs (such as a WS2812b), some of these suggestions don’t apply. Ie, you don’t have to worry about wiring the LEDs in parallel. As ElectricSpidey has mentioned you only need positive, negative (separate from the Arduino) and one GPIO pin for the data. The power supply will need to supply more than 13A (256 LEDs * 50mA per).

 

When he said "array" I envisioned something like a 3D Cube, not a long strand of LEDs... No experience decorating bars and party rooms...

 

If your LED array is a WS 2812B you shouldn't need any resistors, are you sure you know how to drive such arrays?

That type of addressable LEDs have no common cathode or common anodes. (not externally available)

Just positive, negative and data.

I suggest you take the tutorials available at sites like Adafruit.

Every 1cm square of it very clearly has 4 exposed component wires going directly into an RGB LED, I've yet to fire it up an test the thing with a multimeter but I'll let you know what I find, running the coil in parallel with the LED (through 3 transisters and a resister), is part of the plan so far, I've made assumptions there but I'll find out soon enough.

 

Those 4 wires should be positive, negative, data in and data out.

Feature (adafruit.com)

 

When he said "array" I envisioned something like a 3D Cube, not a long strand of LEDs... No experience decorating bars and party rooms...

lol, That's about the only experience I've had in this field, something's you can never truly know until you try, it's exciting to have the time an inspiration to have a go.

 

Those 4 wires should be positive, negative, data in and data out.

Feature (adafruit.com)

Oooh shit.
I hope your wrong, I'll figure something out if not though, probably by getting a different led array I hope.

 

Does you array look like this?

 

Does you array look like this?

View attachment 302659

That's the one.. just had a closer look an it's obvious your right, That's data an power, not direct to LED, I wonder if getting a weld on the three copper plates in the eye of the LED, would be usable for the coil circuit.

 

Those 4 wires should be positive, negative, data in and data out.

Feature (adafruit.com)

And all of the LEDs should be connected in a serial fashion. That is except for the first LED, each succeeding LED’s data_in connects to the previous LED’s data_out. Thus you only need a single GPIO pin from the Arduino to connect to the first LED’s data_in.

 

The array that was pictured in ElectricSpidey’s post takes care of the data_out to data_in connections. So you are left with power +- and the data line.

 

I think I'll get some 3mm flat head drill bit, lend use of a big drill with controllable depth an square footing, then try to get to the conductive layer under the transparent lense on top, then repeat at set depth for the rest of them.. their are 4 copper plates in the eye so my original wiring could still work.

 

I think I'll get some 3mm flat head drill bit, lend use of a big drill with controllable depth an square footing, then try to get to the conductive layer under the transparent lense on top, then repeat at set depth for the rest of them.. their are 4 copper plates in the eye so my original wiring could still work.

You seem to have a misunderstanding of how programmable LEDs work. To me, you seem to imply that you see them as similar to a regular LED. Particularly when you say things like:

.. the LEDs seem to have 4 connections, one is positive and the other three negative for each colour

Nope. Each programmable LED has a built-in LED, resistor and microcontroller.

Besides the power connections, each LED has a data bus, on which it receives commands from another microcontroller. These commands contain an address, a color and a brightness. A software library provides simple functions to control each LED.

I don’t see the benefit in taking all these components apart and turning your programmable LEDs to dumb LEDs.

 

You seem to have a misunderstanding of how programmable LEDs work. To me, you seem to imply that you see them as similar to a regular LED. Particularly when you say things like:

Nope. Each programmable LED has a built-in LED, resistor and microcontroller.

Besides the power connections, each LED has a data bus, on which it receives commands from another microcontroller. These commands contain an address, a color and a brightness. A software library provides simple functions to control each LED.

I don’t see the benefit in taking all these components apart and turning your programmable LEDs to dumb LEDs.

Look, I want the programmable parts, to control an power coreless electromagnets in either polarity, it would have been nice to keep some LED functionality but it doesn't really matter, the 4 connectors under the lense are obviously part of the RGB LED on top of the microcontroller, so I'll just drill to them.

I was obviously wrong earlier, I learn from mistakes I feel no social shame for them if I learn an harm no one undeserving.

 

I've been mulling over my options an brought a teensy 4.1 as speed may be key.

I've changed the shape of the design, you see originally I was gona test on a long flat rectangle but I reckon my best chance to make this electrostatic peristalsis generate enough thrust is to keep the electrostatic charge pinned in by magnetizm all of the time,

So I've opted for cylindrical, so the static electricity doesn't pour out the sides.

An it a occured to me, the prime pattern I want to use.. positive, off, off, negative, negative, off, off, positive, positive, off, off.. etc..

With some deviations at either end, only really need 32 channels an 3 settings for each, positive, off an negative.

So I might be able to just hook it almost straight to the teensy, negating the need for butchering the individually addressable led array,

The teensy says it's outputs can vary in charge by 7 steps, so I've been wondering if I could use that to determine the mode of the electromagnets on that level,

Probably gona need at least one mosfet per level so the electromagnets are powered by a heavier supply but I don't know much else of what this circuit would entail atm, (current level sensitive transistors?)

Any tips welcome!