Hey Guys,
I’m an artist/photographer with rather basic knowledge of electronics and in need of a non ai powered hand with a project.
I’m wanting to integrate 3v LED filament “noodles” into a sculpture. I’m wanting to power it via usb 5v and run 1-2 Filaments 30cm long each. I would like to be able to PWM them to dim/flicker. Main goal is powering them to work and not fry in the first place, stretch goal would be to have them switch to a reactive mode to allow them to respond to environment sounds.
So far i think i need an ESP32 chip that i can load something like WLED onto (though i’m unsure how WLED handles filament strips) and also a MOSFET controller (though i’m not entirely clear on why). I know i’ll need something to minimise the current draw but unsure what or where to put it in the circuit/series. I also think i probably need a step down or something to regulate the power draw. Currently i have 3x noodles hooked up to a variable power supply and at 2.5v they are drawing 125ma, if i go to 3v they suck down 1.1A which sounds like a lot. I would like the electronics to be as miniaturised as possible as i will be attempting to integrate them into the sculpture which is 3d printed so max 256x256x256. Would also like to give them as long a service life as possible so any tips of how to do that (under powering etc) would be much appreciated.
Finally has anyone had experience with using the single ended filaments? The tiny ribbon cables seem VERY close together to get a + and - to without shorting and they seem to small for heatshrink.
Any help very much appreciated
Tom
2 Likes
Hi Tom
Welcome
Lots more info please
What exactly are these “noodles” and how are the LEDs connected in them.
Normally LED current is controlled by a series resistor but to calculate this you have to know the LED configuration and the LED forward voltage drop under required conditions.
If you connected them straight across the power supply without any form of current limiting you are lucky to still have them. I think “fried” is the current general usage term.
The variable power supply could have a current limiting feature but there is a bit of a pitfall here. As the current limiter is adjusted downward (ie for less current) when it goes into current limit the output voltage displayed will decrease. This is the normal way to limit current, reduce the output voltage until the limit value is reached. This will give you no idea of the resistor value required but should indicate the LED forward voltage drop @ the applied current. Knowing the prime voltage value should then enable the current limiting resistor value to be calculated.
Cheers Bob
Thanks Robert!
These are the LEDs,
From what i’ve seen they are all in series and can pull a lot of current if not restricted.
The power supply i’m using now is just for testing them and (possibly very stupidly) i just connected all 3 noodles into the same croc clips, i’m not allowing the voltage to go above 3v so haven’t cooked them yet but i’m sure the final setup needs to be better considered than this.
Any help much appreciated!
Tom
1 Like
Hi Tom,
Welcome to the forum!!
Bob’s caution is spot on here — I would not run those LED’s straight from 5 V or even a bench supply without deliberate current limiting.
For a small sculpture, I’d keep it simple: use a 5 V USB supply, a MOSFET board and current limiting resistor
The MOSFET is there because the ESP32 pin can provide a control signal, but it should not directly carry the power to turn the LEDs on and off (it will destroy the ESP32).
A rough guide from your numbers: if one filament is happy around 2.5–2.7 V at ~100–150 mA, then from 5 V you’d be dropping roughly 2.3–2.5 V in the resistor. That lands somewhere around:
R = V / I
R ≈ 2.5V / 0.125 = 20 ohms
Power in the resistor would be about:
P = V × I
P ≈ 2.5 × 0.125 = 0.31 W
(Check out this guide for the calcs)
So I’d be looking at about 18–22 Ω, 1 W as a conservative starting point per filament, then test at the dimness/temperature you want. One resistor per filament, not one resistor shared between several in parallel.
If long service life matters, under-run them a bit. Barely below the “very bright” point usually helps a lot with heat and lifespan.
For sound-reactive behaviour, an ESP32 can do it, but I probably wouldn’t start with WLED here, so a small ESP32 PWM sketch is the cleaner path if you want to learn from scratch and customise to your hearts content (AI is good here).
To lock in numbers and more specific part recommendations it would be great to get the links to the LED filaments youre after/already have.
I’d also be keen to see those single ended filaments!
Liam
1 Like
Hi Tom,
Sorry, I didnt see your reply come through!
Internally they would rather be connected in parallel (a single LED voltage drop is usually above 2V, and less than 5V) - the amount of current its drawing can increase as there are more LEDs.
Soldering to those ends should work a treat, just make sure to label which one is positive and negative (some small heatshrink ought to cover up the joints perfectly once soldered to - seperately to ensure they dont short like you mentioned
Liam
1 Like
Hi Tom
I note that from your link the recommended operating point to start experimenting is 50mA (0.05A)
That is where I would start.
So using a 5V supply and a quoted 2.6V LED forward voltage drop the resistor would be 2.4 / .05 = 48Ω. The nearest preferred value would be 47Ω, this is OK.
This is one resistor per LED noodle, Total 3 resistors as Liam says.
1W resistors still good. You could get away with 0.5W but they might get warm.
Cheers Bob
Edited
“The nearest preferred value would be 47Ω”
“47Ω” was 4.7Ω. Corrected 24/05/26.
1 Like
