Esp32 wont connect after LED connection

Electronics
1

Hello

I bought the following from AliExpress to play about with.
imageimage

It has standard ARGB LEDs with 3 lines in - 3 lines out (see images) and should run on 5v.
I have connected this to an ESP32 which has WLED installed (0.14.4) but nothing happens. I know the ESP32 is working as i have a couple of LED arrays (WS2812b’s @ 8x8, 8x32 and 64x64) which all work fine.
Once the circular board power is attached, it refuses to load the WLED interface and give a “page took too long to respond” error, immediately returning once disconnected and refreshed.
I have checked my dodgy soldering and all seems fine. I have several of these boards and they are all doing the same thing so I am thinking it may be a faulty batch ?

I have re-installed everything from scratch and started over but am still getting the same issue

Is there a way i can easily test individual LEDs ?
Should there be any continuity between any of the pads on the pcb?

Any suggestions gratefully received

2

The image is far too low resolution to be useful. A link to the listing is better.

2 Likes
3

Also, its useful to provide pictures of your project showing connections.

2 Likes
4

Apologies, didnt realise the images were so bad.

image
image670×618 61.1 KB

image
image664×612 52.1 KB

My dodgy soldering.

20240826_1012_002_0.000
20240826_1012_002_0.0001920×1080 445 KB

these connect to an ESP32 with Dupont to JST connectors ( 5V, GND, Data Pin 16). i am doing nothing fancy with any of it, just trying to ensure they work.

I have connected the same ESP32 to all the LED arrays i have without issue, these have in turn had WLED connected to pin 16 and run 1d and 2d patterns to test.

I was unsure if any continuity should show with them being diodes and if it did whether it should show anything.

Is there a way to “light” the LEDs without a data signal, either individually or as a ring ?

1 Like
5

Thank you!! much clearer now :+1:

No, there is no ‘test mode’ for neopixels. But there is a standard test program.

There are other ‘obscure types’ of NeoPixel available, but the label on yours says it’s a WS2812 unit, the standard Neopixel, and all the libraries should ‘just work’.

Have you tried the standard ‘strandtest’ example that comes with the Arduino Neopixel library? This is the recommended (by Adafruit and others) way to test neopixels. Does this program work? I assume you are using Arduino/Cpp?

Even when fully lit a neopixel only draws 40mA, so the whole strip fully lit is 320mA, so I doubt if this is a power supply issue.

I think this is about voltage levels. You are running the NeoPixels at 5V but the control/data signal from the esp32 is only 3.3v.

This is normally OK (and how I run a pair of rings at home), but is not guaranteed to work:

Read this: Powering NeoPixels from 3.3V - Wrong but everyone does it? - Page 1

  • start at comment #6
  • the OP in that thread is being a bit of a grump, but is fundamentally right. You should have a 5V supply AND a 5V control signal.

So; one trick is to try running the strip’s VCC from the 3v3 supply, so the signal voltage matches vcc; this should allow you to see if the strip works. But as that thread makes clear, color/luminosity may be poor.

Even better; Do you have, or have access to, a 5V microcontroller? Eg an arduino Uno or similar (the atmega328 is a 5V chip)
This will ensure you can drive the strip VCC and Signal at the same voltage.

If it is the logic level it is quite easy to do a 3.3->5v step up using a optoisolator, or logic-level (mos)fet.

As to why some strips work and some dont; not all neopixels are created equal… there are clone pixels out there that are less forgiving than the official part.

2 Likes
6

You can only test continuity of all the 5V pins to VCC, all the 0v pins to gnd, and the data pins between each pixel and the pads at each end of the string.

Neopixels are small 6-pin self contained microcontrollers, with 2x vcc, 2x gnd., data in and data out pins, the data pins are wired between pixels in series. The diodes in them are inaccessible, internally wired and driven by appropriate current-limiting circuits; which is why the colors are so well balanced.

2 Likes
7

I tend to utilise an ESP32 or ESP8266 with a Bench PSU or dedicated 5v 8amp psu. i am still exploring possibilities in my head, so am not running anything on a permanent basis and swap things about regularly.

Probably a stupid question but can i ask, if i run the array from the 8amp PSU and also connect the micro controller to the same leads, if the draw of the LEDS increases to say 3A will this affect the controller Amps? Should this be on separate supply with a lower ampage (i.e the ESP32 uses a max of 240mA so i read)

8

A 5V 8A supply is capable of supplying 5V at up to 8A. It doesn’t force more current through a device that draws less. In fact, “force” here is voltage; voltage is electromotive force. So 5V 8A means that it will supply 5 volts of force while supplying anywhere in the 0–8A range.

Speaking abstractly, if you put a 0.5 Ohm load across thta power supply, that would “try to” draw 10A, and the supply would either supply a lower voltage or trip overcurrent protection.

The LEDs drawing 3A would not change the amount of power the ESP32 would draw. In parallel, you can just add up all the max current draw from the individual components (here, 3A + 0.24A = 3.24A) and as long as it is within the maximum that the power supply can provide, you are OK.

Does that make sense?

2 Likes
9

Hi

Thanks for the explanation, i dont know if this is obvious but i struggle with this:-

They are both using the same PSU and connect using branches from the same line. If the LEDS are drawing 3A why does the ESP32 not get 3A, is their regulating circuitry to reduce the Amps to the ESP32 on the PCB or is it just the miracle of electricity :slight_smile:.
They are using the same cables and one device needs a bigger Ampage, so surely that is the dominant force ?

Sorry if it sounds rediculous but these are the things that peak my curiousity.

10

Please show us a picture of the whole thing.

If you have connected them in series then absolutely the same current has to flow through both, and in addition it absolutely won’t work.

If you have connected them in parallel then the current adds up.

Imagine the confluence of two rivers.

One river is flowing at 1000 cubic meters per second, the other at 2000 cubic meters per second.

Downstream of where they join, the combined river will flow at 3000 cubic meters per second.

Rivers don’t divide as often as they combine, but the same rule applies in reverse in those rarer cases (like a river delta) where they do. If that 3000 cubic meter per second river divides at a delta, and 500 cubic meters per second flows down one fork, 2500 cubic meters per second flows down the other.

Amps are like that. One amp is a coulomb of charge (about 6.24 x 10¹⁸ electrons) passing though a location in a circuit in one second. It’s literally electrons/second.

So if you have 3A worth of electrons going from the LEDs each second, and .124A worth of electrons going from the ESP32, then the wire from the power supply to where you branch apart between the LEDs and the ESP32 is carrying 3.124A worth of electrons per second past any particular point. (Those electrons actually flow very very slowly; there are a lot of electrons in the wires.)

Does that help?

11

Great, that clarifies the Amps distribution perfectly, thank you.

I do not currently have anything to show, i have been playing around and keep changing things like the LED strips/panels and also controllers (ESP32, ESP32-C3,ESP8266 etc).

I had messed up the Pin connection on the ESP32 ( confusing it with another module) and connected the GND to a pin labelled CMD ( im getting old, the eyesights getting worse), this reads a lower voltage output at around 2.6v so shouldnt have damaged anything as far as i can see ?

I will have to get a setup working as expected using the bits i have, then swap out only the LED unit and reset parameters in WLED and see if i can them to work.

12

Ah, CMD is usually GPIO11, I think.

If your ESP32 board has the GPIO6, GPIO7, GPIO8, GPIO9, GPIO10, and GPIO11 pins, you must not use them because they are connected to the flash memory of the ESP32: if you use them, the ESP32 will not work.

I would be surprised if you did damage to the board by pulling CMD down to GND.

1 Like
13

I’m sort of sorry I mentioned Current there, my goal was to note that power supply would not be a problem on most development boards.

I think it’s important to pay attention to voltage instead. The ESP32 is a 3.3v chip and the output signal it sends is 3.3v.

Some neopixels expect the control signal to be above 3.5v, If your rig works with some boards but not with others, this may well be the reason.

Quickly changing the neopixel vcc from the 5v pin to the 3.3v one on your devboard might work. All devboards I have seen can handle 500ma+, which should be enough for testing.

If that does not work you need to provide a 5v signal.

Testing strips using an old school, 5v, Arduino Uno is a quick solution.

A more complex one is to use an optoisolator os signal FET to translate the signal to the 5v supply used by the strip.

Regarding running the whole Neopixel strip at 3.3v; here is 28 Neopixels running at 3.3v on a Arduino Zero compatible board (a SAMD21, 3.3v controller).

IMG_20240827_094943
IMG_20240827_0949434000×3000 2.66 MB

As the thread I linked above says; the specs quote Minimum 4.5v supply, Minimum 3.5v signal. But many units can be run from 3.3v, however YMMV.

1 Like