Hardware related question: Why does it fry the first LED in a WS2182 strip (NeoPixel), if the data pin is connected to an arduino, but the power is not.
Last night I fried an LED by plugging in my ardunio to reprogram it, but the external power supply was not on so the strip was only receiving the data signal from the , now USB powered, arduino.
Also, how to most of you avoid this wire configuration trap when using an external power supply for the LEDS?
My external power has a fan in my lab setup so I always power it off when coding and leave the teensy connected to the data pin of the strip. I never experienced such a problem.
Did you connect the ground pin of the arduino with GND on the strip?
I think the initial inrush of current, although of very short duration is a sufficiently high current to damage the internal input circuitry of the ws2812.
This is why many use a 300 ohms or so current limiting resistor before the first pixel data-in. I have never used that method because I always power the Teensy with the same PSU used for the LEDs. In the case of the Teensy, I just cut the track between the pads that completely disconnect the USB power.
Yes, the grounds are all common between the strip, arduino, and power supply.
I also have a resistor between the datapin on the arduino and the strip.
I heard something regarding this; that when there is data, but not power the strip attempts to power itself through the data pin, causing the LED to break. This doesn’t make any sense to me, but maybe its true. I couldn’t glean any information regarding this from the WS2812 datasheet so I figured I’d ask.
In my experience ws2811 leds will draw curent via the data line, and i have seen strips maintain whatever was showing when power is interrupted, because although the leds were unlit, the chips were still drawing power via the data line, that the arduino had held high during reprogramming.
According to the ws2812 and ws2811, Input current is +/- 1 microamp max. So the input is a typical high impedance input but that is specified under normal Vcc range of 4.5 to 5.5 volts. It may very well react differently with Vcc = Gnd as Stuart experienced.
But the 300 ohms resistor should limit that current to safe levels I would think !?
Adam go to this link it would help you with wiring leds up,when I started I knew nothing about arduino or leds , and lost many no1 leds in the strip.
Thanks for the responses guys. I cut off the first led and tested the strip. Luckily the rest of the strip was fine.
@JP_Roy I believe they suggest including a resistor on the data line in order to prevent ringing. I recall seeing some Oscilloscope readings that showed the ringing of the data signal with and without that resistor.
@Stuart_Taylor That’s what I had heard, but was trying to find some more in depth reading or data sheet that showed why the LED draws current through the data line.
Hi @Adam_Wright ,
My XMAS tree is built around a MEGA2560 and each of the 34 branches is driven by a separate pin and I do not use any resistor between a pin and that 1st LED on a branch.
Note that the wire length between the MEGA and the branches vary between 3 feet and 8 feet and I never experienced issues that would point to any ringing. That is not saying that there is none, just that if there is any ringing, it has no visible effect.
The Adafruit guide mentioned above talks about using a large capacitor across the + and - just before the strip. I read about this elsewhere apparently it smooths out the current going into the strip, avoiding the sudden spikes which lead to a blowout. Since I started using one the survival rate of the 1’st pixel on my strips has been much improved! Also helpful is limiting the brightness in code when trying out new configurations.
@Tom_Hastings the use of a capacitor in this way, is known as a reservoir capacitor.
The principle is that because the internal resistance of the capacitor is so low, it can deliver charge quicker than a battery or a psu can.
Where the psu is down a length of wire, even the small amount of inductance can actively resist the current (reluctance), so the local resevoir reduces any slump.
I’m not sure why a capacitor would mitigate a volt drop from frying the first led. Unless its not the led that fries, but something local to it.