Wiring question - attaching GND from LED strip 1 to LED strip 2

Hello, I have a question regarding wiring. I would like to arrange two strips of WS2812b LEDs (240 LEDs each) as shown in the attached diagram. +5V will be attached to the beginning and injected at the end of each strip. However instead of attaching GND to the beginning of strip 2, I’d like to wire it in to the middle of strip 2 from the middle of strip 1.

I don’t believe this would cause any issues, as GND does not need to be injected into particular points like +5V. But I wanted to double check here. Could this lead to any potential issues, or should it work as expected?

dlovinger:
I don't believe this would cause any issues, as GND does not need to be injected into particular points like +5V.

That is a particularly foolish statement! :astonished:

Ground is even more critical than 5 V supply as it is in general the reference for the data signal.

240 LEDs means a peak current (possible even if the LEDs are less than full brightness as the PWMs combine) of just short of 15 Amps. You should be connecting both 5 V and ground every 60 LEDs or so using fairly sturdy twin cable, not just both ends of each strip.

The foils on the strip are not capable of carrying anywhere near 14 Amps without causing problems.

Paul__B:
That is a particularly foolish statement! :astonished:

Ground is even more critical than 5 V supply as it is in general the reference for the data signal.

240 LEDs means a peak current (possible even if the LEDs are less than full brightness as the PWMs combine) of just short of 15 Amps. You should be connecting both 5 V and ground every 60 LEDs or so using fairly sturdy twin cable, not just both ends of each strip.

The foils on the strip are not capable of carrying anywhere near 14 Amps without causing problems.

Thank you for fixing the image! I hear you, the brightness is significantly limited such that the total max draw is ~5A (measured with a multimeter), 2.5A on each strip. Average draw is about half that. Still, this is a sizeable amount of current. However, GND is at 0V and should therefore generate little to no heat. Heat and voltage loss are typically the reasons to be wary of high current. Do you still believe there would be issues running a single GND wire from strip 1 to 2?

The concern I mention is that the brightness "limitation" is by means of PWM so that each individual LED at some part of the time may draw the full current - 60 mA per LED.

This depends of course on whether you are expecting a single primary colour or a mixture, and whether you are lighting just a few LEDs relatively brightly or many just dimly. In the latter case, if the PWM cycles happen to synchronise - and I do not know to what extent this will happen in a strip of WS2812 - you may still get very substantial impulse current draw.

It follows that if you are generating patterns where just a few of the LEDs are illuminated at any one moment, and with a primary colour, then you may mange with a 5 A supply. These strips as such, lend themselves to "chasing" patterns. :grinning:

I would still advise placing the 470 µF or 1 mF capacitors every 60 LEDs and at each end to suppress these impulses, and you should ensure that 5 v, ground and data wires where relevant, travel as a pair or triplet, generally using "figure 8" or ribbon form cables and connect together to the strip at the same place. The requirement is to avoid open loops between supply and ground or data and ground.

However, GND is at 0V and should therefore generate little to no heat. Heat and voltage loss are typically the reasons to be wary of high current. Do you still believe there would be issues running a single GND wire from strip 1 to 2?

You seem to have a fundamental misunderstanding. Sure, 0V is 0V a the terminals of the power supply but it won't be along the strip, the further you get from the power supply the further from 0V the voltage will be. The conductive strips on LED tape are not very thick and there will be significant voltage drop across them because of the current through them; that includes the 0V strip. The voltage on the negative strip at the end away from the power supply will be significantly above 0V when the LEDs are lit. I suspect you either don't know Ohms law or you have not realised how it applies in this case. If you don't know Ohms law please take a moment to learn it; it's fundamental to all electrical circuits.

It's an interesting point. :grinning:

Having worked with electricity and electronics for many years, we tend to have a "feel" for current and the thickness or wires or in this case, the relative flimsiness of foil on these that are in fact, flexible PCBs. Never had those in "the old days" though of course, these flexible strip PCBs have always been the basis of disk drives, CDs, printers and such.

So we automatically say "that won''t be sufficient" without really going into the mathematics of how wide and thick the traces are.

Thanks for the advice both of you. I am aware of Ohm's law but I neglected to apply it to the GND wire - it has the same current so it has to accumulate a potential difference down the length of the strip as well. That and the other reasons given will just lead to trouble, so I'll take the extra effort to route a GND wire around the project as well.

I am aware of Ohm's law but I neglected to apply it to the GND wire - it has the same current so it has to accumulate a potential difference down the length of the strip as well.

Right, good, glad that makes sense. Sometimes I find it hard to provide the right answer with cases like this; it was clear to me that you were missing something fundamental but not so clear what exactly it was you didn't properly understand.

I hope you get your project working.

You might also want to investigate the terms “ground lift” which is what you are seeing, and also the transitory “ground bounce”.

In other circumstances ground lift can also affect the reading of an analogue input if there is a heavy load on series wiring of a ground. Look up “star wiring”.