voltage drop in LEDs (data pin and power) and maximizing efficiency

jos

Newly Enlightened
Joined
Dec 22, 2020
Messages
2
I am looking to set up a group of 1m LED strips (as many as possible). my power supply is 5V 15A, and I want to use it to power both the Raspberry Pi zero w and the LED strips. The first issue that comes up is that the Rpi only pulls 1.2 amps, and that it actually puts out less, which is not nearly enough to power the 250 LEDs I can reach with a 15A power source. Since amperage gets pulled through parallel circuits, and drops by the amount that is needed by the components in the parallel circuit, I'd assume that I can safely power the Rpi by putting it in parallel with the LEDs.
Assuming that is correct, I'd now only need to solve voltage drop through the LED strips. Each pixel will experience resistance and drop voltage, and after a very short distance (about 90 pixels) the color will get to be way off. But since voltage stays the same through parallel circuits and drops with every component that offers resistance in series, I'd assume that if I power inject in a way that puts all LED strips into a parallel line with each other, then each strip will receive 5V, and voltage drop will be fixed.
assuming that these last two paragraphs are correct, I'll only need a new power source when the amperage has all been drawn, which would occur after adding the fourth 1 meter strip to the circuit. so if I were to want 8 one meter strips, I'd need to get a separate 5v 15A power source and restart the process. I'd love to be able to put the same signal from the first set of 1m strands into the second, thus creating a larger set, using power injection from 1 power source every 1 meter for 4 meters, and using power injection from a new power source every 4 meters for as long as possible.
So here, I have a few questions:
A) is it true that my raspberry pi won't end up soaking up all 15 amps and getting ruined?
B) Is it true that I can run each strip and the arduino all in parallel, so that each component gets 5v power?
C) looking at my third paragraph: at what point will the data line experience enough voltage drop to cause a problem?
D) If I need a way to inject power into the data line (or something like that) so that it continues strong, how would I do it?
Here's a quick sketch that might be hard to understand, just in case anyone finds it clarifying:
 

Dave_H

Flashlight Enthusiast
Joined
Nov 3, 2009
Messages
1,348
Location
Ottawa Ont. Canada
Your sketch didn't come through, hopefully I understand what you are trying to do.

Your Pi itself running from 5v will not try to suck out any more current from the power supply than
it needs, certainly nothing like 15A.

If you are driving RBG line current to LEDs in the strip directly, and being resistance in the lines,
at a certain length the voltage drop is likely to affect different colours differently which can screw
up colour mixing. Joining strips end to end has limits, depending on how well they are designed
(and intended to be daisy-chained).

Not quite sure how you plan driving the strips from the controller end. Sounds like all current driven
to the strips has to run through this driver, which is quite a bit. The Pi can't drive this much directly,
so will need some form of HBD **.

Best bet is to drive as many strips as possible in parallel, and limit the series connections to the point
that colour in furthest out LEDs is not/minimally affected. As for regenerating LED drive somewhere along a set
of series strips, possible but unlikely worth the trouble.

Alterantive is to get strips using RGB drivers in each LED which are controlled by low-current (purely)
data lines; and power is applied separately. These will more expensive than the others.

Dave

** Honkin' Big Driver
 

DIWdiver

Flashlight Enthusiast
Joined
Jan 27, 2010
Messages
2,725
Location
Connecticut, USA
Hi Jos, and welcome to the forum!

If you want to put pictures in your post, you have to put them on a different server (i.e. Google Drive) and get a link to the picture. Then paste the link in the dialog box that comes up when you click the 'insert image' button.

A. +1, the Pi will be fine.
B. Yes, you want to hook them all in parallel so everyone gets 5V.
C. I assume these strips are the sort where you feed them power, ground, and data, and each pixel is controlled individually by the data stream. The point where you experience data transmission problems varies from one device to another. And there are various wiring schemes you can use to mitigate this problem. It's rare that you'd have problems with 0.3V, and the likelihood goes up as the voltage gets higher. By 1.2V you are almost certain to have problems, with most likely threshold being around 0.7V.
D. You would not inject power into the data line, only the power and ground.

To ensure signal integrity, I would aim for no more than 0.3V drop in the ground line. But for color matching, this might even be too much, since you'll have the same loss in the power line, and they add up. Let's say at full power you want the voltage loss to be no more than 0.1V on each line. Now we need to know how long the lines will be, and how much current is required. It sounds like each strip must draw 3 - 3.5A, I'll assume 3A. I'll also assume the longest line needs to be around 10 feet.

For 0.1V with 3A, that's 0.1V/3A = 0.033 ohms. For a 10 ft wire, you need the wire to be 0.033ohms/10 ft. = 0.0033 ohms per foot. 16 AWG wire is 0.004 ohms/ft, while 14 is 0.0025. So wiring each strip separately back to the power supply using 14 AWG wire would meet the design goal.

If your runs were longer or shorter, or your voltage drop goals were higher or lower, you could plug in different numbers and find the proper wire size to use. If the runs were significantly shorter, you could daisy-chain the power connections. This would mean running PS to first strip, first to second, second to third, etc., instead of running each one all the way back to the PS (that would be 'home run' wiring).
 
Top