Hi all,

I have a question about the simplified circuit of the below image.
It is about a 16x16 5V LED matrix, controlled via wifi by a controller board with an ESP 8266.
I do have more details about all components used, but I try to keep my question simple.

I can control the LED matrix by wifi (through the ESP 8266).
The problem is that when I set the LED matrix to white and bright, the LED matrix consumes quite some power. Then, there is not sufficient power left for the ESP 8266 to function. As a result, I cannot control the LED board anymore (since the ESP 8266 is not functioning) and it is 'forever' stuck on this bright setting.

Is there a way to limit the power (or current) going to the LED matrix?
I want to ensure that:

1. The ESP 8266 always gets sufficient power to function and
2. The power supply (5V, 4A) is not overloaded

I tried to include a voltage regulator (5V 2A) between the power supply and LED matrix.
However, the voltage regulator gets considerably hot. I do not think the voltage regulator will withstand this for a long period of time.
Of course, I could switch to a heavier power supply, but this is not preferable.

What would be a good and durable solution to limit the power to the LED matrix?
I am aware that if I limit the power to the LED matrix the LEDs will be less bright, but I have that rather than an uncontrollable LED matrix.

Thank you in advance!

 

Assuming you are using addressable LED's...
(1) reduce your maximum brightness for all the LEDs to 50% or even less.
(2) replace the power supply with a more powerful unit.

 

Purchase a separate Power-Supply for the Controller.
Connect the Grounds together, done.
.
.
.

 

I do have more details about all components used, but I try to keep my question simple.

Information about how you're driving the LEDs would be helpful, as would the time (are they COB/addressable?). If you have a 16x16 matrix, you could choose to have only 1 LED on at a time (R, G, and B in the same LED, so technically 3 but you didn't count them that way) and not consume significant power for them.

 

Information about how you're driving the LEDs would be helpful, as would the time (are they COB/addressable?).

Thank you for your reply!
Yes, the LEDs are individually addressable.
The idea is that they light up in a certain pattern, so finding a solution that does not affect the pattern would be nice.
I prefer a hardware solution (and not limit the max. brightness in the software).

I tried integrating a step-down buck converter with a max output of 5V 3A. The input is 5V 4A.
However, the converter gets considerably hot.
However I do not understand why the converter dissipates energy since the input and output voltage are the same (5V).

 

Yes, the LEDs are individually addressable.

Then this is completely controllable with software. Just reduce the brightness all the time, divide all the settings by 2 for instance.

 

You might try just putting a resistor in series with the LED panel's power input. Without knowing more specifics, it's impossible to determine if this will cause other problems or how to size the resistor.

 

Then this is completely controllable with software. Just reduce the brightness all the time, divide all the settings by 2 for instance.

The problem with relying on a software fix in this case is that the hardware is not properly matched (i.e., the supply can't meet the full needs of the system under all conditions). Hence, if, for any reason, the LEDs pull too much current, possibly even in a short transient, it might cause the controller to become unresponsive.

 

The problem with relying on a software fix in this case is that the hardware is not properly matched (i.e., the supply can't meet the full needs of the system under all conditions). Hence, if, for any reason, the LEDs pull too much current, possibly even in a short transient, it might cause the controller to become unresponsive.

Exactly!

You might try just putting a resistor in series with the LED panel's power input. Without knowing more specifics, it's impossible to determine if this will cause other problems or how to size the resistor.

I am just getting into electronics, so please forgive me if I ask stupid questions.
Integrating a resistor would cause the voltage to drop, right? And because of the voltage drop, it limits the power?
What specifics are required to determine this?

Thanks a lot!

 

What is the distance between the power supply and the matrix and what is the gauge of the wire and where is the micro connected along that line?

 

Why not just provide a separate Power-Supply for the Controller as suggested in Post#3 ????
Is it too simple, so you couldn't possibly be serious ?

If the Peak-Current is too high for the present Power-Supply to comfortably deal with,
simply install a big-fat Power-Resistor after the Power-Supply to limit the Peak-Current to the desired level.
Or, get a bigger Power-Supply.
It's all really just that simple.
.
.
.

 

What is the distance between the power supply and the matrix and what is the gauge of the wire and where is the micro connected along that line?

The length of the powe rcable is approximately 4m. It is a 0.75mm2 cable.

 

Why not just provide a separate Power-Supply for the Controller as suggested in Post#3 ????
Is it too simple, so you couldn't possibly be serious ?

If the Peak-Current is too high for the present Power-Supply to comfortably deal with,
simply install a big-fat Power-Resistor after the Power-Supply to limit the Peak-Current to the desired level.
Or, get a bigger Power-Supply.
It's all really just that simple.
.
.
.

If I understand you correctly, you suggest to use a second power supply/adapter for the controller?
I understand this would solve the problem of the ESP controller not getting sufficient power. But this would mean that I need to use two outlets which is undesirable. Also, this would not solve the problem of overloading the power supply of the LED matrix, right?

A bigger power supply would be the simplest solution. Since each LED can consume max. 60mA at 5V and I have 256 LEDs, I would need a 80W power supply at least. This becomes quite heavy an bulky.
I hope to find a more elegant solution, also since the LED matrix does not constantly ask for a lot of power. I just want to prevent it from being able to ask 80W.

 

Is the Controller at the beginning, or the end of that Cable ?
Is the Main-Power Supply at the beginning or the End of that Cable ?
.
.
.

 

Is the Controller at the beginning, or the end of that Cable ?
Is the Main-Power Supply at the beginning or the End of that Cable ?
.
.
.

The power supply is at the beginning of the cable, the controller very close to the matrix.

 

You don't want to change anything .......... but You want your problem fixed .............
Probably not going to happen that way.
.
.
.

 

The length of the powe rcable is approximately 4m. It is a 0.75mm2 cable.

Where is the micro connected to power...

The power supply is at the beginning of the cable, the controller very close to the matrix.

That could be your problem, can you run a separate cable from the supply to the controller?

 

"" The power supply is at the beginning of the cable, the controller very close to the matrix. ""

This is the crux of the problem.
This is why You need 2 separate Power-Supplies.
.
.
.

 

You don't want to change anything .......... but You want your problem fixed .............
Probably not going to happen that way.
.
.
.

My first aim is to find a hardware solution to limit the power-ask of the LED matrix. If this is not possible, there is indeed no other solution than go for a heavier/second power supply.

 

That could be your problem, can you run a separate cable from the supply to the controller?

I think this is possible. Can you please explain how this solves the problem?