I am working on creating two small circuit boards that can control LED lights using the WS2811 IC and protocol.
The first board is designed to control 1 W COB LEDs, as it maintains a constant current of around 300 mA on the outputs by using the PT4115 chip. By changing the inductor to a 68 µH one and decreasing the current sensing resistor to 0.15 Ω, the same board should also be able to control 3 W LEDs.
The second board is a MOSFET driver that can control regular LED strips.
Both boards use a 78L05 regulator to ensure that the IC is supplied with a constant 5 V regardless of the input voltage, which can be either 12 V or 24 V.
I was inspired by the many great products and open-source projects available on the internet. They mostly shared very similar designs so I thought it wouldn't be too difficult to create my own versions. Even the datasheet showcased this kind of design! I made the schematics and had custom PCBs manufactured without prior testing. Unfortunately, after receiving and assembling the boards, I found that they do not work properly.
Since all the components are working correctly and there are no soldering errors, there seems to be a flaw in the design. Despite my efforts, I cannot figure out what is causing the issues or how to fix them:
On the MOSFET board, the B and R outputs are always off, while the G output is always on. On the other board it's the opposite, the B and R outputs are always on, while the G output is always off. This is strange because all outputs are the identical so I expected them to at least behave similarly.
Although the WS2811 chip is functioning, it behaved unexpectedly while testing it. For this I created a simple circuit using an Arduino powered by USB and a small LED on a breadboard, equivalent to the one shown in the datasheet. When I power both the IC and the LED from the 5 V pin, the LED briefly flashes instead of remaining illuminated. However, if I power the IC from the 5 V pin and the LED from VIN (or vice versa), it works as expected. This behavior is also quite perplexing to me.
These are the schematics in question:
I would be grateful if someone could assist me in identifying the cause of this behavior and help me correct it.
Exactly, in the addressable strips they are connected like this. In my application I'm using the chip to instead switch the BJT and MOSFET which will then handle the bigger loads.
It's just a bare chip in a SOT-8 package. Check the datasheet
The outputs of the WS2811 are by design regulated at a constant current of 18.5 mA, so I think there's no need to limit the current with a resistor.
I have no idea where you've got that information from, take a look at page 2 and 3 of the datasheet. I've used it in the past without issues at 5 V and it's also commonly used with 3.3 V gate voltage.
MOSFET board: if you desolder BJT and connect Base and Collector pads together, do W2811 his job in inverse? i think exchanging the Base pull-up resistor with 470KΩ will let MOSFET board work properly.
Is that powering the LEDs and LED drivers driver? If so you are over-stressing the 7805. It's probably overheating and the 5V probably isn't holding-up, or it may be oscillating.
They are rated for about 1 Amp but the more voltage you drop across it, the more power it has to dissipate and the hotter it gets.
You can power the constant current drivers with a higher voltage so they shouldn't need a voltage regulator. The regular LED strips may need a dedicated power supply or a switch-mode DC-DC converter than can handle the current more efficiently.
No, the 78L05 regulator is there only to supply 5 V to the WS2811 chip and to the pull up resistors for base and gate. The current for the LEDs comes directly from the input (VCC) and doesn't run through the regulator. I think the schematic correctly reflects this.
WS2811 power and ground to an adjustable voltage 60W OEM PSU set to 5V. The PSU ran me 7 or 8 bucks and powers the strip directly.
Power and ground to an Uno or Nano
Data pin on Arduino to data line on WS2811.
The 2811 manages power, you only need a solid power bus and maybe a bypass cap if stability is an issue. I don't refresh more than 60Hz and have had no issues.
Don't COB leds need constant current circuits?
The best extra I have seen are DC converter boards.
@GoForSmoke What you did was using addressable strips based around the WS2811 or WS2812B. What I'm trying to do here is different. I'm making two boards that turns an "analog" strip or a COB LED into one big addressable pixel that can be driven through the WS2811 protocol. That's why in one board the WS2811's outputs drive a MOSFET while in the other one they control the DIM pin of a constant current driver.
Then why not simply using an Analog LED driver instead of trying to use the WS2811? You can turn almost any MCU into an analog LED controller. Why are you limiting yourself by saying you want to control this strip with a chip designed to control a single LED?
Did you try to PWM a FET (I bought logic level low end FETs with an L and and N in the name.) switching power for each junction --- RGB has 3 so you need 3 PWM pins and 3 FETs and 1 PSU to drive 1 RGB led to make 24 bit color.
I ran G4 12VAC bulb replacement disks using Uno PWM FET control and 12VDC (the PCB rectifies AC, DC works) but could only fade/brighten white. Oh yeah, I got 60 IRLZ44N's for $10, plenty for my needs.
