-->
Page 1 of 1

RGBW-Controller

PostPosted: Wed Jun 10, 2015 4:29 pm
by LastSamurai
Hi there, I just finished designing my first circuit for a ESP-12 based RGBW-LED Controller. I have mostly seen only RGB and ESP-01 based stuff her e (which I tried myself too :D ) so perhaps this is interesting for someone else too.

It runs on 12V direclty for the LEDs and uses a small LM2596 buck converter module from ebay for the 3.3V. The LEDs are controlled via IRLM2502's which have handled the loads without a problem in my tests. There is a pinheader for the remaining unused GPIO pin as well as serial and some buttons for restart and flash mode, so that the ESP can be reprogrammed directly on the board if necessary.
Most of this is based on the great help http://www.esp8266.com/viewtopic.php?f=13&t=3271&start=30 from this topic.

I'd love to hear some feedback and get some suggestions where I can improve this.

TODOs:
    - Perhaps I will add some pulldowns to the rgbw control lines to avoid unintentional flickering
    - I am not 100% sure about the capacitors. Are the 2 100uF elektrolytic caps for the buck converter and the ESP enough?
    - The 470uF cap at the restart pin is copied from another design I saw. Would it be ok to use a 100uF one there too? Only one kind for the whole board then ;)

Re: RGBW-Controller

PostPosted: Wed Jun 10, 2015 5:33 pm
by ArnieO
The capa on the reset pin works with the 10 k resistor to form a power-up reset circuit. Its purpose is to keep the reset pin low while the unit is being powered up, to ensure proper reset. By multiplying the resistor value and the cap value you get the time constant (in seconds) of this circuit, which is approximately the duration of the reset (pin held low) from the point of time where voltage is applied. You write 470 uF, but your drawing says 470 pF.
With 470 uF: 470e-6 * 10 000 = 4.7 seconds
WIth 470 pF: 470e-12 * 10 000 = 4.5 microseconds
There is no "right" or "wrong" for this C, although 470 uF looks exaggerated. I would personally choose a 100 nF, giving a time constant of 1 ms . which should be sufficient.

Re: RGBW-Controller

PostPosted: Thu Jun 11, 2015 9:30 am
by LastSamurai
Thanks alot! Here is my second version with pulldowns for the control lanes and a 100nf cap on the reset.
Does anyone else see anything that could be done better? Has anyone experience with caps and the ESP8266? This has to be running 24/7 so power (and hardware in general) should be no problem if possible.

Re: RGBW-Controller

PostPosted: Thu Jun 11, 2015 12:01 pm
by ArnieO
It should be sufficient with one 100 uF capa on the 3.3V power.