I posted a picture of a circuit I created with a handful of PCB boards. The picture only shows one shift register but there's really a total of 8 (meaning 56 leds, transistors, resistors, etc.). The buck converter steps the voltage down to 4.5 volts so the esp8266's voltage regulator can smooth out the voltage even more. Everything worked flawlessly with all of the display sketches I made. However when I enabled a wifi sketch the display bugged out which I assume means the chip failed to execute the code correctly. And by bugged out I mean turned on a few random leds as soon as it was plugged in and that was it). I think it has to do with a voltage problem in the circuit because the code runs on the module with no problem when it's NOT hooked up to the display. Any thoughts on why the module could bug out when wifi is enabled, and only when it's hooked to the display circuit shown in the image?
Another worry I have (which I've already fried a board because of) is the module frying when I program it. I had to order a new one because even with the Vcc line detached from the module's 5v pin, too much current was somehow able to travel through the digital output pins and fry it. A recommended fix for that would also be appreciated.
Check the power supply requirements of your 8266 module and that you satisfy them . some need 3.3-3.6v some more IF they have an on board regualator.
When the 8266 communicates its creates a high current spike, with all those LEDS , it may just be too much for the power supply; you could try a bigger capacitor , 100uf ? but chekc the power supply is ok with that (stable)
Try and measure or calculate the current being used -taking note of what the 8266 needs when commnuicating
It is also possible the shift registers contain random data whcen so powered up, resulting in the leds being on; there may be a way of connecting the shift register on start up ( I've not bothered looking -but disable the ouptuts and purge the shift register with a number of "0"'s , then enable it ).
I also think that it's a power issue. As the ESP operates at 3.3V, the 4.5V powered shift registers don't receive proper signals. Eventually add a level shifter (transistor) between controller and shift registers. Also add a couple of caps, and inspect the supply voltage for glitches, using a scope..
As a test remove all (but one) display module, and check again. Does the wifi part work, even if the shift registers don't work?
I might want to clear things up a little as well. The esp has an on board regulator for 5v. Wifi enabled sketches don't work while ordinary shifting sketches work perfectly fine on the display. So wifi sketches are the only time I have issues with the display running. Once the new one comes in the mail I'll have to see if enabling wifi causes and abnormal amount of current draw because that's the only thing I can think of with it having issues. I'll give tacking in a few beefy caps a go as well.
You mention 8 shift registers, 56xLED outputs
Just curious... why not 64... 56 LEDs only need 7x shifters.
While you’re waiting, have a shot a redrawing your diagram.
There are a few threads on schematics,but a good starting point is power from top to bottom (high to low), and signal flow from left to right.
Nice colours and wiggly lines!
You’ll like it more in the future.
I wanted to lighten the labor of designing the PCB by making it a repeating pastern. The whole thing is for a YouTube subscriber counter and it's been a bit of a headache given that I'm pretty new to all of this. Each PCB contains to digits so by soldering ___ boards you get double the amount at the expense of two pins per board being wasted. Hopefully the picture shows to show what I mean. And using paint to try and explain schematics.... ehhhh I need to move forward. I'll learn to produce better during the wait!
