0

I try to make a system to lift up and down a door for a chicken coop. When I did it on a breadboard, everything was working correctly but now that I use a PCB, when the H-Bridge is connected, the microcontroller doesn't turn on although I did the same routing as the breadboard and the motor turn but slower than it should and after a few seconds the motor speed decrease more and the H-Bridge is burning but if I don't connect the motor to the H-Bidge, it is burning after a few seconds too.

I use a esp32 as microcontroller, a SN754410E for the H-Bridge, the system is powered by four batteries of 1.5 volt each and I use a OKR-T/20-W12-C DC/DC converter to convert the 6 volts from into 3.3 volts for the esp32 and my oled screen. The motor is connected throught the pin head with the name X2. Three push buttons in pulldown are used for the microcontroller and activate the motor in forward or backward. X1 if for a switch to turn on or off the system without touching the batteries.

Schematic PCB enter image description here

I tried to change the routing on the PCB and power the H-bridge by the Vin input of the DC/DC converter but it doesn't work and I have still the same problem although on the breadboard, it works without overheating.

Thank you in advance

ADN
  • 3
  • 2
  • 2
    There seems to be a complete absence of power rail decoupling capacitors. – Andy aka Dec 21 '20 at 10:48
  • How does the board look with those polygons filled in? Perhaps there's some trace with a wrong priority that is merged with a polygon, perhaps theres too little clearance somewhere and things touch? We can't tell from this image. – Unimportant Dec 21 '20 at 10:57
  • @Andyaka I didn't put them on the H-bridge my bad but it works without them on the breadboard. – ADN Dec 21 '20 at 11:02
  • @Unimportant I edited my first message with the pcb and the polygons filled. It was the first thing that I checked to be sure there were nothing touching somelthing else but i didn't see anything. – ADN Dec 21 '20 at 11:07
  • 1
    Let me rephrase this - there appears to be a **complete absence** of power rail decoupling capacitors on **any part of the circuit**. – Andy aka Dec 21 '20 at 11:08
  • @Andyaka OK, I will modify this. Thank you. – ADN Dec 21 '20 at 11:11
  • But the absence of decoupling capacitors shouldn't be a very important part of the H-Bridge heating up and motor running slow. – paki eng Dec 21 '20 at 11:23
  • 1
    @AbdullahBaig yes, you are right but that doesn't mean we can ignore other troubled areas on this PCB. See the link in my answer for reasons that the L293 will likely get too hot. – Andy aka Dec 21 '20 at 11:27
  • You also appear to have floating inputs on your H-bridge IC. You should not leave digital inputs floating. – Tom Carpenter Dec 21 '20 at 12:00

1 Answers1

0

Do your self a really big favour and avoid disrupting the ground plane you have on one side by minimizing things that don't need routing on it. For instance, a few simple modification can achieve this: -

enter image description here

I've just taken your image into paint and redrawn blue tracks as pink tracks. The pink tracks can go on the same PCB side as the red tracks and that means a much, much better ground plane.

Also, as per my comments, there is a complete absence of power rail decoupling capacitors.

See also this Q and A that addresses many problems that the L293 motor H bridge faces on low voltage supplies. In short, try to avoid using it because it will get very warm and burn on motor currents more than a couple of hundred mA.

Andy aka
  • 434,556
  • 28
  • 351
  • 777
  • Thank you for the link and I will modify my pcb and take in account this for my future pcb. – ADN Dec 21 '20 at 11:34
  • 1
    @ADN you should also [take the 2 minute tour](https://electronics.stackexchange.com/tour) to understand the motivation for providing free help to budding engineers. – Andy aka Dec 21 '20 at 11:35