Self-built ESP12-E pulling 700mA and getting hot

Projects General Guidance
1

Hi,
I'm currently prototyping a self-built version of a microcontroller using an ESP12-E and micro-USB + ch340c for communication with my pc.
The ch340 is connecting properly to my pc and is recognised by it.
when I connect the ESP though it draws constantly more than 600mA and gets hot really fast. I once got it to start uploading the blink sketch but got an error at about 27%. On this one the current draw was at a normal 80mA. The other times I try to upload the code I get this error message:

Serial port /dev/cu.usbserial-1434440
Connecting....

Chip is ESP8266EX
Features: WiFi
Crystal is 26MHz
MAC: e8:db:84:8b:61:34
Uploading stub...
Running stub...
Stub running...
Configuring flash size...
A fatal esptool.py error occurred: Timed out waiting for packet header

I already checked all the soldering connections and the wiring.
One weird thing I noticed is that on its own the A1117 regulates power to 3.3V properly, but when connected to the circuit it outputs about 3.6V. On one try it was at 3.8V without me changing anything.

Does anyone know if there is a reason for the high current draw and if it correlates with the power voltage and/or the temperature? Also why did it upload up to 27% percent once but never again, with the exact same setup.

This is my schematic:

schematic
schematic1520×1006 33.1 KB

Few Notes to the wiring:
Power is supplied from 2 sources:

  1. the USB-Port (V+ Net) through a Diode into the main +5V net.
  2. The connected batteries also supply into the +5V net.

The voltage regulator outputs a +3.3V net for the ESP and the CH340.
The connections to the ULN, Servo and Stepper only use IO Pins and should not be relevant for this issue.
Both resistors are 10K, C1 is 100nF, C2 is 100uF

Thanks for any help!

2

What is the 16 pin chip. Pcf8574 or uln2003?
What about capacitors on the voltage regulator (see datasheet)?

3

No capacitors (or the wrong ones) on the regulator could make it oscillate.
And that's a bad thing for the ESP.
Leo..

1 Like
4

Hi, @ghost_y

Can you please post EXPORTED images of your PCB from the CAD IDE?
So we can see your component layout.

Can you please post some images of your project?

Thanks.. Tom.... :smiley: :+1: :coffee: :australia:

5

Hi, @ghost_y
Where are your 1117 bypass capacitors?

image
image356×218 3.92 KB

image
image894×415 25.4 KB

Very essential components, read the data/app sheet.

Tom.... :smiley: :+1: :coffee: :australia:

1 Like
6

Based on what's been said already about the lack of caps; If there's oscillation your multimeter won't catch that, but gives you a mean.

You should consider that ESP a thing of the past, maybe other components too.

7

It's uln2003

8

Currently I'm just using breakout boards and a breadboard, here are some images of the setup:

IMG_1742.HEIC
IMG_1742.HEIC4032×3024 2.57 MB

IMG_1741.HEIC
IMG_1741.HEIC4032×3024 2.49 MB

IMG_1740.HEIC
IMG_1740.HEIC4032×3024 2.31 MB

IMG_1739.HEIC
IMG_1739.HEIC4032×3024 2.24 MB

IMG_1738.HEIC
IMG_1738.HEIC3024×4032 988 KB

The VR being on the bottom of the ESP breakout makes it kind of weird to access, but I know added the 10uF Caps I initially forgot directly to the bottom.

Since adding the caps the ESP draws constant 40mA and I can upload up to 25% every time, until the current spikes to over 600mA again and I get this error message:

A fatal esptool.py error occurred: (22, 'Invalid argument')

So adding the capacitors certainly made it more consistent. Could the not going past 25% be due to some other issue or just the ESP being broken from the earlier inconsistencies?