Strange character on the serial monitor

[dyeatman]

All of this is why, years ago, I ditched crystals entirely in all my designs and
started using clock modules. Takes a slight bit more room but worth it for the
stability and accuracy.
_________________
Google and Forum Search are some of your best tools!!!!

[temtronic]

In the picture of the btm side of the board, the 6th pin under C14 doesn't look properly soldered...
Kinda curious as to the 'zigzag' routing of that port of pins...

[PCM programmer]

Here is a quick critique of the power and ground layout.

The red lines are the +5v power. They're bad. Skinny little tracks going
all over the place. Not gridded.

The yellow circles are ground problems. The big one is the ferrite bead
in the ground path in the upper left. You want a low impedance ground path.
Don't put a ferrite bead in the ground path.

The bright green line is the ground path for the 18-pin PIC. The ground path
from the crystal circuit to the Vss pin on the PIC is too long.

[erpgc82]

[erpgc82]

[erpgc82]

[erpgc82]

Hello my friends, sad about this problem.
Finger on the GND is necessary to operate the serial communication.

As a recommendation from some, I did these procedures in order:
- 1st, I removed FERRITE BEADS, no positive result;
- 2nd, reinforced GND trails, no positive results;
- 3rd, I removed all the components related to the LM2576, and started using the 7805 with 2 capacitors of 220uf, one at the input, another at the output, no positive result;


Start, without touching the GND;
perfect, finger on the GND;


reinforcement in the GND;
removed ferrite beads;


removed the entire circuit of the LM2576;
using 7805 (2 electrolytic) 220uf;


     I don't know what it can be anymore, I never had this problem before.
     Even considering the flawed, problematic layout, I could not have such a critical problem in such a small and simple circuit.

     With so many modifications, no minimum improvement, except by touching the circuit's GND with your finger.

     I will continue studying the book, placing my finger on the GND to visualize the result on the SERIAL MONITOR.


_________________
Gradually you will go far with persistence, will and determination!

[PCM programmer]

The basic problem you have only this one test setup.

You need to get another complete test setup. Another board, another PIC,
another crystal (from a known-good manufacturer), etc.
Maybe borrow it from a friend. You need to test with known-good
hardware. The board should be made by a company, and be known to work.

[erpgc82]

[Ttelmah]

So the obvious suspicion is you just have a faulty MAX232.

It's worth saying that when building a 'protoytype' board, I usually make three
or four. Always minimum two. It is so hard otherwise to rule out hardware
problems....