-->
Page 1 of 2

What's wrong with this circuit?

PostPosted: Sat Oct 28, 2017 7:05 pm
by Millstone
I can rarely programm th ESP here. Ist there anything wrong with the ciruit?

What I found is that at the beginning I had a 12k pullup on GPIO2. It was continously oscillating, so I changed to an 820Ohm resistor et voila I could programm the esp and open an accessport/log into wifi. But the overall performance was not ok. It crashed very very often. 2200µF between 3.3V and GND had no effect. :(
I already exchanged the chip one time. The other chip also oscillated on GPIO2. Is there anything I could do better here? Please help me!

Re: What's wrong with this circuit?

PostPosted: Sat Oct 28, 2017 7:19 pm
by rudy
A schematic shows connections but does not show how you made the connections. Can you take a picture of the setup?

Also. What are you using to power it?

Re: What's wrong with this circuit?

PostPosted: Sun Oct 29, 2017 2:02 pm
by Millstone
Thank you for your answer. It's a pcb in eagle. Here I file you the picture.

Re: What's wrong with this circuit?

PostPosted: Sun Oct 29, 2017 3:44 pm
by rudy
I can't see anything specific that I could point to as being the cause of your problems. But I will tell you my thoughts on the routing on this board.

A - If you will me using the on board antenna then you should not be placing conductors (ground) underneath it.

B, C - You have a lot of vias connecting islands of copper that go nowhere. If there is not going to be any current flow then what is the point of the connections? It may not be hurting anything, unless it is giving you a false sense of comfort. Lots of vias to copper patches = impressing that you are doing something good. It doesn't work that way.

If the vias at B were placed to provide thermal conduction then that purpose would have been valid, although I doubt that is the case here.

D - All the current that is going into the ESP8266 is going through the one lone via at D. If you were going to get via crazy then this is the place you should have done it. You have a choke point on the most important pin on the board.

Imaging a current path that starts at the ground pin, continues to the bypass capacitors, through the other side of the capacitor, then continues on to the Vcc pin. How direct is that path? Is it short and direct or is it long? Power routing should be a primary concern rather than an afterthought. When I design boards it is the first thing I take care of.

What voltage regulator are you using? And what is the input voltage?