ESP is not starting up first time power is switched on

I did this electronics based on a Doit ESP32 Devkit V2 processor board.

To power my circuit I have an 12V/4A power supply after that is a buck converter which gives me 3.3V for the circuit and ESP.

Power Supply: https://www.aliexpress.com/item/1005006678748569.html?spm=a2g0o.order_list.order_list_main.173.427f1802Qtbi1n

Buck converter: https://www.aliexpress.com/item/1005006431660781.html?spm=a2g0o.order_list.order_list_main.4.427f1802Qtbi1n

The entire circuit is switched off by a switch in the 230v AC voltage line that powers the power supply.


Basically my circuit works fine but I got one remaining problem:

When it was off for a longer time (so all capacitors in the power supply are discharged) and I switch power on the ESP wont start up.
So I have then to switch off and on again THEN it will start up fine.

By now the 3.3V from the buck converter is connected to the 3.3V pin of the ESP.
Earlier I had the buck converter set to 5V and connected to the VIN pin of the ESP having the same result. Also pushing the voltage highter to 7V didn't change anything.

I also tried to attach a 4.7uF capacitor from 3.3V to ground at the buck converter which also didn't help at all.

Therefore.....

  • What might be the reason I need to switch power on twice to make the ESP start correctly?

  • What can I do to resolve that problem?

Is this true no matter what sketch is uploaded?

Also, nice diagram. Say you have the grounds connected not shown.

a7

Some pins of the ESP32 must be high or low at boot.
Did you check that. See this page.

Disconnect one device/sensor at the time, and see if it starts up then.

I would prefer 5volt on Vin (safer).
Leo..

What is the voltage on the ESP's 3.3V pin when it is refusing to start up?

Try removing the soft-start capacitor circled in this image:

See datasheet page 5: https://cdn-shop.adafruit.com/datasheets/MP2307_r1.9.pdf

Alternatively, try a different step-down converter.

My reasoning is that the slow-rising soft-start feature of the converter causes problems with the controller coming out of Reset while Vcc is still insufficient for reliable operation. It's likely that the buck converter you use, uses the recommended 100nF soft start cap for a slow voltage ramp-up of 15ms during switch-on. If you remove it, or replace it with something much smaller (e.g. 10nF), power on will be quicker or virtually instantaneous and you may prevent this startup problem.

1 Like

Since I didn't use them tjey should all be fine.

About the 5V VIn as I said that's what was my idea at first as well but it had rhe same Problem and my LEDs would have had too much voltage at the ESP Pins which work at 3.3V so I changed it.

You mean while the starting should happen or after it didn't start and hangs itself up?

Would need to measure it.

That sounds like a reasonable ideaof the problem solution.

Is that known for having caused problems before?

Will try that this evening once I get home.

I've run into similar problems at times, yes. That's how I got in this train of thought.

Ok I removed that capacitor and it instantly worked.

I need to test it multiple times after all capacitors have discharged again but for now it seems that solved my problem.

I would never ever have figured out that theres such capacitor I suppose.....
Thanks very much!

I will give feedback if it still wont start up in the future tests but as said I think its fixed now!

1 Like

Hurray!

Startup behavior can be confusing/frustrating sometimes.

1 Like

Yeah that little capacity really was the problem!

Thanks again!

1 Like

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.