Current Peak of power cause shutdown

Hi every one !!

I have a roller blind controlled by a Arduino (a ESP8266 to be precise), and those two are powered by the same source of electricity.

My problem is that when the roller's motor is starting, the peak of current causes a kind of surcharge of power which interfere with the power supply of the Arduino, which reset the Arduino.

So I am looking for a solution, as I was explained a capacitor should do the trick, but I don't understand where and how to put it. I will try to post a image for better understanding.

I am sorry if my explanation seems poor, my English and knowledge in electronic are not that good.

Best regards

The same problem was solved in the industry. We installed a diode, in serie, on the positive wire to the controller. I suggest a Schottky diode in order to loose as little voltage as possible. Between the diode and the controller we added a bulk capacitor.

For better understanding

Do You say that the 220 volt line goes low at motor startup?

What do you means by "goes low" ?

and the 220 to 5 volt converter drops the 5 volt at motor start?
Try a bulk capacitor between +5 volt and GND (minus).

Yes I am pretty sure there is a drop a current somewhere at the same time that the motor start.
When I try to powering the arduino on a différent circuit (same house but different circuit breaker), I don't have this problem. All my blinds are isolated on the same circuit breaker.

Do you think a 6.3V 1000uF will do the trick ?

The power supply I am using for the arduino already have a bulk capacitor on the output.

The voltage drop is only a theory unless you measured it.

How long are the 5 volt cables, from this supply to the controller?
Are they twisted?
If You have a bulk capacitor You can try, but as You tell, there is one already.

That would be a good reason to post images.

I can see all the LED go off then on during a fraction of seconds when the motor start.

barely 2inches, not twisted.

Does a 6.3v 1000uF will be enough ?

FYI I put some RC snubber circuit to prevent the contacts in the relay to get stuck.

There is no shielded, grounded enclosure? That is a serious shock and fire hazard.

Considering your problem, though... you should not run mains wires so close to the relay board.

But, a reset condition can do that... it can also be produced with a pulse, not just a pause in power.

The enclosure is programmed, I am still in the testing phase.

Have you tried it with a different source of 5V?

Yes, it is stable.

Then you need to re-engineer the supply.