I have an Atmega328p based node with RFM69HW radio, controlling a relay. Power is supplied using a cheap 220V to 5V Chinese power supply (see photo). A 30W led bulb is connected through the relay to mains power.

The problem is, when I run test, which turns on and off the relay once per 5 second, sometimes (once per 5-10) the system just reloads. I checked every part with an oscilloscope and found that the power supply has voltage drop to near zero, which causes the system restart. Checked with different power supply, everything works perfect. Also, without the bulb it works.

Can anything be done to fix the problem?

enter image description here

enter image description here

The schematic: enter image description here

  • 1
    \$\begingroup\$ The power supply is only 3.5 W. So I do not see how you can use a 30W led bulb. You seem to overload the circuit. Please add a schematic to prevent misunderstanding \$\endgroup\$
    – Decapod
    Commented Oct 29, 2016 at 5:56
  • \$\begingroup\$ A schematic will help us help you. \$\endgroup\$ Commented Oct 29, 2016 at 6:40
  • \$\begingroup\$ The bulb uses mains power. \$\endgroup\$ Commented Oct 29, 2016 at 6:41
  • \$\begingroup\$ Added the schematic \$\endgroup\$ Commented Oct 29, 2016 at 6:51
  • \$\begingroup\$ Are you sure the PSU and MCU match? \$\endgroup\$ Commented Oct 29, 2016 at 7:26

1 Answer 1


The situation is more clear now.

I expect that the incomming mains supply causes the problem and that switching the bulb creates a voltage drop high enough to make the power supply restart.

Remember that the led bulb has a very high inrush current due to the rectifier on the input of the build in converter. This could be more than 20 A. The connecting cable could be the cause

To verify this.

  1. Try a short and big connection cable.

  2. Separate the two systems and feed the buck converter from a different outlet.

  3. Try a standard bulb.

PS. I use the same power supplies and have no problems.

  • \$\begingroup\$ Thank you, will try these. When I change the PSU to another, no drop and restart occur. That's why I concluded that the PSU is the cause. \$\endgroup\$ Commented Oct 29, 2016 at 7:51
  • \$\begingroup\$ Changed the cable, testing now. No restarts so far. Previous cable was really thin and long. But why other supply had no problems? Better design? \$\endgroup\$ Commented Oct 29, 2016 at 8:15
  • 1
    \$\begingroup\$ The inrush current of the bulb is the problem with a long cable. That made the incomming voltage drop for a very short moment. \$\endgroup\$
    – Decapod
    Commented Oct 29, 2016 at 8:16
  • \$\begingroup\$ It does restart sometimes anyways. Is there anything else I can do? \$\endgroup\$ Commented Oct 29, 2016 at 8:52
  • \$\begingroup\$ Tried standart bulb. Does the same. \$\endgroup\$ Commented Oct 29, 2016 at 9:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.