I have an embedded system and one of the PCBs in the module should be connected to a 5V power supply -> positive(+) and neutral(-). The NEUTRAL knob is directly tied to the GROUND knob in the power supply. I had a power supply with the current knob turned to maximum and a set voltage of 20V.

After turning on the power supply, I connected the ground wire from my PCB to power supply's negative (-). I left the VDD wire from my PCB unconnected to the power supply, it was just hanging around. The PCB required 5V input but the power supply was set to a voltage of 20 V. I thought since the VDD is not connected to the power supply yet, it should not be a problem.

Immediately, the ground connector on my PCB sparked, causing a decoupling capacitor connected to it to badly burn out, followed by a few resistors and LTC3404 step-down converter chip on the PCB. Then I pulled out the ground wire of the PCB from the power supply.

I do not understand how the connecting of just the ground wire from my PCB to power supply neutral could cause my capacitor to burn out, even though the power supply had higher set current and voltage than required by the PCB. So, was it the current on the neutral wire that burnt out my capacitor or is there anything to do with the higher voltage?

NOTE: I have been using the same power supply and PCB with 5V and the current knob turned up to maximum, for months and had no issues before.

  • \$\begingroup\$ Did you check the galvanic isolation of your power supply? \$\endgroup\$ – Jeroen3 Oct 5 '18 at 14:17
  • \$\begingroup\$ No I haven't done that. \$\endgroup\$ – Madhu Oct 5 '18 at 14:19
  • \$\begingroup\$ if you've connected a -20V power supply to a ground connector on a PCB and had a VDD connector on the PCB floating if the VDD wire touched off a grounding metal (the leg of your table) then you would have had 20V going through the Cap from ground and out the VDD connection that sounds like the frying that you described, id also say that you've fried more than just the Cap. Did you have a floating wire that could have made contact with something? \$\endgroup\$ – Jack Oct 5 '18 at 14:26
  • \$\begingroup\$ Perhaps you wasn't careful enough with the positive terminal or you shorted something on your PCB while the capacitors are still charged up. \$\endgroup\$ – Zy Gan Oct 5 '18 at 14:26
  • 3
    \$\begingroup\$ @Madhu. There may be a language problem in the question. "Neutral" means one of the mains wires that is connected to earth at some point in the electrical power distribution system. "Common" or "ground" is the reference point (usually supply negative '-'). Can you hit the 'edit' link under your question to clarify what you mean. \$\endgroup\$ – Transistor Oct 5 '18 at 16:21

The most plausible explanation is that you have both

  1. A faulty power supply, bench power supplies are nearly always designed to isolate the output from the mains inlet, it sounds like that isolation failed in your case.
  2. Something connecting your circuit to mains earth, possiblly a data connection, possiblly an osciloscope, possiblly just part of the circuit touching an object that is connected to mains earth.
  • 1
    \$\begingroup\$ I did have my embedded setup plugged into a picoscope and more wires to my computer's USB port. But I don't know if it was the scope that acted up! \$\endgroup\$ – Madhu Oct 17 '18 at 22:15
  • 2
    \$\begingroup\$ That was essential information. Your question says that you got sparks with only one wire connected to your circuit. You have many including one with a 5 V supply. Big edit required with schematic. \$\endgroup\$ – Transistor Oct 17 '18 at 22:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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