0
\$\begingroup\$

I am using this large 24V power brick (rated for 9.2A) to power a small circuit (an arduino controlling some RGB LED strip, drawing less than 0.5A).

If I physically pull the plug out of the wall, the circuit runs on the transformer's stored energy for about a second before shutting down, which makes sense to me.

If the power supply has been plugged in for a long time, and I switch it off at the wall outlet, then the circuit continues to run for a very long time-- more than an hour-- before shutting down. But if I physically pull the plug, it still shuts down in a second or so.

  • The switch on the wall socket is working correctly (tested with a desk lamp), and as I understand it, this switch interrupts the 240VAC live conductor.
  • The power supply connects to the live, ground and neutral conductors on the AC side, and on the DC side provides a +24V and an AC ground connection
  • Because I use an Apple laptop, I know the ground and neutral wires in my home are not at the same voltage

So, where is this circuit getting its power from when the live AC is disconnected?

\$\endgroup\$
3
  • \$\begingroup\$ To clarify, yes, 0.5A is a guess, and probably on the high side. But it's running the arduino itself (through a linear regulator dropping 15V), plus 70 LEDs in the RGB strip. I don't know how that strip is wired internally, but it seems to be the arduino that turns off first, and according to the docs it will power off when the supply drops below 7V. \$\endgroup\$
    – bobtato
    Commented Feb 13, 2015 at 19:50
  • \$\begingroup\$ Is there a surge protector or wall dimmer in the circuit? \$\endgroup\$
    – Spehro 'speff' Pefhany
    Commented Feb 13, 2015 at 19:54
  • \$\begingroup\$ No, although obviously there are various other devices connected to the ring main... \$\endgroup\$
    – bobtato
    Commented Feb 13, 2015 at 20:25

2 Answers 2

Reset to default
1
\$\begingroup\$

Okay, given the comment that it's just a normal UK outlet, I'm going to take a wild-a** guess and suggest that your outlet may be wired incorrectly so that the wall outlet switch is breaking the neutral and thus the Y capacitor in the supply is continuing to conduct some current to through the ground connection and hot.

This is based on some assumptions about the UK residential electrics that may not be valid, and I'm sure a right-pondian correction will be supplied if necessary.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
1
  • \$\begingroup\$ The switch is connected to the live pin ... they are part of the same unit so you can't get that wrong ... however, which wire (L or N) is connected to them is anybody's guess without testing the socket or pulling it off the wall to look... I can't think of a better explanation. \$\endgroup\$
    – user16324
    Commented Feb 13, 2015 at 23:58
0
\$\begingroup\$

Capacitors in the power supply.

\$\endgroup\$
6
  • \$\begingroup\$ Although your description of the problem does sound odd. Have you measured (safely) the AC voltage at the input when the switch is turned off. Also I remember there are saftey requirements for large power supplies, maybe there's a circuit in there that's draining the caps when you unplug. Mostly conjecture though. \$\endgroup\$
    – Some Hardware Guy
    Commented Feb 13, 2015 at 19:07
  • \$\begingroup\$ Or depending how that switch is wired maybe there is a sneak path from line to gnd \$\endgroup\$
    – Some Hardware Guy
    Commented Feb 13, 2015 at 19:12
  • \$\begingroup\$ An hour seems like a long time for the caps to delivery 0.5A of current... (Though it's not clear, it maybe only the arduino that runs for an hour.) I like your sneak path idea. \$\endgroup\$
    – George Herold
    Commented Feb 13, 2015 at 19:16
  • \$\begingroup\$ I was thinking it's probably not really drawing 0.5A but he's just guessing that. But yeah it reminded me of getting zapped debugging a circuit where the neutral was broken... There was still a path from live to gnd through me. \$\endgroup\$
    – Some Hardware Guy
    Commented Feb 13, 2015 at 19:18
  • \$\begingroup\$ Yeah, it's the change in behavior when the plug is pulled that seems suspicious-- the power supply shouldn't even be able to tell the difference. I thought about investigating the situation on the AC side, but then I had visions of a tombstone reading "I should have just posted a question on Stack Overflow". It's tricky to jam things into fortress-like UK power sockets because of the safety shutters. \$\endgroup\$
    – bobtato
    Commented Feb 13, 2015 at 19:59

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.