2
\$\begingroup\$

I am confused about the ground part of a voltage converter.

enter image description here

This is a typical voltage converter. As you can see the the main voltage is 230V, and the output is 5V. Look at the output 0V, which is connected to the diode node, so I know, yes, this is the ground node.

Here also comes the question: As we know, for the main voltage power plug, especially in Europe or Asia, there is a ground line in the plug, so how I am going to do with the GND line in the main gate? Where is it in this circuit? There are just two lines input!

Also for Europe, there is only two inputs in the plug, but also, if you look carefully, there is actually a GND node in it.

enter image description here

If I simulate the main power voltage in the circuit by the sinusoidal source, how I am going to do that?

  1. One sinusoidal source power, one end is grounded, or
  2. Two sinusoidal source power, opposite connected, and the node between them is grounded
  3. Any other option?
\$\endgroup\$
5
\$\begingroup\$

The voltage converter circuit shown, like most (safe) DC adapters in domestic or retail use, is an isolated power supply - in other words, the voltages are isolated from any reference to the power supply lines, including the building earth line.

From the circuit's point of view, the only reference voltage is the 0 Volt on the output side. Voltages are measured from a convenient reference, and electronic devices (usually) only care about this reference level. This does not need to be referenced to building earth, and should not be.

The isolation protects the users of the electronic device, since touching any exposed metallic part of the device will not cause a shock: The voltage will not find a path back to the mains supply through the body.

Yes, you will find some transformers, especially high power ones, to have a high resistance connection from the secondary side to the supply earth, so as to allow a leakage path for the secondary side voltages to reference themselves loosely to building earth.

This kind of high resistance path, or a similar mechanism, could be used to reference your "isolated" output side of the power supply, or the circuitry connected to it, to building earth. However, this is of limited merit and very occasionally done, other than for relatively high power equipment such as a drill. Note that this won't leave the output circuit properly isolated any more.

\$\endgroup\$
4
  • \$\begingroup\$ thanks a lot, there is a ground in the wall plug(maybe it is the building ground),I mean for the building wall outlets, maybe they are all designed to automatically connected to the building ground, right( or isolated , so then what the ground node for? ) then if I design the first circuit, so I just not connect the building earth, that is enough ? \$\endgroup\$
    – alan
    May 10 '13 at 17:08
  • \$\begingroup\$ @alan Your circuit should simply not connect to the wall-outlet ground. \$\endgroup\$ May 10 '13 at 17:18
  • \$\begingroup\$ @Anindo : not always : there are some circuits that should be connected to safety ground. \$\endgroup\$ May 11 '13 at 14:39
  • \$\begingroup\$ @BrianDrummond Agreed. I was handing out the simple answer, without the caveats and exceptions :-) \$\endgroup\$ May 11 '13 at 15:05
2
\$\begingroup\$

The ground connection of wall outlets is a safety ground, not a reference ground. I believe you can connect safety ground to the metal enclosure or omit it for double insulated appliances.

\$\endgroup\$
2
  • \$\begingroup\$ so you mean the metal connection in the outlet----the safety ground, actually not the reference ground, so the safety ground voltage is not zero? then what is the voltage on the safety ground.. \$\endgroup\$
    – alan
    May 10 '13 at 16:56
  • \$\begingroup\$ @alan: Voltage is relative, not absolute. If you jab the red probe of your voltmeter on the safety ground, the voltage your meter measures depends entirely on where your black probe is currently connected. Voltage is potential difference. labelling earth as 0V is completely arbitrary. \$\endgroup\$ May 10 '13 at 22:05
2
\$\begingroup\$

It's usually easier to think of the input side safety "ground" as PE, or protective earth - it's more representative of what's actually going on.

enter image description here

This is a typical TN earthing system. There is a hard connection between the neutral and the protective earth at the generator location.

In real life:

In a non-isolated power supply, anything on the primary side of the isolation transformer (referenced to the lines or neutral) are to be kept separated from the protective earth by an isolation barrier.

On the secondary side of the isolation, it's OK to connect the low-side (the 0V path) to earth if you wish. Most small wall-wart types of power supplies omit this connection (they're said to be floating output) - many PC power supplies have this connection.

In the simulation world:

You can simulate without the protective earth connection. You may need to add a 100 Meg resistor from one of the primary side inputs to the 0V rail if you're getting weirdness or convergence issues.

\$\endgroup\$
2
  • \$\begingroup\$ This is a wonderful answer. A side aspect of this is that power supplies with floating outputs won't cause ground loops. I am not sure though if this is the main reason why it is now almost universal to omit the connection to earth. \$\endgroup\$
    – Alexander
    May 10 '13 at 19:37
  • \$\begingroup\$ @Alexander Earthing through the chassis usually means that there isn't any ground loop issue (it's often a better conductor than the power supply itself). \$\endgroup\$ May 10 '13 at 21:08
1
\$\begingroup\$

While many LV circuits are isolated as Anindo says (and should then be "double insulated" and marked as such) there certainly are cases where your circuitry should connect to wall socket ground.

If the LV circuitry connects to exposed metalwork, for example, then it ought to be connected to the wall socket ground - not via a high resistance but by a "fault current rated" (high current) connection.

A consequence of this is that if the isolation in the transformer breaks down, connecting the LV side to AC, the result is a high current flow to ground, immediately blowing fuses or tripping breakers, instead of a piece of equipment at 240V waiting for you to touch it...

A lot of hi-fi equipment is (or used to be) built this way. You can generally call the safety ground voltage "0V" but take care that in your hi-fi system, only one unit (usually the amplifier) connects the audio 0V to the safety earth. Multiple connections to ground form loops, and magnetic fields intersecting these loops induce voltages which usually appear as hum.

I would simply simulate the AC supply as a 110V or 240V sine voltage source, with the "neutral" end connected via a low resistance ( 1 or 10 ohms) to your safety ground. Whether I connect that to my LV circuit 0v depends what I'm simulating - a double insulated circuit or not.

\$\endgroup\$

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.