Floating DC System

If the power supply is floating, there is no connection to ground. A voltmeter takes a very tiny amount of current to read voltage. If the power supply is truly floating and you measure to ground, there is no complete circuit, so you should measure zero. It would be the same as if you only connected one lead of the voltmeter.

If the power supply is not grounded but there is a high resistance path somewhere in the circuit, you would get some reading on the voltmeter, which would depend on the resistance to ground and at what point in the circuit this leakage path occurred.

Dear sir

thank you for your reply

but if the multimeter read +12v between positive and ground,and -12v between

negative and ground are this mean the system is grounded and how i know which

terminal is grounded positive or negative

thanks

What you have is dual supplies, +12v and -12v. This is common with analog circuitry (like amplifiers) where the signal needs to swing both positive and negative with respect to ground.

Case 1: If you place a fused jumper between the positive terminal and earth and then measured from earth to the negative terminal, what voltage will you measure?

Case 2: If you measure from the same earth point to the positive pole with the jumper still in place, what voltage will you measure?

Answers: Case 1: 24V and Case 2: 0V.

Measuring across either pole to earth should yield +/- half the output Voltage, if the voltages are unbalanced then there is leakage current flowing from the lower voltage terminal to earth. The only exception is if the leakage impedance between each terminal to earth is identical, a very unlikely event.

Dear sir

if the DC power supply system is not floating and its grounded how i know which

terminal is grounded positive or negative,and when we use floating system and ground

system.

thanks

Use the polarity or your voltmeter, normally it will read positive upscale if the red lead is plugged into the plus plug on the meter, and the red is on positive.

Dear kasrop,

This is a comment I made to a previous question on CR4....

It depends if one pole is earthed!

If you have a situation where a battery and its connected loads are well-insulated & not earthed at all, except by an earth fault detection relay, then I have known a plant sitting for years [with an earth fault alarm (One pole 40V to earth, compare 65V fault-free) - disconnected from the annunciator!] until the battery cells were washed externally to remove the traces of damp electrolyte which were causing earth leakage (more on one pole than the other).

The relay connected each pole of the 110V battery to earth by a 50,000 ohm resistor. The relay alarmed if the voltages to earth were unbalanced by more than a preset amount. Voltages of battery poles normally +/- 65V to earth (battery float charge 130V) with no leakage - consequently possible current, about 2 mA, into earth fault is harmless to most equipment. Compared to a "solid-earthed" battery, there is also a personnel safety advantage.

There is also a system (e.g. telephone supplies) where the positive of the battery is biased negative above earth by the earth fault detection relay integral DC source. This was because electrolytic corrosion due to leakage currents is worse at the positive end. Since the telephone end may be the fine wire of a coil, it is far more vulnerable to corrosion than a large earthed metal case. By biasing everything in the system negative a version of "cathodic corrosion protection" was achieved. There is also the advantage that everything in the system has a definite voltage to earth, to allow earth fault detection - the centre tapped system has the possibility that a point of the load which was at zero potential could be earthed without causing an alarm.

The current permitted by the earthing resistors is selected to be too small to spuriously operate or hold-on a relay or PLC input, but enough to give warning of a problem. Also the resistance is low enough to avoid an alarm due to small currents which are sure to exist in a real plant.

67model

I gotta say I'm a bit confused.

Your last paragraph states that "Measuring across either pole to earth should yield +/- half the output Voltage," (thus indicating that earth is the mid point of the power supply in your scenario) which I agree with if the supply is, as it appears to be, a dual supply both sides of ground/earth.

Your first case though, states to "place a fused jumper between the positive terminal and earth" - surely the fuse would blow if you were to do this as you have now shorted out the +ve side of the supply and the resultant reading between earth and -ve would be 12v not 24v as you suggest.

In Case 2 the measurement between earth and +ve should again be 12v, but only because the jumper fuse has blown.

A reading between +ve and -ve would yield 24v.

What am I missing? Perhaps I need re-educating here.

There might be a moderate resistance (multiple kohms) or even a nonlinear impedance between the ±12V power supply return and ground in an effort to reduce ground loop effects. That would be one plausible scenario that agrees with the OP reports. Then again the OP doesn't understand what they actually have so they may only think they are testing things correctly. We may also have a complete communication breakdown leading to the great shrug, I don't know.

A truly floating supply has no connection to earth and only two rails, positive and negative, therefore any measurement from either pole to ground would represent the the voltage drop across any stray impedance to earth. A dual voltage floating power supply would have 2 power rails and a neutral, all of which must remain isolated from earth.

The fuse will not blow if both poles are properly isolated from earth, since there is no path for current to flow. However, if there is already a path to earth from one rail to earth, a fault on the other pole will be the equivalent of a short across both poles. The jumper is fused with a properly rated fast-blow current-limiting fuse to protect any personnel from arc-flash as the jumper is applied.

Thank you for your reply.

You make this point -"The fuse will not blow if both poles are properly isolated from earth,"

However, in your earlier post you made this point "Measuring across either pole to earth should yield +/- half the output Voltage". This statement suggests that "earth" is the midpoint of the power supply in this instance.

It was this point that I have a problem with as, if your first point above is true, then how can you have any voltage reference to that isolated earth as you have stated there will be in the second point?

Unless I am misinterpreting the meaning of your earlier post as it reads, then the information given is incorrect.

If there is a leakage to an otherwise isolated earth then, unless it is of a very high impedance, or the power supply is a very large capacity, then any reading from earth to that compromised pole will show a depressed voltage to that of the other uncompromised pole, therefore your statement that "Measuring across either pole to earth should yield +/- half the output Voltage". is incorrect as the compromised pole voltage would be somewhat less than the other pole reading.

If the leakage is from the neutral point of the power supply to earth and is of some impedance then both pole voltage readings to earth would be lower than half the power supply pole to pole voltage. If the leakage is a full short circuit then the fuse in the jumper will blow, and the voltages will be as I suggested in my first post.

Dear sir

thank you very much for your reply

It is capacitive coupling to ground, generally you want to see equal voltage to ground on either leg, unless you know you have a good reason for unbalanced capacitance to ground ( i.e. one wire length is significantly longer than the other)

You also see this on ungrounded 120Vac circuits, you can read about 60V to ground on either leg if it truly is ungrounded.

The reading is indeterminate and will depend upon the make and model of both the power supply and meter, as well as their internal circuitry and condition. For example, a modern high impedance digital meter will pick up and measure any capacitively coupled electrical noise between the power supply and ground, but an old analog low impedance d'Arsonaval moving coil meter will immediately drain any noise and show only the true leakage voltage/current.

The best way to view the nature of any spurious voltage in a truly isolated system is to use a high impedance oscilliscope to determine the source of any leakage voltage/current.

Weird. This Q appeared about an hour ago - I replied to it, and even got a GA, then the Q and my reply disappeared, only to reappear sans my reply.

I made the points covered in #1 and #2, and added that a floating system will not be affected much by a single short of either rail to ground, while grounding the ungrounded rail of a grounded system can have nasty consequences (blown fuse/PSU, sparks/fire depending on protection and PSU), and will definitely remove power from the load.

I gave you the GA. You were spot on.

Something odd is happening behind the curtain.

Thanks - odd indeed!

"Something odd is happening behind the curtain."

I’ve had that happen too. My suspicion is that the server here has a bot that looks for specific words and automatically deletes posts containing them. I’m not sure what those are, but in another form I visit, what it does is to change the words to be all asterisks and the ones they do that on seemed very odd until someone explained to me that some words spelled in English are close or identical to those used as pejoratives in other languages. I can’t think of them right now, but if I do I’ll post them (in a way that doesn’t result in deletion).

This is indeed strange. I'm moderating this afternoon and I haven't done anything other than keep an eye on things.

We do have a list of spam terms that result in automatic deletion but only for users with fewer than 10 posts. Spammers generally only post once or twice. For anyone else a post will be reported as spam, but the moderator has to read the post and decide whether to delete or, more often, edit out something. So that's not what happened.

Please send a note to CR4 Admin if any of you notice the same thing happening, in this thread or elsewhere. Thanks!

Anything it flippin’ well likes, Mildred. That’s why it is an isolated system.

I got an electric shock the other day when I touched the door handle of Lieutenant Gruber’s armoured car. Does that help?