Grounding a dc power supply on the negative side...what's the purpose? 2

[bdn2004]

A DCS vendor has sent out an instruction that all of the 24VDC power supplies should have their negative terminal connected to ground.
I think they are marked 24V+ & 24V-. This system up/running no issues for a couple years now. What would be the purpose of such an instruction?

 

[HamburgerHelper]

Prevent your voltage from floating in reference to ground. I think by doing that you would limit th current that could be pulled during a short.

 

[Skogsgurra]

Sorry HH, I cannot follow your reasoning there.

A more likely explanation could be that there have been problems with interference and someone found out that there are less problems with the DC- grounded. A very common practice in Europe and many other parts of the world.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ScottyUK]

Better to have a known reference and be able to fuse the ungrounded pole at appropriate points than to have a floating supply where you have to fuse both poles because 'something' will eventually get grounded and you just don't know what or where.

 

[bdn2004]

I've always seen the + and the - on dc systems fused on both legs. Do we need to remove the fuse and replace with a dummy straight through copper bar in the grounded leg?

 

[itsmoked]

Yikes! It's lame to mark the 24V return (24VRTN) as -24V or 24V- as that implies it's a negative supply.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ScottyUK]

In floating systems both legs need to be fused because you can't predict where a ground fault will occur. Sod's Law says that the fault will occur on the opposite pole to an existing one, and if you use links then you find that the conductors act as fuses to protect the links. From experience, when there is a very large battery involved the conductors first emit smoke, then light. ;-)

Generally I would only fuse the ungrounded pole of a grounded system and use a link in the grounded pole.

 

[Skogsgurra]

Yes, Smoked. That mentioning of Minus has caused problems in many places. I think that it is there because a car battery has "-" and "+" and people constantly look for familiar things to relate to. Also, there ARE things like positively grounded systems.

So, who puts a fuse in the ground cable of a battery or power supply? Not me. And I guess no-one else. I don't put it in the ground lead (even if the terminal is marked "-") of a PSU either.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[krisys]

Grounding of power supply is required for the purpose of providing the reference. Now the question is why the negative terminal to be earthed?

After some years I understand that the +ve terminal being adapted to prevent the electrolytic corrosion of the conductors/terminals connections.

Generally for the telecom power supplies, power supply +ve grounding is customarily adapted.

 

[waross]

Why is the negative grounded and not the positive?
Metal case flash lights have always had the negative battery terminal grounded to the case.
Billions of flash lights can't be wrong.
Joking aside, almost-
Urban legend had it that this was to prevent corrosion.
Actually, historically, the first telegraph systems used a ground return. It was found that the ground electrodes acted as a crude electroplating process. The negative ground electrode tended to lose metal and the positive ground electrode tended to gain metal.
The central station grounded the positive conductor and all the remote stations grounded the negative side.
This gave a complete circuit, negative to positive.
The central station grounding electrodes always stayed in good shape.
The remote, negative, ground electrodes did tend to erode, but very slowly and if performance deteriorated due to a poor ground electrode, the operator could repair or replace the station electrode.
The combined area of all the remote ground electrodes at the telegraph stations down the railroad lines was collectively much greater than the area of the central station electrode so erosion of the negative electrodes was quite slow.
The central station electrode may corrode but it did not erode.
That was a long time ago.
For the last 75 years or so, it hasn't mattered whether the negative or the positive conductor was grounded.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[marks1080]

There's a good rational for using an ungrounded (floating) DC system. If this is what you currently are using I would think twice before changing philosophies. Sounds like you're dealing with electronic telecom equipment, so it's probably something the manufacturer is saying to do to protect the equipment. It's your responsibility to make sure that you are protecting the batteries.

 

[ScottyUK]

Ungrounded systems are fine in organisations with the skills and resources to hunt and repair ground faults down. In organisations which can't do those things they can be a complete pain.

I left one of the former organisations to go to work in one of the latter, and some very odd things were happening to their gas turbine control systems: three machines share one central battery and ground faults on one machine would cause problems on the others. For example, occasionally a fuse in the I/O would blow without apparent cause, and the fault would only occur in very specific set of circumstances during machine startup when certain I/O was energised. Ground faults existed on both positive and negative poles, but only made contact when certain relays were closed.

 

[GroovyGuy]

Good question,
I always ground the negative side of a 24Vdc power supply. I like the predictability of blowing only one fuse on a short-circuit. Also having the one side at ground potential makes trouble-shooting a tad simpler.

Having said that, 125Vdc systems (ie station batteries) are almost always floating. But these systems are always equipped with ground detection for both the positive and negative sides. If properly monitored, the ground can quickly be located and repaired.

Also the telecom industry uses a grounded positive 48Vdc system. I believe this practice dates back over a 100 years ago (and had something to do with corrosion?).

 

[mls1]

DC distribution systems are typically ungrounded for reliability. In any system, all ungrounded legs must be protected so ungrounded DC panelboards have two pole breakers (or three in the case of center tapped 250VDC systems) and fused protection has fusing on each leg.

What is missing from this conversation is why the vendor would call for grounding a DC system. The answer is most likely because grounding one leg will VASTLY reduce common mode noise in the system. You can talk to just about any process control person that has had a noise problem that after extensive troubleshooting the only solution they could find is to ground a leg. It is common. Before you choose to go against the vendors instructions you might want to get their buy off if you expect to have any kind of warranty support.

 

[Skogsgurra]

That aspect is not "entirely missing", see 1 Aug 16 16:26

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[bdn2004]

Not going against the vendor's instruction whatever gave you that idea? I'm just asking the question as to why. The answers here certainly seem reasonable.

 

[submonkey]

Hi,

We generally "earth" the midpoint of a 110V DC
system via the coil of a ground detection
relay. The system tolerates a ground fault
until the alarm from the ground detection
relay is acted on.

Apart from master battery fuses (200A), the
negative side of outgoing circuits is linked
and not fused. There are very few (any?) real
faults which will not be cleared by a fuse in
the positive. On the other hand, a blown negative fuse can result in some truly bizarre
behaviour as loads start interacting via
bits of conductor which ought to be the negative.

Thanks,
Alan

 

[krisys]

submonkey (Electrical),

Mid point earthing thru the ground detection relay does not make the system grounded. It is still considered to be ungrounded.

So in your case, not providing fuse in the negative side is actually a flaw. In my view, you better provide the fuse.

 

[ScottyUK]

The scheme you describe is safe only if the largest fuselink is smaller than the smallest conductor's fault withstand capability. Consider the following example:

A part of the DC system protected by a large fuse - say an emergency lube oil pump for a turbine, with a fuse of 300A or thereabouts - develops a solid earth fault on the positive (fused) side of the supply. There's no problem as long as it is the first fault and everything works normally. The earth fault alarm comes in but it's nothing to worry about, right?

Now add a second fault, this time on the negative side of a much smaller supply to, say, a protection relay cabinet with a 20A supply. That leg only has a link protecting it on the negative side, not a fuse. What is going to melt to clear the fault - the 300A fuselink, or the panel wiring?


Having seen this exact situation for real, the answer was the panel wiring melted through, disabling the protection for the generator and caused a fair amount of damage to other wiring within the panel. The 300A fuselink was unharmed.

 

[submonkey]

Hi Scotty and Krisys,

All the outgoing circuits have 32A fuses, and all the wiring is large enough
to safely blow a 32A fuse. If an earth fault occurs on one circuit #1, everything
keeps going. If a second earth fault develops on the negative side, the positive
fuse in circuit #1 blows.

Like I said before - it's hard to find a real case which is a problem.
Your 300A circuit example is interesting, do you often have DC supplies that large?

Thanks,
Alan