Earthing rectified DC

Thread Starter

shortbus

Joined Sep 30, 2009
10,045
Having a disagreement with some one on "earthing" the out put of a bridge rectifier in a circuit. He has a transformer from the mains, with the secondary going to a bridge rectifier and is then connecting the "negative" of the rectified DC to the chassis of the item. The chassis is also wired to the earth/ground of the mains. He claims that negative is ground so it has to be done. The rectified voltage is ~100VDC.

I always assumed that this was a bad idea. That it would then eliminate the 'galvanic isolation' of the transformer. Neither of us are professionally trained in electronics, so I want a second(or more :)) opinion to settle the disagreement.
 

Sensacell

Joined Jun 19, 2012
3,444
The secondary winding of a mains transformer is (should be) isolated from the primary. The secondary is a floating voltage source- the energy is transferred by the magnetic field alone.

Ground away.
 

nomurphy

Joined Aug 8, 2005
567
When creating a single-ended supply, such as +100V, then the negative side of the bridge can be grounded. However when using a bridge to create a split-supply, such as +/-50V, then the bridge will not work properly if somehow grounded -- but, the repective side of the filter caps can be grounded.

When I use a three-prong AC plug, I always connect the circuit side to ground (green wire), which means the secondary side of the xfmr is "earth" ground. Otherwise, the secondary or "circuit" side will float and there will be a potential between the circuit ground and earth ground which can cause a shock.

With dual prong AC plugs (look at old guitar amp designs), there is typically a high voltage cap from one of the lines to "common" and a switch to reverse the polarity. However, these are not as effectively safe as the three-prong plugs that include "ground".

In DC-DC isolated systems, there is usually a high-voltage safety cap that connects one ground to the other.
 
Last edited:

crutschow

Joined Mar 14, 2008
34,418
As Sensacell noted, the transformer output is isolated so you can connect either output polarity to ground or not as desired. Ground in neither negative or positive, it's neutral.

Often the chassis is connected to earth ground (mains ground pin) for safety but the electrical common is not (but the electrical common is connected through a large resistor to the earth ground to drain away stray charges). That way if you touch the chassis and a hot voltage on the circuit, you will not be shocked. A small capacitor is also sometimes added between circuit and chassis ground to minimize hum in audio circuits.

Connecting the electrical common to the chassis is usually only done when the circuit involves RF signals and you are using the chassis/case as a shield to minimize stray RF radiation.
 

#12

Joined Nov 30, 2010
18,224
Short answer: Yes. You can attach any one(1) point of the secondary circuit to the earth ground carried in a 3 wire, single phase system, as used in the USA.
 

Thread Starter

shortbus

Joined Sep 30, 2009
10,045
Thanks guys for the answers! I guess I was wrong. Again.

I read some where it was not a good idea to ground the secondary circuit, in case of a fault in the mains connection. Without the secondary side grounded there would be no connection to a mains fault. So much for information on the net.

Crutschow's idea of the large value resistor seems to make the best sense to me.
 

radiohead

Joined May 28, 2009
514
Instead of putting your faith that the ground lug on the outlet is properly grounded, I would recommend pounding at least a 6 foot ground rod, then using that as a proper earth ground.
 

vk6zgo

Joined Jul 21, 2012
677
Having a disagreement with some one on "earthing" the out put of a bridge rectifier in a circuit. He has a transformer from the mains, with the secondary going to a bridge rectifier and is then connecting the "negative" of the rectified DC to the chassis of the item. The chassis is also wired to the earth/ground of the mains. He claims that negative is ground so it has to be done. The rectified voltage is ~100VDC.

I always assumed that this was a bad idea. That it would then eliminate the 'galvanic isolation' of the transformer. Neither of us are professionally trained in electronics, so I want a second(or more :)) opinion to settle the disagreement.
The Mains Earth is connected to the Neutral & an Earth stake at the Mains inlet to your house.
The Earth (originally called a 'safety" Earth) is to prevent accidental electric shock.

With a metal cased device with the Mains Earth connected to the case,if there is a fault inside the device which connects the Active line to the case,the result will be a blown fuse,not a dead owner.

You can connect one side of the rectifier output,or one side of the secondary,even if you don't rectify it,to the case,& hence Mains Earth,without losing your safety protection,as the Active line is still isolated.
On the other hand you can float both sides if you like.

If someone has fouled up the Mains cord & connected the Active to the case,you have lost isolation,but if that has happened,you have a lot more problems than that!:D

Many devices are double insulated,& do not use safety Earths.
 

Lundwall_Paul

Joined Oct 18, 2011
236
radiohead has a great point about ground rods. I have a house that was built in 1970. All of the outlets have hot, neutral and ground, but I discovered that half the ground wire within the electrical box was cut off. The outlets had no ground. I also discovered that the main house ground rod by the meter had a very loose connection.
 

Thread Starter

shortbus

Joined Sep 30, 2009
10,045
Didn't want to open this again but have been thinking about it. I don't think I explained things correctly in the beginning.

This circuit is for the spark generator of a EDM machine, a machine that"erodes" metal to make a shape or remove a broken tap or ETC. Basically it uses a pulsed DC voltage between a shaped electrode and a work piece, with a dielectric fluid to remove the unwanted metal. See the attachment for a better understanding of what I mean.

Due to the need to change electrodes during the cutting process, there are times that the working parts (electrode and work piece) will need to be touched. While this may look like an unsafe thing to some, it is a very common industrial process and almost anything that is molded of plastic, is made in molds that use this process.
 

Attachments

Last edited:

vk6zgo

Joined Jul 21, 2012
677
Didn't want to open this again but have been thinking about it. I don't think I explained things correctly in the beginning.

This circuit is for the spark generator of a EDM machine, a machine that"erodes" metal to make a shape or remove a broken tap or ETC. Basically it uses a pulsed DC voltage between a shaped electrode and a work piece, with a dielectric fluid to remove the unwanted metal. See the attachment for a better understanding of what I mean.

Due to the need to change electrodes during the cutting process, there are times that the working parts (electrode and work piece) will need to be touched. While this may look like an unsafe thing to some, it is a very common industrial process and almost anything that is molded of plastic, is made in molds that use this process.
Yes,your original posting was pretty much a red herring!
It would seem that your idea of "floating" the rectified output would make the 100V pulses slightly safer.
The idea of touching the electrode and work piece while operating seems strange.
Wouldn't you have to cease operation to do this,& turn off the device?
Aren't the working parts hot?
 

#12

Joined Nov 30, 2010
18,224
I'm having difficulty with the top drawing. The "machine" is shown as earth grounded but the "electrode" seems that it must be isolated from the earth ground of the "machine". The pulse circuit does not show whether any part of it is earth grounded or what impedance (resistance) might be between the pulse terminals and the case or the most negative rectified voltage. That is the crux of the biscuit (if you are familiar with Frank Zappa).

If the most negative rectified voltage is about zero impedance to the negative pulse terminal, the top drawing is perfectly good, but I would build this so the "electrode" is earth grounded and apply negative pulses to the "work". That would put me in direct agreement with your second drawing. The internal ground on the most negative rectified point must be removed to set this up so that there is no voltage anywhere except the "work".
 

THE_RB

Joined Feb 11, 2008
5,438
...
The idea of touching the electrode and work piece while operating seems strange.
Wouldn't you have to cease operation to do this,& turn off the device?
Aren't the working parts hot?
EDM doesn't really "touch" the electrode, it charges a cap and dumps the energy across the electrolyte gap which is very small, which removes a tiny amount of metal.

The electrode is constantly cycled up and down to flush the electrolyte and waste, and the electrode height is set at the point where the energy pulse is just able to jump across.

It can do some impressive stuff like carve a square hole through something extremely hard like a magnet. Generally it doesn't get hot as there is lots of electrolyte and it's constantly flushed past the hot spot.
 

Thread Starter

shortbus

Joined Sep 30, 2009
10,045
Thanks RB for explaining it better than I seem to be able to.

The Voltage in the gap ionizes the fluid and then the ionized path allows the current to follow and erode/melt the metal. Then the gap spark is extinguished and fluid rushes back in and causes the metal particles to be flushed away.

The gap resistance during this is around 1 -1.5 Ohms, depending on the length of the gap.

In the drawings I was trying to show that by connecting either side od the DC circuit to earth ground it actually made it able to get shocked accidentally . The originator of the circuit shown claims that by connecting to earth it makes it more safe. And I still don't agree with him and his disciples. This circuit is from a book that is being sold to the public.
 

THE_RB

Joined Feb 11, 2008
5,438
Thanks Shortbus I understand better now what you are asking. Your diagram as from the book relies on the electrode being iisolated from the machine, which is how I would do it.

That way the metal target (work) and the machine can both be connected to ground -ve and likewise the electrolyte and all pumps etc.

Your suggestion diagram at the bottom would mean the metal "work" item has to be isolated from the machine, that is hard as it is probably a slab of metal bolted to a metal machine, also it means the electrolyte would be "live" with a -ve voltage and could never touch the machine or grounded metal body pumps etc.
 

Thread Starter

shortbus

Joined Sep 30, 2009
10,045
The fluid is petroleum based so no problem there. It only ionizes at the gap which is ~.001-.002 inch.

The electrode holder is usually the insulated/isolated point in a machine like this. And for some types of materials the polarity is changed on the machines.

I guess my real reason for saying not to earth the DC side of the circuit is, to me it just seems better to only expect the electrode and work to have voltage on them.
 

THE_RB

Joined Feb 11, 2008
5,438
You can isolate the work if you like but there is no need when it can just be grounded, you said you would have the electrode holder isolated anyway.

Often EDM is used for things like eroding a broken thread tap from within a hole in a metal block or a cylinder head etc so it's good to have both the work and machine grounded as then workpiece can just be bolted or clamped to the metal machine using the normal metal bolts and clamps.
 
Instead of putting your faith that the ground lug on the outlet is properly grounded, I would recommend pounding at least a 6 foot ground rod, then using that as a proper earth ground.
I know this post is old, but here it is. It is a very BAD idea to drive a 6 foot rod into the ground to bond your appliances or your breaker box to ground. You said NOT to have faith in the ground lug within the breaker box, though by introducing a better ground than that of the power company's utility ground located at the base of their pole or underground feed is a horrible idea and I will tell you why.. It is the same reason why you NEVER puncg a ground rod on the side of a swimming pool. It creates an attraction to lightning and could be deadly.

Here is why you never do this... The motor on all swimming pool pumps or any other appliance are connected in several locations. Every connection to the ground wire, in the home at the ground bus, then again in your meter box, then another connector at the weather head, then again at the pole at resistance to the ground circuit. There isn't a ton of resistance, but the truth is that lightning electricity follows the least path of resistance. If the lightning were to choose a path, you want it to choose the utility pole because all transformers have lightning arrestors connected to them which protects the system from feeding into one's home.

This ground rod at the pole is where you want lightning to hit, but obvious choices for lightning will be to trees instead since their roots are moist. If you connect ground rods to your electrical system anywhere short of the utility pole, you are only asking for trouble.

It's best not to earth many thingas even though the NEC tells you to, in fact with new 4 wire feeds, you don't have to , but the old 3 wire feeds require this and it is stupid. The neutral is grounded at the pole and it is not intended to be a backup if the neutral breaks, it is only supposed to be bonded to ground so it has no voltage and has a reference to ground at 0 volts.

there is absolutely no advantage to the additiona of any ground rods unless you get shocks from the electrical system's ground wire. If this is the case, you call the utility company and ask them to add a better ground at their pole. The NEC code says that (2) 10 fot ground rods be added within 6 feet of the base of the utility pole for proper grounding. Sometimes linemen cheat because the ground is too difficult to punch a rod and they cut it in half before driving it in. This is bad, they need 10 feet for a reason, that is so it touches the moist ground and conducts. If the drive it into sand there will be no advantage, so it is advised to spread the rods out and ground them deeper even if they have to dig a hole first to touch the conductive layer of Earth's crust instead of the dry sandy and non conductive areas.

This is a dangerous thing to tell folks. Always follow the connections in a grounding system to add resistances so you are not the chosen one for the lightning. it is critical for you and your child's safety!
 
Top