Why Would You Want an Electrical Isolated Ground?

It sounds like they want a separate isolated ground bus system with a single connection between it and the main site ground. This star point earthing (or grounding) configuration is common (in power, instrumentation, electronic and hazardous area earth systems) and simply put is used to limit or prevent ground loops, and voltage potential differences during earth faults. Some example information as follows.

http://www.google.co.nz/search?hl=en&num=100&q=instrumentation+earth+ground+star+point&meta=

http://www.idc-online.com/pdf/Papers/Joe_Zullo.pdf

etc. It should also be explained in your local electrical wiring regulations handbook or standard (it is in ours).

Hats off to Joe Zullo's wonderfully wriiten paper. Buy that man a beer!

Here is my question.

How am I to deall with the grounding of the negative connection of the remote 24VDC battery system providing 24VDC power to the control panel?

If I tie it to dirty ground at the battery, then bring 24VDC to the control panel and tie the instrumentation to ground then a massive ground loop is formed.

Is the solution to install an isolated ground beginning at the control panel and run back to the battery negative grounding point?

Thanks,

Bill

If I tie it to dirty ground at the battery, then bring 24VDC to the control panel and tie the instrumentation to ground then a massive ground loop is formed.

Could you not ground this connection (if required) at the control panel end (thus limiting any potential ground loop problems). Do you need to tie the negative 24V battery terminal to ground at all, as given the information you have given for the application I wouldn't in this case?

What do your local electrical wiring regulations (and applicable equipment manuals for your control panel) say about your specific application and grounding requirements?

Ground loop current form voltage in resistive ground wires and become a serious reference zero (common) voltage problem. Isolation lets the reference potential become a new one to which it is connected and trusted.

I agree but as a electrical panel builder we had to modify the implementation of the "isolated ground" a bit for "practical" purposes.

We do have a separate grounding circuit for the instrumentation but we now ship the panels with an obvious link to the "safety" ground. This is needed because the customer's electrician often forgot to tie the "isolated" ground to the clean grid.

Now, if they forget, the instrumentation is still grounded. If they do know what they are doing and have a separate "clean" ground, they connect it to the "isolated" ground point in the panel and remove the link to the "safety" ground. (hopefully)

If they forget to remove the link , it is still safer than having all the instrumentation floating.

I agree with Lhannon in that terms are confusing. Rather than using 'insulated' rather than 'isolated', though, I'd use:

- 'ground', pure and simple, to mean safety ground (against dangerous voltages); and

- 'common' for signal ground.

'Common' is ... uuuhhh ... commonly used to avoid all this trouble, but 'isolated ground' is even more frequently used, unfortunately.

Thus, we'd have 'ground all enclosures and panels' meaning that they'll be connected to a ground grid; and 'tie all device commons to a single, insulated, common point within each panel and ground, and ground that common point through a wire used for specifically that purpose'.

DZ

think of a simplistic repair, if each system has a separate ground system, you don't have to interrupt the other systems, nor does failure interrupt other systems, electrical, that is...

Nooooooo ... :O

Grounding has nothing to do with being able or not to shut off separate systems. The POWER feed is what determines that (one power feed to separate systems shuts them all off, no matter what the grounding is).

It's perfectly OK to have several systems tied to a single 'signal' ground point that's NOT the system-wide 'common' ground point. The thing to understand is that the 'signal' ground is only a reference ... ideally, there should be no current running through a signal-ground conductor to the ground-grid that it's connected to, since that would cause the signal-ground voltage in a panel to be different from the 'true' ground voltage at the ground-grid (voltage = current in ground-conductor times resistance of ground-wire).

For similar reasons, every source of 'signal' power should be signal-grounded to a single point in a panel .. no daisy-chaining grounds from one device / rack / power supply to another, else their ground-voltage-references could become different from each other.

And again, NEVER connect safety-grounds and signal-grounds.

DZ

I agree, but I got the impression that the comment referred to the possibility of breaking a connection inadvertently while removing some component of the circuit. This was in fact addressed in the NEC from the equipment grounding perspective and there is now a requirement for the equipment grounding conductor (a non current carrying conductor) and the grounded conductor (current carrying conductor commonly referred to as a neutral)to be connected in such a way that removal of a device on a multi device circuit, like an outlet or switch does not break the connection for the remaining devices on that circuit.

But getting back to the three referrences to ground. One being current carring, one being the equipment grounding conductor, and the third being the signal ground.

If these are not referenced together at some point as a common point of connection by being bonded together, then almost anything including the conductivity of the earth itself within a few feet of identical ground rods will always cause a difference in potential between the two seperate ground systems. That always results in a flow of electrons.

And the first thing someone does to try and eliminate the interferrence that the difference in potential causes, is to lift the equipment grounding conductor resulting in a serious and life threatning situation should something go wrong.

A equipotential ground plane distributed through out the facility assures that everything that references earth, does so on the same plane so that there can never be a potential between any conductor that reference ground for whatever its purpose. Be it current carrying, non current carrying, or signal reference.

I am not saying however, that there is ever the need for an inter-connection at the device itself. Say, by the use of a grounding strap between the isolated ground and the equipment ground in a device or panel box. The need to stay seperated remains the same. And failure to do so results in major issues for both the electrical distribution folks and the signal distribution folks. but just as many, if not more issues arise form the failure to properly bond and reference all of these systems in the proper manner at the proper place in a distributed and interconnected system.

I have seen more issues caused by a incorrect application of the intention of a isolated ground system, than any other code violation. Specifically when it involved the use of a seperatly driven ground rod not referenced to anything else, being used to create a "Isolated Ground".

It is only isolated in the sense that it uses insulated and seperate conductors that are in addition to, but not in place of the equipment grounding conductors and are not run seperatly or outside of the raceway containing all of the other conductors. And must be referenced by mechanical means to the same reference used for all other references to ground. The earth by itself cannot be used as a reference unless the points connect to it are bonded together as a single system.

Distribute it however it is needed, but reference it all to the same potential. Earth is not the same potential even within a few feet of itself.

Good answer . It needless to say that thereshall be a common ground for electronic circuitsused in controle panel. Suppose it is totally isolated , thereis every possiblityfor entry ofa Spike during Groud potential rise in in regular earthing system for safty. To avoid this only this common ground of Controle circuitry in connected to to the safe earth at only one point to have reference to common earth.Whenever ground potential rise even in Kilovolt range, the common earth for controle circuitry also got elevated to the safe earth's potential and concequently thereis no possiblity of flash or spike jump over to the controle circuitry.The intention of providing insulated wire also for providing reference only at one point. If you use bare conductor, It again become a multiple reference and make the purpose defeated.

A good example is A crow sitting on a power line near crossarm. Nothing happens as long as it sits only on enrgised wireonly. even if Ht line touches the LT Line, nothing happens to the crow. but if crow attempt to touch the crossarm, it become fatal.

Here our Electronic Circuit is similar crow and we are placing this stand on the potentially dangerous common earth system( Safe earth system) through a single common reference connection through an insulated wire.This will safe guard the controle circuit componants from the abnormal ground potential rises in safe earthing sytem

AN Isolated ground bus bar allows you to control the RF(noise) of your power supply: if your systems are totally common you will experience difficulty isolating RF interference in unique components. ie.... computer control moduals and electrostatic discharge....

LCD monitors and Audio/Video systems a extremely sensitive to Radio Frequency interference they typically display it as static with audio or snow with video. not a problem if you suffer from glocoma or tineitis. but if your in a airplane and you can't make out the land below or above, you would think you might like to be able to clear the snow from the screen or be able to hear the towers tone inflections for control of your craft.

Hi rabat,

It looks as if the designer of the control panel has experienced problems with 'noise' being induced into the processor side of the panel from field wiring having multiple earthing points.If the processor is sensitive to this sort of disturbance then all sorts of glitches, hang ups or even damage can result.

It would be interesting to know what sort of an electrical environment the panel will be used in. If there are further out stations with distributed I/O modules the isolated ground becomes an essential feature.

As 'QUEUE' rightly points out there is also the potential problem with rf signals and other transients disrupting the processor.

Finally there is also the issue with protecting the processor side componentry under fault level conditions within the power circuits. A large fault to earth can create a substantial surge in the earthing system which if shared with the processor could easily spell disaster.

Massey.

This is why I am saying that there must not be a separately derived reference. In the event of a fault or a lightning strike, having all ground planes bonded causes the entire plane or planes to rise at exactly the same potential. If there is no potential difference, then there is no flow regardless of how high they rise. If there is any difference of potential between the sources, then they will rise at different values and there will be flow between them.

Hi lhannon,

Indeed you are correct the issue with ground rails(note the plurality) acting in parallel is not only the generation of inductive loops (you can't have a loop with one end open circuit),but, when a power side fault situation arises there will be no chance of some of that energry transferring back to source via the network of grounded conductors. Therefore a reduced risk of a potential difference presenting to the low volt DC rail infrastrucure componentry.

Sounds simple but to be absolutely certain that no 'bridges' to the external earth have occurred anywhere in a large installation requires a meticulous approach to the field wiring and a rigorous testing prior to commissioning. Even the sheaths of signal wires must obey being connected only ot one end and the other must be properly insulated in the sensor heads etc. A job for Controls Electricians not 'Floor board Divers'.

We were doing this 30+years ago in DC operated signalling cubicles!!

Nothing is really new,experience is the key to avoiding pitfalls generated out of ignorance in the practical application of technology ....and always will be so !

Think on.

Massey.

3 out of 97 Aye!ΦR√ corn bread r round.

YOUR 5.2 MESSAGE READS VERY MUCH LIKE THE INSTALATION PAGE FROM A LIGHTENING ARRESTER SHIPPING BOX. WE NEVER THOUGHT OF THIS SPECIFIC SLIGHT DIFFERENCE IN A GROUND CONNECTION, UNTIL WE HAD A LIGHTENING STRIKE AT ONE OF THE COMPANY MICRO WAVE TRANSMITER TOWERS.

IF YOU HAD SEEN THE HOLE BURNED IN THE STEEL CONTROL CUBICLE WALL WHERE THE GROUNDING LUG WAS BOLTED IN. YOU WOULD BE INSTALLING AN INSOLATED STRANDED BUS WIRE FROM YOUR COPPER BUS BAR TO THE ACTUAL GROUND ROD DRIVEN IN THE EARTH, THAT HAS BEEN TESTED WITH A MEGGER.

AMEN!

It sounds as though this is the same thing that is labelled "instrument earth" or "clean earth" locally.

The principle is to ground/earth all the heavy power-consuming devices, including the <...back panel...> to one connection for electrical safety, and then a separate connection from the <...field wires...> onto an electrically separate ground/earth, then running each of those separately to the plant's central ground/earth point. What it means is that a short duration fault current of some thousands of amps on the safety earth/ground doesn't inadvertently appear on the instruments via the <...field wires...> and fry them while the circuit protective device(s) clear the fault.