[stirling]

I have my plasma table (and a couple of routers) up and running and thus far working well. It has a THC300 and a pilot arc plasma unit and is under Mach3 control. Running it in my shop I have no problems - but when I took it to a friends (much bigger) shop with *lots* of other machines of all types running, it kept throwing a limit error in Mach.

Motor drive cable from controller to steppers is a single 12 conductor screened cable with screen connected at controller end only to chassis. Limit and home cable is a single 4 conductor screened cable, again with screen connected at controller end only to chassis. Noise affects only the limits - motors work fine and don't miss steps.

I've read here and elsewhere about using a ground rod as the one and only ground point for *everything* but I'm confused as to how this would figure given that all equipment, computer, plasma unit and controller box are already grounded through their (UK) three pin mains plugs.

Like I said - confused as to how to stop the limit from triggering all on it's lonesome.

[jimcolt]

If the problem is with RF noise from the plasma (and...there are pilot arc plasmas that ionize with high frequency......and pilot arc plasmas that use blowaback, non high frequency start...you did not elaborate on which type you have)....the RF noise wants to find its way to earth ground as efficiently as possible. If it has to travel throught the mains cable with the power conductors....it has a good long distance to travel in parrallel...where it can inductively couple to the mains.....which allows for voltage anomalies to get back into your system.

By using a simple grounding circuit for cable shield drains and for all equipment chassis....taking a short route to a closely placed ground rod....you minimize the possibility of inductive coupling. Noise can enter a systems through inputs, outputs and through the electrical mains. Often the source (transmitter) of the noise is the torch leads...between the plasma and the torch....if this is routed through a power track with low voltage control cables (like limit switches?)...then the problem could be starting here...and the fact that you moved the machine and somehow changed the grounding scheme....could have affected the situation.

Electrical noise is tough to deal with.....the most proven method is with very simple grounding circuits with an eart ground rod.....although there are few, if any people that have a good understanding of why this works!

Best regards, Jim Colt Hypertherm

[stirling]

Thanks for your reply Jim. Thing is I know it's not the plasma that's causing the trips. a) It doesn't happen in my shop - I can cut all day without issue, and b) it happens even if the plasma is switched off or even diconnected in my freiends shop. It's only the limit switch circuit that suffers. If I turn off the floating head switch it will sit there all day without tripping with or without plasma running so it HAS to be RF inducing in the floating head switch cable. The screen of that cable is connected to the chassis of the controler box and all earth connections are identical in both locations i.e. via the mains connectors.

[jimcolt]

I'm not familiar with the electronics package that you have.....doesn't make a lot of sense why it worked in one location....and not in another!

Jim

[stirling]

Hi Jim - like I say - all I can think is that in my shop the *ambient* RF is lower/non-existant but in the busyer shop with welders and such constantly firing up it's rife. Thing is I thought with the screen of the floating head switch cable earthed it would all be taken care of. Back to the drawing board I guess.

[jimcolt]

You may have to figure out a way to optically isolate or filter the inputs from the limit switch. I have run into situations over the years where a cnc machine was simply moved from one bay to another....a machine that had been working flawlessly for years...and all of a sudden it is beset with electrical noise issues. We don't have noise issues now with the cnc controllers that are specifically designed for industrial plasma cuttig applications (such as Hypertherm Automation and Burny controls)...these units are industrial hardened...and are designed to work in high frequency/industrial environments.

Most of these new class of controls are even wirelessly networked......files can be downloaded from office computers....and with the Hypertherm system...up to 12 technicians can get online and drive the controller (but not fire plasma or move the machine) from anywhere in the world fro troubleshooting.

Good luck getting it sorted out! Keep us posted if you find something definite!

Jim

[Al_The_Man]

Electrical noise is tough to deal with.....the most proven method is with very simple grounding circuits with an eart ground rod.....although there are few, if any people that have a good understanding of why this works!

Jim Colt Hypertherm

You may want to check for the implications of a ground rod set up in U.K.
The practice used to be that a ground conductor was never supplied by the service provider and the service was grounded at the supply transformer only, which can be some distance away from the end user, and a Earth leakage trip at the panel was used.
This could mean that the E.L.T. ground conductor could be bypassed in a fault condition.
This is different to the system used in N.A. where the service supplier provides a system ground and it is grounded at the service entry.
Al.

[stirling]

Al - a good point well made. Certainly I've never personally come accross a dedicated ground rod for an individual piece of equipment. Pardon my ignorance of Stateside ways but I must admit whenever I've seen instructions for using ground rods I've wondered why and assumed there was something different in the ways ground is implemented.

[Al_The_Man]

It has been some years since I was involved in the Electrical industry there, but I remember when ELT's were mandated due to the prevalent use of non-metalic water supplies which ruled out the then practice of using the water supply pipe as a ground conductor.
Al.

[mkweiss222]

Im an electrician in Texas. There are multipal ways to supply service power to a shop (house, etc.) from the ''pole'', or service drop, or whatever your situation may be. The codes differ from one city to the next. Also if you live outside of the city limits there are generally no inspections therefore alot of stuff gets ''bootlegged''. Some peoples mentality is if the lights come on then its good. Now on a single phase service, ive seen (and ran) both 3 and 4 wires. The three wire service uses a separate ground rod/plate at the building. In the 4 wire type, the ground has to travel through its conducter out to the transformer, where it is then earth grounded. Although both of these are acceptable methods (depending on where you go), the 4 wire type system may not necessarily be good for our cnc plasma operations. Also in some instances where subpanels are used the neutrals and grounds get bonded together. What im getting at here is that there are many ways to wire a shop (right or wrong). And just because its a large commercial/industrial shop doesnt necessarily mean that it was wired correctly. I know that in the U.K. your wiring is different than ours but it is based off of the same basic principles. So I would just run an earth ground to the table and be done with it.

[Al_The_Man]

I have never come across the UK method in N.A. That is the only place the service supply is grounded is at the star point of the supply transformer, wherever that happens to be.
If the service supply DID happen to include a ground conductor it could NOT be attached to the installation by any means.
I did not mean to infer a ground should not be run to the table, it was just the question of local regs now as to separate ground rods.
Al.

[mkweiss222]

I have never come across the UK method in N.A. That is the only place the service supply is grounded is at the star point of the supply transformer, wherever that happens to be.
If the service supply DID happen to include a ground conductor it could NOT be attached to the installation by any means.
I did not mean to infer a ground should not be run to the table, it was just the question of local regs now as to separate ground rods.
Al.

Al,
I think you took me wrong. I wasnt trying to step on your toes at all.
I was just illustrating how one shop could be wired differently from another.
Now im a little confused about what you mean by the service ground could not be connected to the installation by any means...could you elaborate on that?
Thanks

[Al_The_Man]

I am quoting the situation that existed some years ago, so it could have changed.
The typical residential system in the UK was that a 3 phase distribution transformer with a grounded star neutral which fed an area using a phase and a neutral for every 3 homes, the next phase and N the next three etc.
The service installation consisted of live and neutral only, no ground conductor, per-se.
The INSTALLATION ground conductor was connected to a qualified ground, IOW if a water supply pipe was used, a ground path resistance reading was obtained using the pipe ground and the neutral as the two conductors, the circuit was completed through to the transformer, through the the star point to ground, through ground, back to the pipe.
A resistance megger was used to obtain the necessary low ohmage reading.
Only then would the service be connected, any local connection between neutral and the installation pipe ground would result in a non-connection.
With the advent of non-metalic water supplies, the regulation was that a ELT or GFI unit was installed at the service entrance and a ground rod was used for ground return to the star point.
I would be interested to hear if things are still that way from any UK 'Sparks'.
Al.

[wildwestpat]

Hi folks

The UK regulations have expanded over the last few years to take into account the chages in materials used. Where the incomming mains is via a lead or other metalic covered cable that covering may be used as the earth for the installation. However the incomming mains supply might be a plastic or pitch protected cable or by over head supply and thus a local earth rod / tape or plate must be provided to the requirements as laid out in the current IEE wiring regulations. To carry out the full testing of the installation the local earthing is tested using earth rods and the appropriate instruments.

Also in the UK the other utilities gas and water might be supplied via non conducting plastic pipes and instead of the metal ones used in the past. The wide spread use of plastic for pipes has given rise to the changes in the wiring regulations which now require all exposed metal work to be bonded together as well as to bond to the incomming earth of the supply or the approved earth rod / burried plate if the supply is not by a cable with a metal outer sheath (lead) - pitch covering or plastic sheath on the mains supply means an effective earth rod is required.. The bonding is to be provided by a conductor of the correct cross sectional area and the use of metal conduit as the earth conductor is not permitted. The outer cases of all machines - water supply points - metal sinks etc are to be bonded and connected to the main incomming supply earth point.

I hope this clears up the role of the UK earth rod for permanent installations for equi potential earth bonding. There is a cop-out clause that requires RCD to be fitted where the supply might 'reasonably be used outside and many electricians will opt to fit an RCD to protect all power sockets. To this end the supply of consumer units (fuse board replacement in the new parlance!) are fitted with one or two high current RCDs. Good practice is to have the lighting separately connected so that an earth fault does not disconnect the lights - sensible as the lighting should be connected to a lower rated circuit breaker / fuse.

The UK regulations are complex and require authorised people to carry out the work. Each country will have their own regulation and I have only outlined the UK requirements - if in doubt get proffesional advice - electric shocks can kill and CNC is too much fun to risk not getting all those projects done.

Albert is basically correct but the regulations have altered a lot with the advent of non metalic supply pipes and PVC covered underground cable. The earth impeadance testing is now a specialised task and some very fancy instruments have been developed to speed up the testing.

Regards - Pat

[Al_The_Man]

One thing you do not see in N.A. is an instance I had where I spec'd in a storage heater system for a customer in an older (UK) residence, the service supplier insisted that 3 phase be installed to cover the load, for which they brought in the 3 phase.
The 1 phase dist system here precludes the advantage of 3 ph installation, residential anyway.
Al.

[wildwestpat]

Hi Al

Here in the UK the use of three phase brings another raft of regulation. The switch board requirements are different which is kind of logical but the rules on using three and single phase in the same room are a bit more stringent and the equi-potential earthing becomes all important. For my own satisfaction I all ways bond to gether all exposed metal work including any water or gas supply pipe work. This prevents any unpleasant shocks. For good measure for my own workshop I use an earth spike as a belt and braces approach along with residual current breakers on all supplies.

I also fit emergency stop buttons on the wall at convenient places. These are wired to the main switch board where they are used to switch off all electrical supplies. Obviously these are not used often but are tested once a week as part of the health and safety checks. Also these wall mounted emergency stop buttons are not to be confused with the 'E' stop buttons mentioned earlier in this thread. Emergency stop buttons for the switch off everything require a mains contactor and a means of resetting the contactor when it has been deactivated. A minor but again a safety feature is that the power remains off in the event of a mains failure until reset. I use a key switch to reset and the key is held in the safe not the switch!

Hope this helps others stay safe. Regards Pat

[Al_The_Man]

I recall when the reg was brought in to bond across all water outlets sinks/basins hot/cold/waste etc, in those days the supply was copper and the waste lead.
An old plumber taught me how to wipe a lead joint, this guy could make it look like it was turned on a lathe.
Came in handy later for lead sheathed cable.
Al.

[Torchhead]

First try increasing the Inputs Signal Debouncing (MACH General Config) to a higher number. That will filter out noise on inputs. The THC inputs (which HAVE to be high speed) are excluded from the debounce parameters in MACH versions after 1.84. If that fails to do the trick you need to look at the type inputs on your breakout. Most BOB's use opto inputs for all the input lines and that offers better noise immunity if you keep the INPUT ground isolated from the table.plasma ground. True opto isolation has it's own floating voltage and return (incorrectly called ground) and won't transmit the noise across a common ground because there is none.

Ground is part science and part art. What works for safety and code may not work for grounding noise for machinery. Any arc cutting/welding process is noisy. The trick is to either shunt the noise to a local ground or keep the two sides isolated. You can get several volts of noise (and even AC) between one side of a shop ground and the other.

TOM caudle
www.CandCNC.com

[stirling]

Thanks guys for your illuminating discussion, but what I was really after was: if I use a ground rod and the system is already grounded via the mains connectors etc. whatever the scheme, am I not liable to create a ground loop?

Tom, I've tried upping the debounce in Mach to the point where it actually affects whether the floating head triggers when it should, but with no effect on stopping the noise. Re: optos: The motor drivers I use have them built in but I don't have optos on the floating head switch line which is where the problem is. That said I'm confused as to how this would help. If a signal is being induced that is enough to create a *rogue* active high then surely the optos would just pass this through. I understand where optos have a role in protection by not allowing damaging voltages through, but don't understand how they would help here. I've tried to keep my signal *grounds*/returns, separate from true ground and tried to ensure all screens are connected at one end to true ground but I guess I need to re-visit the whole thing.

[wildwestpat]

Hi Stirling

The best way of looking at the extra earth rod is as an insurance policy.

Yes there are earth curents flowing in the soil. In urban areas the voltage between earth rods inserted several meters apart can be quite large.

If you have the time and inclination you could push several earth rods into open ground and measure using screened leads the potential between them. That you are most likely to find is a differnece in potential - a small direct current difference and a larger alternating voltage component at mains supply frequency. This should be less than a couple of volts if the soil is damp and the rods are of the same material. However in the event of a fault the earth current may rise particularly if the fault is a supply line fault in the locality.

The reason of earth bonding is to get all exposed metal at the same potential. This makes it certain that contact with machines or other metal work will not cause an electric shock which is unpleasant. This is called equipotential earth bonding and is covered in the regulations of most electricity supply companies and may also be covered in local health and safety regulation or recommendations.

The local wiring regulations will specify the gauge of wire to be used - a good rule of thumb is greater than the current carrying capacity of the supply fuse or circuit breaker plus a bit. Also do not rely on metal conduit for the earth continuity but use a separate wire.

The local earth rod serves as an extra measure of safety against abnormal earth current faults in the locality - or - failure of the earth connection to the public supply. Failure of the public supply earth is not unheard of - the earth connection is made via a clamp to the metalic outer casing of the incomming mains cable if it is an underground supply and the outer metalic covering is in intimate contact with the ground i.e. is not plastic or tar covered. In the case of an over head supply on poles the 'earth' connection is either provided locally by a rod or plate in moist ground or may be by an extra conductor from the supply company. The mains meter and supply earth are the province of the electricity supply company and should be respected as such. However a close inspection - but not interference - of the electricity supply companies earth bond may be appropriate as the connection may have been damaged. If necessary the connection should be remade and tested by a competent electrician recognised by the electricity supply company.

There will always be earth currents between any burried metal objects due to the currents flowing in the upper few meters of the soil. All your 'extra' earth rod is doing is ensuring that the exposed metalwork of all machines in your workshop are at the same potential and the potential is the same or nearly so as the soil outside the workshop door. If you have ever worked in a mobile workshop you will be aware of the need for the earth rod or mat as it is possible to get a shock on stepping down from the vehicle due to the supply earth not being at local soil potential.

Sorry if this is long winded but I hope this clears up any confusion. Belt and braces are best for pesimists - earth rods are the braces for the electricity supplys earth belt!

Added a post script - The earth loops talked about in electronic terms are not the same. In electronic circuits a single point is normally designated as the zero signal voltage point and all supplies used in the equipment and signals refferenced from that point - OR - care is taken to avoid earth loops inducing unwanted signals into other parts of the equipment. Since this 'earth loop policy' is isolated by the mains power supply to the electronic equipment by the transformer or electronic regulator that derives the various voltage supplies required. There is some times confusion about mains and signal earths - treat the mains earths as bonded together and the individual electronic systems as unique to an individual machine. The electronics then being 'earthed' to the machines supply earth by a single wire or flexible braid. This will avoid most if not all earth loop problems.

Kind regards - Pat