Harmonics

I am in full agreement with tcmtech.

From my experience too many get by on and take for granted electrical feeds that will work somewhat independent of proper grounding as long as the neutral will carry the backfeed, but always there are problems if there is not a good solid ground, especially when major amp draws are involved.

Good point tcmtech. You also get a GA from me.

Keep in mind that all ground currents should be undesired or fault currents.

There is no simple magic wand that can be waved to solve an interference problem like this. I would start by identifying the interference path by running the welder with all of your computers running on the batteries of their UPS. If this still causes interference then no amount of power filtering will help. You might find that your internal wireless communication network cannot cope with the RFI produced by this TIG welder. But clearly the least expensive method will be to identify how the interference propagates (broadcast, communication cables, power cables) before buying hardware. Once you find this out, you will have to decide on a course of action. Making the PC's tolerant to TIG welder noise maybe the lowest cost per application but will have to be reevaluated with every new piece of office equipment. Preventing the TIG welder from polluting the air waves or power grid will require more expensive equipment, but need be done only once.

I don't know what "PCs have fits" means, but most UPS systems are "off line" and won't go "on line" unless they detect a loss of 60/50 Hz for a certain number of ms. Then, they go "off line" again very fast if the AC is restored. This will work well for "normal" loss of power, but might not do anything for you with very dirty power. As a last resort, a "true on line" UPS that is actively producing the AC output 24/7 would give you clean power.

Note that this is the "last resort". Checking grounds AND NEUTRALS first is a much better way to go. If poor grounding of the welder allows the welder to screw up neutral then you might be equipped with the right idea but still barking up the wrong tree.

It might be helpful if you add a few details about what "PCs have fits" means (reset, data/program corruption, power down, run with bad video, etc.)

A couples of questions...

1) How is 120 volts Ac provided for the PCs? You said it was closed delta, as such, there is no 120 volt available. Is there step down transformers?

2) Is this a corner grounded delta system,? If not, that is more then likely your problem.

I did my apprenticeship in a very large printing plant fed from a 240 volt closed delta 3 phase system that was not grounded. When we installed new laser graphic scanners (this was in the 70s - the scanners were HUGE), we had all sorts of problems with ground loops.

We solved the problem by installing a grounding resistor between "ground" and one of the phases. This gave us a "corner grounded delta system" and the resister limited available fault current to a manageable level.

We then tied all of the "grounds" together, including the neutrals of our step down transformers for lighting and hotel loads and the problems went away.

Thanks for all the input,It is a corner grounded system,no xformers,Welder install was done buy an electrical contractor,but also same guy that told owner,installing a sub panel just for welder would cure problem.As the Pc having fits,the small ups go to battery and monitors show distortion,don't know if data is lost,will look at grounding system closer,at this point making and assumption that licenced contractor knows what he was doing as far a grounding,but we all know about that. Have s[spoken to some other people and they suggest,changing service to a 120/208 system.

I would not argue with the statement that changing to a 120Y/208 system would not help the situation. Having said that... I would question if such a change is needed at this point, as not enough data has been presented yet.

You still have not told me how the 120 volts is derived for the PCs. There has to be transformers if it is infact a 240 volt delta system as there is no way to get 120 volts other then to center tap one of the three phases. Which does not go with a "corner grounded" system.

More investigation work is called for...

All of Millers welders have considerable line noise filtering built in. But all of that filtering is very dependant on the primary earth ground circuits being adequate. In some applications adequate literally means an solid earth grounded point within so may feet of the machine with a large enough size of wire from the machine to the ground rods and the correct size and number of ground rods for the local soil conditions.

Without the correct ground circuit those machines basically become high powered RF hash/noise transmitters!

As far as what the contractor says, To be honest I have met more electrical contractors that have never opened an electrical code book or electrical physics book since the day they got their licence than those who have by a very wide margin!

Having a contractors license is very similar to most people having a drivers licence. They passed the test 20 years ago and have kept the renewals current but that doesn't mean they have read the rules and regulations books since the day they got it! If anything most couldn't come close to even passing the test again if they where given it on the spot!

to tcmtech

SLAP!!!!

from an electrician with a license that closes the books long enough to eat, sleep, and work

Hehehe! Some one is offended by my comments and had to post as a guest!

A good licenced electrician would have never thought twice about my comments being that they would have considered it was their competition I was referring to and not them!

As I referred to MOST. If your offended that likely puts you on the wrong side of the MOST list.

Since your so sensitive here is a joke for you.

How many licenced electricians does it take to change a light bulb?

9 plus one inspector, $17002.98, and four months time do it.

It takes a month for the first electrician to show up and give you an estimate. Then it takes 8 more electricians two months to put in a new service from the utility pole, through the whole house up to the light socket for the cost of $17000.

Then one month more before the inspector comes by and signs off on their work.

He then hands you a new bulb and tells you that you have to do it yourself because of a liability clause in the work contract that stipulates that a light bulb is not the contractors responsibility due to hazardous material handling regulations ($2.98 CFL bulb of course) but is the end user/home owners responsibility.

(Oh, and you still have to wiggle the light switch 'just right' before it will come on just like it always did before anyway.)

If you have ever had any electrical work done by a licenced contractor you know full well what I am referring too!

I appreciate tcmtech patience in posting. I too was a licenced electrician now working as a senior manager. I have faced inspectors who gave licences to people who bribed them but not to those who really deserved; that was too bad. I think tcmtech did not offend any one; what he said is some realities.

Thanks. And thanks to everybody else as well it is appreciated!

In my working career I have met many very good knowledgeable and honest contractors in many fields that if needed I would recommend to anyone.

Unfortunately I have met more unskilled, unqualified, and dangerously corrupt and crooked ones in the same time line plus a few bad inspectors included as well. I have seen licenced contractors steal customer property, and other contractors property, bribe incompetent inspectors to pass horribly wrong and dangerous work, and far worse. I have even been asked and expected by a former employer to play along because it was good for everyones business. And no I didn't play along. All the while the customer was getting shafted from front and back by everyone.

To me this thread likely points out the OP has a questionable contractor (likely incompetent and or greedy too) on his hands.

In my line of work experience the service system is the last thing that gets upgraded when odd electrical problems occur with office equipment, not the first. I would have started with taking a basic look at the welders grounding system and the buildings grounding systems first.

After that I would have likely put in an isolation transformer or line filtering system for the office supply lines. A 1 - 2 KVA isolation transformer, or a number of smaller individual ones for each computer, is far cheaper than a 20 - 30 KVA transformer for a welder system. Plus small ones could be set up as a simple plug in and go devices as well. No service line work would have been needed.

If the problem went away the noise is in the lines and can be cured at the welder with a some off the shelf commercial line filters installed in the welder AND in the service panel systems. If it didn't the noise is in the grounding system which any competent electrician should always suspect first and never last!

LOL!

I like your style.....

The "low life" will probably not come back.....

Who is also not man enough to post with his CR4 name.....!!!!

Your comment was simply rude, probably inaccurate as well, so it brought nothing to the problem fix......eg. Totally useless!!

wouldn't the noise come from the large currents in the welding cable like a spark gap transmitter?

Is the welder 3 phase or single phase?

Single phase supplied from a closed delta will unbalance the delta. If the system is a grounded B-phase more problems will orrur. Grounded B-phase is no longer allow per NEC.

Check the wire sizes to small of wire will create problems to.

From: Tim Hawley Master Mech.

All of you gave great feed back to dg1258

I would recommend as well, a ground rod drill through the cement floor, drove down minimum of 4 foot into earth and connected to ground bar in service panel.

Also look for floating grounds on AC connections on cold water pipes etc...

If this doesn't solve the problem, use an isolation transformer to feed the AC to the welder.

Problem solved.

Best Regards,

Tim

I didn't feel you were being quite fair, nor accurate in what you wrote.

I am placing here what I understand for earthing, correct me if I am wrong please.

Correct earthing, each unit (motor, PC or welder for example) is connected to the earth busbar once.

But if an earth path developes as in the red wire shown in this diagram, some nasty things can happen as in here:-

There are other possibilities that can cause loops, I have just picked a simple version....

Tell us what you think please.

PS I liked also the idea of rolling the cable completely out, especially if on a metal drum. My wife nearly burnt our house down many years ago when she used a rolled up cable for the wash dryer!!!!

If your getting a voltage differential on the earth grounding circuit between your two devices you have something wrong in one or more of the devices.

Multiple earth grounds are just that. There are more than one point of physical contact between the devices grounding wire that goes to your buss bar and the actual physical earth/ Tera firma/ the big thing we live on.

Basically the short answer would be that every device in your drawing can have a earth grounding rod next to it that ties it directly to the physical earth itself in addition to the ground line going to the buss bar you have shown. This is fairly common procedure in correctly designed electrical systems as I have seen it.

I cant be more specific to any or all instances or applications without actually printing the whole NEC code books article 250 section which is pages and pages long.

Its situations like this that are why stuff gets screwed up in real life implementation of designs. Someone takes an incomplete, inacurate, or condensed version of just part of the rules as given by someone else instead of consulting the rule book like they should have and then builds their system on incomplete misunderstandings of information.

If you need to know the details get a copy of the 2008 NEC code book and read it. Or wait until the next one gets printed in 2011. It gets updated every 3 years.

Sorry but thats the only real and correct answer I can give.

Bravo. Precisely the point of this and many other comments here at CR4, follow your local code. Not the code of another location, nor somebody's vague recollection of what a code states. Buy the code book itself for your area, maybe even take a class on how to translate the code into English (very helpful class IMHO).

I have had this occur before and accidentally solved the problem by laying the stinger cable out as straight as possible on the floor. The cable had been rolled up on the hook on the cart. Maybe it acts as a big antenna when coiled up?????

Beware the misplaced decimal point.......

dq1258,

I'm afraid there will not be a cheap solution to this. I can't see how you will avoid installing an harmonic filter.

The welding equipment will continue distorting the supply voltage to the building and creating harmonics. This will be 'seen' by everything else in the building and possibly by some of their 'electrical' neighbours. The contractor should be ashamed of himself. A separate sub-panel was never going to be the solution.

Rob

Driving Ground rods next to each device can work, but this solution can create problems. High ground resistance can create ground loops between the devices.

The ground rod should only be used at the power supply source for the building.

Tig welding is a given around nuclear power plants. Whenever any Tig welding was underway, the nuclear instruments would go crazy. Not good. However, the Navy never found a solution and, as far as I know, the instruments still have problems.

Essentially a Tig is a spark transmitter and the spark is not at all constant and it tends to be of a fairly high frequency IIRC.

Given the relatively poor shielding of a standard office PC, I am not surprised they react. There are shielded PC enclosures designed for use on factory floors. Hard to imaging those working well in an office environment, however.

Maybe the office, if it isn't too large, can be placed in a Faraday shield. Probably not practical.

Maybe a faraday shield around the work area?

Just wondering if you solved your problem with the TIG vs computer trouble? What fixed it?

beware the misplaced decimal point.......