Grounding Rod For Portable Generators Necessary?

[ThomasH]

This discussion seems to have slowed down, but I hope someone else can comment on the above quote. I am currently looking into welders for running with a generator. The comment above has me curious about whether inverter welders need grounding rods connected to generators more than non inverter based welders...

It has been a while since I have seen equipment with a parasitic filter which is implemented with a capacitor linking an internal power source with a ground. This type of equipment can cause the chassis to have some voltage, depending on the exact voltage wave of the incoming voltage. The purpose of this type of filter is to filter out any transient spikes from the power source, and it requires the metal chassis (which can be touched by a human) to be attached to a physical ground.

So as long as the generator is well behaved (and I would look for one with a sine wave output or simulated sine wave), it should be fine. If it is one of the cheap ones, which can generate a bunch of noise (voltage spikes) from the dirty brushes, burned commutators, and a malfunctioning governor, it is best to ground it *if* you are plugging in equipment that needs a real ground to keep from possibly shocking you. This equipment could include power tools with a metal case or electronic gear.

A sine wave (like the power company supplies your house) looks like the black line in http://www.marine-deals.co.nz/images/inverter_sinewave.gif
The red square wave exaggerates what a stepped square wave looks like, but shows the difference in the type of what is generated. Better generators either generate a sine wave, or a stepped square wave with lots of steps. Typically, equipment can't tell the difference, but some more sensitive equipment can, and some can even be damaged.

Almost all "standard" generators under 10 kw certainly and quite possibly most under 25 kw are inherently "noisy" in the electrical sense and most of them simply do not have the rotational mass to ensure close frequency control OR close voltage control. On the other hand, generators which contain inverters have crystal-controlled frequency and will indeed hold the frequency within 1/2 Hz. from no load to full load. The voltage regulation is far tighter than "standard" generators at about 1% from no load to full load. In addition, it is the elimination of the transients that are common with non-inverter generators where the name brand inverter generators have the leg up.

Hope that helps.

 

[ishiboo]

I hope you take these slight disagreements as constructive and not offensive, as nothing is meant. But there are some thoughts for discussion here.

It has been a while since I have seen equipment with a parasitic filter which is implemented with a capacitor linking an internal power source with a ground. This type of equipment can cause the chassis to have some voltage, depending on the exact voltage wave of the incoming voltage. The purpose of this type of filter is to filter out any transient spikes from the power source, and it requires the metal chassis (which can be touched by a human) to be attached to a physical ground.

Two things to think about here:

1. Assuming you mean "physical ground" to mean ground rods, since power goes back to its source, this doesn't quite make sense... it would just return via the ground wire to the generator head.

2. If this concept were in practice, and if there was enough voltage and current available on the device chassis and thus the ground wire to injure or kill someone, none of these devices would work on GFCIs, since that would be a ground fault! All my computers, power tools, etc. work fine on GFCI receptacles. If it were not enough to trip the GFCI, it would not be enough for a person to be at risk. It would also likely not pass UL standards.

A sine wave (like the power company supplies your house) looks like the black line in http://www.marine-deals.co.nz/images/inverter_sinewave.gif
The red square wave exaggerates what a stepped square wave looks like, but shows the difference in the type of what is generated. Better generators either generate a sine wave, or a stepped square wave with lots of steps. Typically, equipment can't tell the difference, but some more sensitive equipment can, and some can even be damaged.

All mechanical generators produce a regular, true sine wave like the power company supplies and the black line. This is because AC power has a sine wave because that is how power is mechanically produced from an alternator - it's the natural output. Versus an inverter which has to make it using solid state components. The only thing that produces modified/square waves are less expensive inverters. More rotational mass and larger generators mean better response to sudden load increases, however, which will decrease low voltage situations when large loads are suddenly applied. Surge ratings are just as important as running ratings when warranted.

Realize small inverter generators like the Honda EU-series may not have a neutral/ground bond. The Hondas to my knowledge DO have a main GFCI breaker.

Once you add plain inverters to the mix, you run into more interesting issues, such as many cheap inverters producing 120v by having the neutral leg at +/- 60v, and the hot leg at +/- 60v, so the neutral has potential!

EDIT: The EU-series do not appear to have GFCIs.

 

[Tractor Seabee]

My Gen-Set is a Generac 17.5 KW. It runs power tools and my 50A input rectifier AC/DC welder just fine w/o a ground rod. I tried the ground rod deal and checked the potential from the N and G conductors and 0 volts, same from the gen frame. When connected to the house circuits it is bonded through N and G cord conductors to the house neutral and grounding systems. I have had no problems running all my electronics including the computers and various solid state appliances and power supplies, probably 50 HRs in the last 2 years. I have not had a GFCI trip on anything for a long time, but I test them once a year along with fire alarms and all that stuff you need to do on a schedule. I will leave all the theory to others and believe in the proof in the pudding I have experienced. Yes, I do know the codes also, both NEC and OSHA; no grounding of portables and only one grounding system for house connections.

Ron

 

[CobyRupert]

As I said, its main purpose is lightning/static dissipation... check out HiV's post above. If that was true, then why does a 100 amp service have a different size grounding electrode conductor than a 400 amp service?

The NEC requires a maximum resistance to ensure that the ground rod can participate in that fashion effectively, or two rounds must be driven if you cannot test.



The grounding rods serve no purpose there. They will not flow enough current to trip a regular breaker as they have a very high resistance compared to a "good conductor".

Where a metal item becomes insulated, the EGC connected to the metal case should allow the fault current to flow back to the main panel, meter can, or wherever neutral/ground are bonded. When this happens, the short travels over the EGC, the breaker trips and everything is fine. The ground from the ground bar to the ground rod, and ground rod itself, are not used for clearing the fault and unnecessary for that particular purpose. True! ...but, if not for the ground rod how would we get 2-120V phases from a 240V service? Also, when that short occurs and there is no ground rod, only a bonded "EGC" and the "floating" neutral like you talk about, if you are touching the metal, it's hard to say what voltage (and current) you will experience. Actually, if you can limit the coupling capacitance between the source and earth , it may be safe to touch it, as there will be no ground fault current using your body and earth as a path, although I wouldn't recommend it. I believe some European systems are ungrounded. Often in an industrial setting, on an ungrounded (delta) or high resistance grounded system this type of system is purposely installed so that the system can continue to run if 1 phase becomes accidently "grounded".


GFCIs are used to better protect people from wet circumstances or ground faults. Despite the name "ground fault", a GFCI does not use the grounding rods and need not have a grounding wire to function... in fact, they are the accepted way for replacing a 2-wire ungrounded outlet when no ground wire is present and its not feasible to get one there.

*****

 

[ishiboo]

If that was true, then why does a 100 amp service have a different size grounding electrode conductor than a 400 amp service?

Dunno, code and science are closely related but different. I'm guessing the assumption is a larger service would service a larger potential area. I'm not an electrician, one could probably explain better the reason for the code. Check Mike Holt's forums.

if not for the ground rod how would we get 2-120V phases from a 240V service?

Single phase 240V in the United States is brought in using two 120V hots (180 degrees apart) and one neutral (0V). The difference between those two hots is 240V RMS. Ground has nothing to do with whether 240v is available unless you're on a VERY archaic single wire service, used only in incredibly rural areas.

Also, when that short occurs and there is no ground rod, only a bonded "EGC" and the "floating" neutral like you talk about, if you are touching the metal, it's hard to say what voltage (and current) you will experience.

Neutral does not float, it's 0v potential. If 120v electrifies something metal, the purpose of the EGC is it completes a circuit to neutral, creating a short. Once you have a short, the breaker servicing that circuit blows.

 

[CobyRupert]

Dunno, code and science are closely related but different. I'm guessing the assumption is a larger service would service a larger potential area. I'm not an electrician, one could probably explain better the reason for the code. Check Mike Holt's forums.



Single phase 240V in the United States is brought in using two 120V hots (180 degrees apart) and one neutral (0V). The difference between those two hots is 240V RMS. Ground has nothing to do with whether 240v is available unless you're on a VERY archaic single wire service, used only in incredibly rural areas.
That's true but backwards (IMO). A single phase 240V is the output of one (1) utility transformer's secondary coil. The 2 legs of 120V are obtained by grounding (pegging to 0 Volts) the center (tap) of the transformer coil. Perhaps we are talking about the same thing, tomato/tomata, but the difference is the 120V is derived by splitting the 240V output of the transformer, rather than constructing 240V from (2) 120V sources like you state. Again, this may be tomato-tomata


Neutral does not float, it's 0v potential. If 120v electrifies something metal, the purpose of the EGC is it completes a circuit to neutral, creating a short. Once you have a short, the breaker servicing that circuit blows.

Unless there is a ground rod bonded to the neutral, the neutral will float and is NOT 0 volts with respect to a person (on the ground) touching it. If neutrals are not grounded (pegged to 0 volts) as you originally spoke about, now your (2) "120V" legs may be not be split 120V/120V , but the voltage split (totaling 240V) will be a function of the load (resistance) in each leg (e.g. L1=160V/L2=80V, etc..)

 

[ThomasH]

Neutral does not float, it's 0v potential. If 120v electrifies something metal, the purpose of the EGC is it completes a circuit to neutral, creating a short. Once you have a short, the breaker servicing that circuit blows.

Not true. V = IR. The current in the neutral wire (I) is greater than zero, and the resistance of the neutral wire (R) is greater than zero, therefore voltage (V) must be greater than zero.

In addition, neutral is the middle of what could be an unbalanced load, so it can contain some voltage potential due to the unbalanced load. Since it is the return leg of 120 volt loads, and if many of those loads are on one of the 120v leg and not the other, it can have additional voltage.

 

[ishiboo]

Not true. V = IR. The current in the neutral wire (I) is greater than zero, and the resistance of the neutral wire (R) is greater than zero, therefore voltage (V) must be greater than zero.

In addition, neutral is the middle of what could be an unbalanced load, so it can contain some voltage potential due to the unbalanced load. Since it is the return leg of 120 volt loads, and if many of those loads are on one of the 120v leg and not the other, it can have additional voltage.

Huh? It wouldn't be 120v if the neutral was not 0v. Voltage is the potential difference between two points. 120v - 0v = 120v. (And, irrelevant but, the resistance of the neutral wire is nearly 0, or it would be a poor conductor and a better space heater )

EGC and neutral are directly connected and you usually use the EGC to measure the voltage on the neutral, so of course the potential between the two (being connected) is 0v.

Say you have 120v, and the neutral is at 20v as you propose. The voltage you would measure would then be 100v. This comes in handy when you have multiple voltages and need something "in the middle" - for example, a computer fan. Connect one half of the fan to +12v and the black negative wire to +5v, and the fan sees +7v and operates happily.

("Neutral does not float" as I put it is in the discussion of a discrete system/circuit)

 

[Tractor Seabee]

It's amazing how a resurrected thread can take on a new life. All this theory stuff gives me a headache like it did when I first was exposed to it. The majority of our readers could care less as long as they do get a tingle when they touch metal. I at least can measure it with my volt meter.

Ron

 

[DarkBlack]

Dunno, code and science are closely related but different. I'm guessing the assumption is a larger service would service a larger potential area. I'm not an electrician, one could probably explain better the reason for the code. Check Mike Holt's forums.



Single phase 240V in the United States is brought in using two 120V hots (180 degrees apart) and one neutral (0V). The difference between those two hots is 240V RMS. Ground has nothing to do with whether 240v is available unless you're on a VERY archaic single wire service, used only in incredibly rural areas.



Neutral does not float, it's 0v potential. If 120v electrifies something metal, the purpose of the EGC is it completes a circuit to neutral, creating a short. Once you have a short, the breaker servicing that circuit blows.

You're right, and your wrong. A house is fed 240 volts single phase, center tapped. No need to confuse yourself, or others here with the notion of "2 120 volt feeds 180 degrees apart". "
Neutral is derived from ground which comes from the center tap of the pole transformer. It's potential is midway between the 240V legs.