Why is safety ground dangerous as reference for 115V control voltage?

5ubtle · Jul 17, 2018

George Carlson's notes on RPC design (posted here) twice states a warning (IT CAN KILL YOU!) about not using safety ground for a reference for 115V control voltage. In the section titled "Simple Rotary Converter", the diagram is:

The warning that I'm referring to is contained in note 6 below the diagram.

"6. In the drawing, the potential relay uses AC Neutral as a reference. DO NOT USE SAFETY GROUND INSTEAD OF AC NEUTRAL, IT CAN KILL YOU! Since rotary phase converters require a fixed installation, availability of AC Neutral should not be a problem. If it is, use the dual primary transformer trick shown in the section on static phase converters."

The section on static phase converters has this diagram:

The second instance of his warning is in note 4 below this diagram. His workaround is in note 5.

4. In the top drawing, the potential relay uses AC Neutral as a reference. DO NOT USE SAFETY GROUND INSTEAD OF AC NEUTRAL, IT CAN KILL YOU!

5. The bottom drawing uses a dual primary (120/240V) transformer to generate a reference. This is handy if the AC Neutral is not available. Any small transformer will work. EPO or Ace Electronics would probably have these. The voltage on the secondary is not important since it is not used.

I see that Fitch Williams also uses a transformer to obtain 115V control voltage. I assume it is for the same reason.

If safety ground is used for the control voltage reference, what situation could arise causing the operator to be killed?

JST · Jul 17, 2018

Because ground wire may be open, and then the chassis of the machine, that is expected to be grounded, will actually have 120V on it! One single failure (the ground wire) can cause a dangerous, potentially fatal, condition

When you use a neutral and ground, if the neutral is open, the thing does not WORK, but it will not shock you if you touch the chassis. It takes more than one failure to create an unsafe condition, and failure #1 is usually the thing not working, which is hard to ignore.

sfriedberg · Jul 18, 2018

Not to contradict JST in any way, but to provide a general principle: The safety ground should never have any current on it under normal circumstances. That means you can't use it for a reference, because connecting a potential to the safety ground will cause current to flow.

"No current on the safety ground" is one principle behind ground fault circuit interruptor (GFCI). Even a few milliamps of current on the safety ground should pop a GFCI device. (As should that level of current imbalance between the two/three load conductors.)

Note that you can connect one end of an isolated transformer secondary (a typical Xfrmr, but not an in-line boost/buck device) to your device chassis to establish the device-local reference for the secondary winding. In this situation, loss of your neutral/return supply conductor will not cause current to flow through the safety ground.

9100 · Jul 18, 2018

I posted this once before, but it has been some years. An out of work electrician was looking for something to do so I had him replace the two prong outlets in my office. I handed him a box of outlets and a roll of green wire. He said, "We don't run ground wires. the conduit will ground it. Sometimes we hook the ground prong to the AC neutral." The argument degenerated to "I own this place and am paying you and here we run ground wires." He went to work and I went to other things. Months later, I had occasion to open a box. No ground. Even though he would have made more money, being payed by the hour, he was so pig headed that he wouldn't do it. The conduit was old when I bought the building and had multiple boxes with, of course, joints at each, so it was a dangerous "ground." I did it right and resolved never to take his word again.

I wondered how dangerous it really was so I did some tests, at a lower voltage, naturally. In a shock from hand to hand so it goes through the heart, the couple of gallons of saline solution in between is a good conductor so most of the resistance is skin contact.

If you touch two dry finger tips, you will get a tingle but are in no danger. Sweaty fingers get a much bigger tingle.

Sweaty hands with more contact area are dangerous.

In a worst case scenario where you are holding a metal cased electric drill plugged into a receptacle with the ground pin connected to a faulty AC neutral, you have worked up a good sweat and grab a water pipe for support, your chances of survival are very poor. People often are contemptuous of 120 V but under the right conditions it can be fatal.

Electrical codes are written for good reason.

Bill

JST · Jul 18, 2018

I recall going to the NAMM show, which is the big music biz trade show (filled the Anaheim convention center, or McCormick in Chicago, with leftovers). One show, an exhibitor had a humongous PA amplifier for subwoofers (they typically need lots of power). It was a fancy "digital" (PWM-based) amplifier and was something over 2000 watts, but had no big power transformer as most do.

In talking to the folks I discovered that they had decided that they did not need the transformer because they were just rectifying the AC line, and using the neutral as their ground.... the wires to the speakers were the AC neutral and the audio hot wire.

I was "rather surprised", and did mention that I thought that was about as unsafe as it could be, never mind that it would not pass any UL or other safety inspection. The sales guy there did not understand what I was saying, and was convinced it was OK, because "neutral is ground". I said my piece and moved on.

Someone must have clued them in more forcefully later, since I never saw them again, and at that time we went to two Namm shows per year.... January, and June or July. They were not at the next one.

thermite · Jul 18, 2018

9100 said:
I posted this once before, but it has been some years. An out of work electrician was looking for something to do so I had him replace the two prong outlets in my office. I handed him a box of outlets and a roll of green wire. He said, "We don't run ground wires. the conduit will ground it. Sometimes we hook the ground prong to the AC neutral." The argument degenerated to "I own this place and am paying you and here we run ground wires." He went to work and I went to other things. Months later, I had occasion to open a box. No ground. Even though he would have made more money, being payed by the hour, he was so pig headed that he wouldn't do it. The conduit was old when I bought the building and had multiple boxes with, of course, joints at each, so it was a dangerous "ground." I did it right and resolved never to take his word again.

I wondered how dangerous it really was so I did some tests, at a lower voltage, naturally. In a shock from hand to hand so it goes through the heart, the couple of gallons of saline solution in between is a good conductor so most of the resistance is skin contact.

If you touch two dry finger tips, you will get a tingle but are in no danger. Sweaty fingers get a much bigger tingle.

Sweaty hands with more contact area are dangerous.

In a worst case scenario where you are holding a metal cased electric drill plugged into a receptacle with the ground pin connected to a faulty AC neutral, you have worked up a good sweat and grab a water pipe for support, your chances of survival are very poor. People often are contemptuous of 120 V but under the right conditions it can be fatal.

Electrical codes are written for good reason.

Bill

Factory installed or added LATER in some shop, I know not, but..my 1942 10EE worklight had but ONE wire to its box. The other lead went to the box & base casting only.

Part of why I simply gutted it, 100%. All wire, conduit, switches, MG, DC panel.

100%.

M.B. Naegle · Jul 18, 2018

I was working on a machine once that had 115v supplied as described above: one leg to 220v and the other to ground. Never realized what was up and It never shocked me cause I stood on a rubber mat, but one day I touched the machine beside it while my hand was on the machine in question and that gave a little jolt! :wrong:

My body essentially linked the 'live' ground to an actual ground.

I've seen it in other cases too. Always non-OEM "shade-tree" electrical hacks.

JST · Jul 18, 2018

M.B. Naegle said:
....Always non-OEM "shade-tree" electrical hacks.

Those guys do some things I can barely believe.... if I had not seen it 14 ga aluminum wire hooked to a 40A dryer circuit, a connection (to a ceiling fan) with the wires twisted together and wrapped with scotch tape.....that sort of thing.

jim rozen · Jul 18, 2018

Note that you can connect one end of an isolated transformer secondary (a typical Xfrmr, but not an in-line boost/buck device) to your device chassis to establish the device-local reference for the secondary winding.

No. That is to say, not CAN but rather MUST make this connection to obey code rules. The secondary of the transformer is
defined as a separate service entrance, and as such, MUST have one end of the secondary winding (or CT if a centertapped
transformer) bonded to the groundING (green wire) buss.

sfriedberg · Jul 18, 2018

jim rozen said:
Note that you can connect one end of an isolated transformer secondary (a typical Xfrmr, but not an in-line boost/buck device) to your device chassis to establish the device-local reference for the secondary winding.

No. That is to say, not CAN but rather MUST make this connection to obey code rules. The secondary of the transformer is
defined as a separate service entrance, and as such, MUST have one end of the secondary winding (or CT if a centertapped
transformer) bonded to the groundING (green wire) buss.

Jim, I was referring to things like control transformers, which do not count as separate services. It is certainly legal, and in some situations eminently sensible, to leave isolated secondary windings unbonded. In fact, there are situations where leaving the secondary unbonded explicitly for user safety is the entire reason for using the (isolation) transformer.

The NEC Articles covering separate services do not apply to all transformers in all situations. If the transformer secondary is supplying a branch circuit, your statement is entirely correct. But not all transformer secondaries supply branch circuits.

JST · Jul 18, 2018

sfriedberg said:
.....

The NEC Articles covering separate services do not apply to all transformers in all situations. If the transformer secondary is supplying a branch circuit, your statement is entirely correct. But not all transformer secondaries supply branch circuits.

And if one does not, and stays within the equipment, as it seems is referred to, then it becomes an issue for UL, and not the business of the NEC. The NEC specifies "equipment listed for the application", and UL, etc, do the listing, typically according to UL rules. In a few cases the NEC "reaches inside" equipment, but generally it defers to UL, etc.

medsar · Jul 18, 2018

sfriedberg said:
In fact, there are situations where leaving the secondary unbonded explicitly for user safety is the entire reason for using the (isolation) transformer.

My workplace uses a totally isolated system and it is great for our application.

This next statement is not in response to the great answers on this thread -- it is for people who are reading/searching threads. The code is appropriately nebulous when taking about transformers. They assume that the installer knows when to use an isolation transformer and when not to. They also assume that the installer knows the appropriate monitors/maintenance for isolated systems. Just because the code says that you *can* run isolated system and per 'up to code', it doesn't mean that it is the right thing to do for every application.

5ubtle · Jul 19, 2018

Thank you all for the great replies.

M.B. Naegle said:
I was working on a machine once that had 115v supplied as described above: one leg to 220v and the other to ground. Never realized what was up and It never shocked me cause I stood on a rubber mat, but one day I touched the machine beside it while my hand was on the machine in question and that gave a little jolt! My body essentially linked the 'live' ground to an actual ground.

So, considering this scenario (as described by M.B. Naegle), where machine "A" has 115v control voltage referenced to safety ground, and he received a shock when he bridged machine "A" to machine "B" (machine "B" having a proper ground). It seems that machine "A" had a proper ground as it is implied that the machine was working correctly (if the ground was open, the control voltage would not have worked). Why did he get a shock? Is it because the ground on machine "A" was a poor ground (a high resistance connection back to the breaker panel)?

JST · Jul 19, 2018

5ubtle said:
Thank you all for the great replies.

So, considering this scenario (as described by M.B. Naegle), where machine "A" has 115v control voltage referenced to safety ground, and he received a shock when he bridged machine "A" to machine "B" (machine "B" having a proper ground). It seems that machine "A" had a proper ground as it is implied that the machine was working correctly (if the ground was open, the control voltage would not have worked). Why did he get a shock? Is it because the ground on machine "A" was a poor ground (a high resistance connection back to the breaker panel)?

That is a good question. Presumably one of them was not actually at ground potential, even though it was thought to be. My bet is on the one with the bad wiring. Possible that the various parts of the machine were not connected together very well, and that the voltage on at least the part he touched was not at zero volts, but at something higher. depending on how sweaty your hands are, you can feel a tingle at a fairly low voltage, maybe 20 or so volts, maybe higher. The voltage on the 115 volt might have been low, but still able to work the controls or whatever was powered by it.

jim rozen · Jul 19, 2018

sfriedberg said:
Jim, I was referring to things like control transformers, which do not count as separate services.

Been a while since I read the code but I believe they do, and they do need to be bonded (and fused as well....)

9100 · Jul 19, 2018

Most of the time electricians talk about feedback and the like, it is because they didn't do their homework but sometimes a surprising amount of energy can be coupled between wires in a long conduit. One time an electrician asked why a conduit was smoking. It turned out that he had run two conduits with boxes on both ends and the power to a pair of induction heaters, L1s in one pipe and L2s in the other. I wasn't involved with them but apparently they drew serious current for single phase units. Of course the two conduits made a single turn transformer secondary. Switching wires to make one going and one return in each pipe cured it.

Bill

blcksmth · Jul 19, 2018

9100 said:
I posted this once before, but it has been some years. An out of work electrician was looking for something to do so I had him replace the two prong outlets in my office. I handed him a box of outlets and a roll of green wire. He said, "We don't run ground wires. the conduit will ground it. Sometimes we hook the ground prong to the AC neutral." The argument degenerated to "I own this place and am paying you and here we run ground wires." He went to work and I went to other things. Months later, I had occasion to open a box. No ground. Even though he would have made more money, being payed by the hour, he was so pig headed that he wouldn't do it. The conduit was old when I bought the building and had multiple boxes with, of course, joints at each, so it was a dangerous "ground." I did it right and resolved never to take his word again.

I wondered how dangerous it really was so I did some tests, at a lower voltage, naturally. In a shock from hand to hand so it goes through the heart, the couple of gallons of saline solution in between is a good conductor so most of the resistance is skin contact.

If you touch two dry finger tips, you will get a tingle but are in no danger. Sweaty fingers get a much bigger tingle.

Sweaty hands with more contact area are dangerous.

In a worst case scenario where you are holding a metal cased electric drill plugged into a receptacle with the ground pin connected to a faulty AC neutral, you have worked up a good sweat and grab a water pipe for support, your chances of survival are very poor. People often are contemptuous of 120 V but under the right conditions it can be fatal.

Electrical codes are written for good reason.

Bill

100 milliamps (1/10 amp) through the chest cavity is almost always fatal.

Bob
WB8NQW

JST · Jul 19, 2018

9100 said:
Most of the time electricians talk about feedback and the like, it is because they didn't do their homework but sometimes a surprising amount of energy can be coupled between wires in a long conduit. One time an electrician asked why a conduit was smoking. It turned out that he had run two conduits with boxes on both ends and the power to a pair of induction heaters, L1s in one pipe and L2s in the other. I wasn't involved with them but apparently they drew serious current for single phase units. Of course the two conduits made a single turn transformer secondary. Switching wires to make one going and one return in each pipe cured it.

Bill

That's covered i n the NEC, which even shows cutting slots between holes for wires in a box if the wires for a circuit cannot be run through one. I'd have to look up the exact section.

ifixcnc · Jul 19, 2018

blcksmth said:
100 milliamps (1/10 amp) through the chest cavity is almost always fatal.

Bob
WB8NQW

Its all about the time in the heartbeat cycle. If you have seen an EKG pattern, there is a long down slope (the T wave as I recall). It is the short instance that the heart is not "in cycle". Shock at that point and it loses sync, no defibrillator around? 4 minutes until the end.

medsar · Jul 19, 2018

ifixcnc said:
Its all about the time in the heartbeat cycle. If you have seen an EKG pattern, there is a long down slope (the T wave as I recall). It is the short instance that the heart is not "in cycle". Shock at that point and it loses sync, no defibrillator around? 4 minutes until the end.

That's somewhat true. The T wave is the repolarization of the heart muscle. It is still 'in cycle' at that point. You are correct that a small amount of energy delivered during the repolarization phase (T wave) can result in very bad things including death. However, with a larger amount of current delivered over a longer period of time, the heart can go into the same very bad things that can result in death.

Why is safety ground dangerous as reference for 115V control voltage?

Plastic

Diamond

Diamond

Diamond

Diamond

Diamond

Diamond

Diamond

Diamond

Diamond

Diamond

Cast Iron

Plastic

Diamond

Diamond

Diamond

Cast Iron

Diamond

Hot Rolled

Cast Iron

Similar threads