What is the best equipment and method

[Tonyrand]

I have a customer that is getting shocked when in the vicinity of a pool light , and pacemaker goes crazy. The pool lighting is 12V, on a GFCI breaker. All wiring and bonding looks good. Removed wet nitch light and found the #8 bonding conductor not connected to the shell. Customer replaced old metal shell with a plastic one and I connected the #8 to the metal ring on the shell.
Did not solve problem.
What is the best type of equipment to check for voltage in a pool (besides sticking your hand in)and the best method?

 

[SG-1]

The trouble goes away when the light is de-energized ?

When does the customer feel this ? Siting on the edge with their legs in the water or swimming ?

What is the pool made from, cement, fiberglass ?

AC or DC ?

If the GFI is supplying a small stepdown transformer then it cannot "see" any current leaking on the low voltage side of that transformer.

Try using a sheet of tin foil attached to some conductor so you can "troll" the pool looking for voltage, that should give you plenty of surface area. Then a DMM ( digital multi-meter) with a Lo-Z setting to take the readings or an analog type voltmeter with a needle.

Welcome to the Forum !

 

[dmagyar]

Were you joking about the pacemaker?

Were you joking about the pacemaker?

Like SG-1 said, if the low voltage end of the transformer feeding the light is leaking that could account for the shocking; I'd remove the light, then take it home and do some testing on it. Check with tester as suggested by earlier post, not by relying on poor guy with pacemaker to see if shocking is absent when light is gone. Replace light with new unit if the light is the culprit as it seems to be.

If the guy really has a pacemaker I would also be careful about slamming the gate on the way out of the yard.

Did you open the above ground junction box for the light to check on the continuation of the bonding there? http://www.westwaysupply.com/fw028731.html

 

[Tonyrand]

No pacemaker testing

No pacemaker testing

Like SG-1 said, if the low voltage end of the transformer feeding the light is leaking that could account for the shocking; I'd remove the light, then take it home and do some testing on it. Check with tester as suggested by earlier post, not by relying on poor guy with pacemaker to see if shocking is absent when light is gone. Replace light with new unit if the light is the culprit as it seems to be.

If the guy really has a pacemaker I would also be careful about slamming the gate on the way out of the yard.

Did you open the above ground junction box for the light to check on the continuation of the bonding there? http://www.westwaysupply.com/fw028731.html

Yes the problem went away when power was switched off.
Pool is concrete gunnite restored 5 years ago right on the ocean.
To clarify, customer with pacemaker was in pool when he found the problem, not using to test.
Installed plastic wet nitche with metal ring and new pool light. Connected #8 bond to nitche ring.
Filling up pool now.
Then will re-test.
NEC 680.5.B.b appears to say there is no requirement for the #8 bond to the nitche if it is a low voltage (12V) system. Does this mean that it (bond) from the pool JB does not need to be there at all just connected to the pool steel? Yes checked connections from both side of transformer to (brass) pool light junction box.
Low Z analog tester the best method? Didnt know if there was a better tester.
If problem still exists should I be looking at the transformer?

 

[gar]

110313-1545 EST

I disagree with with using a low impedance meter. You want a high impedance meter, in this case a 10 megohm input impedance.

Most tap water is moderately conductive. In a room temperature pot using my tap water I read 50 kohms with a Fluke 27 from the stainless pot to a Fluke probe in the middle. DC resistance tests on water are not very good because of polarization. So it is better to use AC, but DC gives a quick indication. You want the measuring instrument to have a much higher impedance than the impedance of what you want to measure.

To measure the electric field in water or any other material you want a high impedance meter so as to not modify the field. For insulators this very difficult because it would require ultra-high input impedance. It is a relative matter.

If I take my pot of water and put it in the sink (an unnecessary requirement to be in the sink) and measure the AC voltage from the pot to any point in the water the reading is 0 millivolts resolved to 0.1.

Next I take a 16 volt transformer and connect one secondary side of it to the stainless pot and the other secondary side to an electrode in the middle of the water. The electrode is supported so it does not touch the pot. One side of the meter is connected to the pot and the other is the investigation probe to observe the voltage at different points in the pot. Because of the geometry you will see a greater voltage gradient as you approach the 16 V electrode.

You can run similar experiments in a glass dish with different electrodes connected to see how the voltage gradient changes with electrode and vessel shape and positioning.

In the swimming pool a reference point needs to be picked. For convenience this might be a 12" screwdriver driven in the ground close to the pool. Maybe better to have a large stainless mixing bowl in the pool water possibly about 5 feet from the light. The greatest voltage gradient is probably near the light. All sorts of positions can be checked in the pool. Ideally I would expect 0 millivolts between any two points in the pool.

Possibilities for other real or apparent pool current could be --- normal earth currents, and high frequency broadcast signals. A Fluke AC response drops off around 20 to 50 kHz. So if it was broadcast stuff, then it would result from rectification from probes and water and the demodulated signal.

"Kitchen sink experiment" this is my term for a simple experiment using easily available items. Does not mean it has to be run on the kitchen sink. As an aside --- Henry Ford ran his first internal combustion engine experiments on his wife's (Clara) kitchen sink on Bagley Ave. in Detroit. In the backyard was a small shop, now in Greenfield Village, where the first Ford car was made.

On to my "kitchen sink experiment" --- An 8" by 12" glass tray with an 1" of tap water, and 12" apart in the center of each of the two short ends is a 2" wide copper plate. These two plates are attached to my 16 V transformer.

If you put the two voltmeter probes on opposite sides of the long sides of the tray and hold one fixed, and move the other along the side you will find a null point. It is going to be very close to 0 V. This is where you are on an equipotential field line. If you had ever had A. D. Moore's senior class you would know how to sketch the electric field map of this tray. You can move the first fixed probe to a new point and then you will find the other probe has a new position for the null point.

At the extreme corners on one side of the tray I measured about 10 V. Keep one corner fixed and move the other probe close to the center of its near electrode and the voltage goes to about 12 V. Move the two probes close to the injection electrodes and without touching I can get to about 15 V.

You can use this type of experiment to determine the map of equipotential lines.

.

 

[G._S._Ohm]

What is the best type of equipment to check for voltage in a pool (besides sticking your hand in)and the best method?

I'd use resistors across your meter to simulate a human body.

Almost no one reads as low as 300 ohms and most people are lower than 3000 ohms so this puts most people between these limits.

0.5 mA is usually not detectible, 1 mA is detectible, 10 mA is painful and the let-go threshold is around 20 mA.
From these currents you can figure out your corresponding resistor voltages.

 

[SG-1]

Yes the problem went away when power was switched off.
Pool is concrete gunnite restored 5 years ago right on the ocean.
Then will re-test.
NEC 680.5.B.b appears to say there is no requirement for the #8 bond to the nitche if it is a low voltage (12V) system. Does this mean that it (bond) from the pool JB does not need to be there at all just connected to the pool steel? Yes checked connections from both side of transformer to (brass) pool light junction box.
Low Z analog tester the best method? Didnt know if there was a better tester.
If problem still exists should I be looking at the transformer?

It is encouraging that de-energizing the circuit caused the voltage to go away. That rules out a primary neutral problem & should make this problem easier to deal with.

What were the inital test results that you will use for a comparison ? What model meter were they made with ?

The highest voltage reading will be nearest the nich if it is the source.

If the trouble persists then it is most likely between the nich & the transformer, including the transformer, provided this circuit does not supply other pool side equipment.

If no bonding is required the new luminairie will be listed as such. 680.23(B)(5) seems to require any luminary with non-current carrying metal parts to be bonded.

gar gives very good advise.

 

[hurk27]

I would question the transformer and its wiring?

The reason is any non-auto type transformer is isolating, and 680.23(2) requires not only this but also requires it to be of a shielded type, to add to this the secondary is not allowed to be grounded, now if a underwater light is fed from a 12 volt isolated supply then it can in no way reference Earth, or any other grounding, also as Gar's experiment was pointing out, the only voltage gradients would be between the two contacts connecting the lamp.

With that said, the only things that I can think that would cause this, is someone bonded the secondary to ground or the wrong transformer has been used, or this voltage is showing up on the #8 ground that shouldn't be necessary, some where the isolation has been compromised.

Simple test is to place a small short rod into near by Earth and using a high impedance meter such as a DVM, test between this rod and the #8 EBJ, turn the light on and off, and see if there is a difference, if there is a difference, then this pool has not been bonded correctly, most likely missing the EB grid in the gunite.

The problem with older pools, is there is no easy fix, I don't see a reason for the #8 to be at this fixture, so it might be a fix to isolate it if it shows to be the problem.

Another problem I have found is the installer will bond the EBJ (#8) to the EGC of the circuit feeding the transformer, there should only be one connection between the EBJ and the EGC's and that will be because of the pool pump, This is one of the reasons why the NEC requires all EGCs to pools to be insulated.

Not saying this is the problem, but thinking the circuit out to me only points in this direction?