# 3-phase neutral wire inquiry

Hello all, first time posting here. Just to give a quick background, I'm NOT an electrical engineer; I'm a motorsport mechanic by trade. However, in the motorsport world electrical work typically doesn't go beyond "start replacing parts and see where it goes", so I try to learn as much as I can because I actually find it interesting. I have a question regarding three-phase stators that will probably be easy for you guys.

In a system using a permanent magnet alternator, the stator does not have a ground or a neutral. My understanding is that if there is a neutral lead off the wye connection of a wye-wound three-phase stator it should hold that phase-to-phase voltage=phase-to-neutral*the square root of 3, ie if the phase-to-phase is 50 ACV the phase-to-ground (yes I am aware I'm using neutral and ground interchangeably here, I'll get to that in a second) should be ~28.866 ACV. And that is what you get when you measure phase to chassis ground. But here's my question: the stator doesn't have a neutral off the wye OR a ground connection, it's just the three AC output leads! If you measure phase-to-ground resistance it's an open line, but if you measure phase-to-ground ACV you get essentially what you would get with a neutral line off the wye. Can someone explain why this is? I'd appreciate it.

Your totally correct about the phase and ground relationships. As to why you ~28 volts to car frame? Well ur right, the frame could be floating at any Fixed potential (v) and hence measuring between them would give you some potential difference and not ~28 volts. However, my estimation is that since the car is in close proximity to the earth and it's weather, any bit of charge the frame may have had gets leaked to ground(earth) and this effectively keeps the potential of the frame a near 0.

is that what ur getting at?

Hhhmmm...don't know about those alternators and don't know if I am going to be able to answer your question; but I would share a couple of things.

In a 1-phase system, one wire is supposed to be the one that has voltage and can push current; the other wire is just the neutral that offers an easy path for the current to go back and close the circuit.

In a 3-phase system, each wire has voltage and is able to push current; BUT, these voltages and currents are out-of-phase in such a way (120 degrees) that when you connect them together, the currents add up to zero...so, there is no need for a neutral, in fact is you had one, there wouldn't be any current flowing through it in a good, balanced working circuit.

So, maybe, just maybe, the "neutral" of the stator winding is connected to the alternator housing with a very high resistance or some other device designed to work as a fuse or something, should the alternator ever fail.

Hope this helps.

I'm not familiar with automotive alternators, but I've got two thoughts.

1) Maybe it's not a wye connection. It could also be a delta connection, which doesn't have a neutral.

2) Regardless of whether it's wye or delta, the output is probably rectified with 6 diodes. Which means that for half of each cycle, each lead is one diode drop (0.5 to 0.7V) away from the negative terminal, which in probably always connected to ground.

Thanks all for the input. A couple points I forgot to mention in the first post:

1. I've measured the phase-to-ground current as zero, though the meter I was using at the time was an inductive ammeter that only goes down to .1A.
2. If you simple float the other end of the ACVmeter it reads as zero (or rather random fluctuations around 1-2 volts), but as soon as you connect it to the frame you get essentially phase-to-neutral voltage.

My working theory thus far has been that although the generator winds themselves are infinite resistance to ground due to the insulation, the metal itself of the stator core is ferrous and therefore has magnetic properties, and that the current being induced in the windings by the magnetic field of the spinning flywheel is in turn inducing a field (albeit a negligibly small one) in the metal core, and since the metal core is bolted to the frame, when the meter is connected phase-to-ground this is what it's actually reading. Basically phase RMS to the neutral of the X-axis on an oscilloscope since it is a balanced three-phase wye. Is this reasonable or am I way out in left field?

1.- I don't know what you mean when you say you "measured the phase-to-ground current"...careful, now, if you place an amp-meter between phase and ground you will create a nice short circuit!
2.- Yes, that's fine.

You can forget about the currents being generated in the stator, they are pretty much enclosed and don't come out. The stator core is made out of laminations which are insulated from each other and, thus, the induced currents travel axially (as in the winding wires) but can only go back and forth the short length of the lamination...this entire laminated stator core should, in turn, be insulated from the housing.

The housing itself could have small current flowing in it, too, depending how much magnetic field from the magnet is escaping beyond the stator core...I am thinking that , if they designed the alternator right, this effect would be very minimal. You could try to measure the voltage from one end of the housing to the other during operation and when off and compare.

1.- I don't know what you mean when you say you "measured the phase-to-ground current"...careful, now, if you place an amp-meter between phase and ground you will create a nice short circuit!
2.- Yes, that's fine.

You can forget about the currents being generated in the stator, they are pretty much enclosed and don't come out. The stator core is made out of laminations which are insulated from each other and, thus, the induced currents travel axially (as in the winding wires) but can only go back and forth the short length of the lamination...this entire laminated stator core should, in turn, be insulated from the housing.

The housing itself could have small current flowing in it, too, depending how much magnetic field from the magnet is escaping beyond the stator core...I am thinking that , if they designed the alternator right, this effect would be very minimal. You could try to measure the voltage from one end of the housing to the other during operation and when off and compare.

1. Please disregard, very poorly worded on my part. I started off referring to the current from one AC output to the other and ended up on phase-to-ground voltage instead. I meant the stator is open line to ground so I know no current can flow from the wind to the ground. That's what happens when thoughts run together...
2. I appreciate you're pointing out that it's correct, but the question I'm trying to answer is if there is an open line phase-to-ground why does an AC voltmeter read phase-to-ground voltage as essentially being the same as phase-to-neutral voltage if a neutral were led off the wye? In one case the phase wind has a clear path to a neutral connection, in the other it has no path to anything other than the other two phase outputs, yet measuring ACV yields the same number either way. That's the part I don't understand.

..., in the other it has no path to anything other than the other two phase outputs, yet measuring ACV yields the same number either way. That's the part I don't understand.

Doesn't each phase have a rectifying diode connected to ground? Then there is a path.

Doesn't each phase have a rectifying diode connected to ground? Then there is a path.

Yes but it's a permanent-magnet system different from a car's alternator. The stator plugs into a rectifier/regulator and the AC test is done with the stator unplugged.

Here's a diagram of the type of system I'm talking about (they're all pretty much the same in the charging system).
http://www.factorypro.com/wire_diagrams/GSX-R_600_K1,K2_Wiring_Diagram.jpg
Stator and regulator/rectifier are in the lower right corner (everything else is just simple DC application). You can see there's a connector between the stator and RR. I'm referring to the AC output of the stator unloaded from the rest of the system.

jim hardy
Gold Member
Dearly Missed
If you measure phase-to-ground resistance it's an open line, but if you measure phase-to-ground ACV you get essentially what you would get with a neutral line off the wye. Can someone explain why this is? I'd appreciate it.

With what are you measuring that voltage?

Motorcycle alternators indeed may well be just three wires with external rectifier/regulator.
So when you unplug it you have an honest-to-goodness floating AC system with no intentional connection to frame.
So strictly speaking the voltage could be anything, for example what if lightning just struck it and charged it to a zillion volts?
But that's exaggerating just to make a point.

The wires have capacitance to the metal alternator poles they're wound around.
So when you connect one wire to earth through your voltmeter, the other two wires which connect to different windings in the armature are both at some nonzero voltage.
A little bit of current will pass from those other parts of the armature to earth (er, uh, make that) to frame , through the capacitance i mentioned earlier,
and then through your voltmeter to get back into the armature.
So that little bit of current will make your voltmeter read some voltage.
DMM's are high impedance. A DMM may read near full phase voltage in that situation because it requires so little current to make it indicate.. maybe only a few microamps for a good DMM .

A lower impedance meter like this cheapie i always use

will read lower because it is less sensitive to current. This one needs a few milliamps...

You can find out if this is what gives you the effect you noticed by making your phase to frame connection through say a small light bulb
it'll easily pass more current than the stray capacitance of the armature windings so won't light
and if you connect the bulb while your voltmeter is connected you'll see the reading drop to zero.

Grounding is one application where "El Cheapo Meters Rule ! "

.

Last edited:
Jim Hardy you hit the nail on the head. My reasoning had been along the same lines but I never thought to use a cheapo meter instead so I went to Wal-Mart and bought a $10 meter, repeated the same test and voila! 0 ACV phase-to-ground on the same connection I was getting voltage on with the$250 fluke.

Important lesson learned and I've tucked the $10 meter away in the toolbox. Much to learn, this one has! Thanks for all the input folks. I'm basically trying to teach myself the basics of electrical applications and forums like this one are a great resource. dlgoff and jim hardy jim hardy Science Advisor Gold Member Dearly Missed The$10 Walmart GE meter does 99% of my work.
I soldered 15 foot test leads into mine with insulated alligator clips, makes it a lot handier to use. One can always grab a nail when he needs a pointed prod. Soldering them keeps them from falling out, and when i drop it off a ladder i can pull it back up.

My expensive Fluke comes out only when i need precision.
Found some medical Lithium 9V batteries for it in a thrift store. Seems they've hit mainstream now, Walmart has 'em.

Thanks for the feedback !

old jim

dlgoff