Found a problem with my ac voltage

[jarrod9155]

  Today I was trying to do some data logging ac volts verses rpms to work on my mppt tables  and found that the the volts were diffrent from leg to leg not a little alot and only had this result when the two out side legs were connected that read 110 volts ac and any other way you measured two legs it read 85 volts . I dont remember ever seeing this before  but what would be the cause and the effect it might have on performance .    always something

[Flux]

Were you measuring open circuit volts or was it connected to something? On open circuit volts are directly related to rpm so you only need one point, not much point in logging a lot of data.

If connected to something then it all depends on the circuit it is connected to. Any loading of one phase more than another will alter the volts, the alternator will not be very stiff electrically.

Obviously you don't have any short circuits or it would drag to a stop so it will be something to do with the load or you have a connection problem that has always been there and you misses.

Make sure these readings are on open circuit and if the difference is still there you have a connection problem. If you can get to the star point then check each phase to the star, if all are equal then you have a reversed phase.

It will certainly affect performance if it is a connection issue.At worst it will de-rate the alternator as one phase will take most of the load. You may be able to programme the mppt to still get the loading but it will be at reduce maximum power rating and almost certainly it will have most of the vibration issues of single phase and most of the time it will be running as such.

Flux

[jarrod9155]

It was open circuit not connected to anthing just spinning in a light breeze the mill does vibrate under load also

[joestue]

At 3/4th's the voltage it sounds like one coil is shorted out.

[Flux]

Sounds as though you have a reversed phase. I can only imagine this has been there from the start. Did you do any form of hand cranking tests on the alternator alone. If you only did tests with the rectifier connected for cut in speed you would not have spotted it was single phase unless you shorted the rectifier, in which case it would turn in lumps and that is a dead giveaway.

Not much you can do without getting it down.

Flux

[jarrod9155]

 I tested it this morning  under load and got the same results loaded there is one leg that is 10 volts lower at 60 rpms and as the rpms increase the diffrents increases as well . This would explain why the tower grouls so dame loud the vibration has actually broke my wind meter mouted on the pole .

[jarrod9155]

I did a bench test with it before install to test volts verses rpms but not on all legs and not with a rectifier  . But your right this problem may ave been here from the start becauce it has always had a slight cogging as if maybe uneven load phase to phase . The weird thing is it still makes power not great power but something .

[jarrod9155]

Could it just be one phases has less copper in its windings . I was pretty carefull but it was my first stator ever .

[Flux]

It is not going to be a construction issue, slight differences in amount of copper will go unnoticed.

This will be a connection issue, either in the connection of the coils in a phase ( a reversed coil will knock 2 coils out), or one phase is 180deg reversed at the star point.

Yes it will make power and under light load will behave as a single phase machine with all the disadvantages.

If you load it down hard it will work better than a single phase winding as the low voltage phase will conduct when the IR drop of the better ones exceeds the battery voltage. For mppt you should never load it that hard and you will be effectively reduced to single phase.

A fundamental first test should be to short all 3 leads and turn it, then compare it with 2 leads shorted. The lumpy single phase resistance to turning shows immediately. That and a voltage test on one phase should be good enough but checking all 3 phase voltages is a wise move.

Use the rectifier to determine cut in speed, measuring dc volts, again short after the rectifier and if you get lumpy torque and the machine is ok directly shorted then you have a diode problem.

Flux

[ChrisOlson]

I would agree with flux, and my best guess is that you accidentally welded one coil in backwards.

I build all my stators with six leads out.  I was working on a turbine this morning in my garage so I just tested this for you after reading this.  I connected it wye quickly with a piece of solder as a jumper and reversed one phase deliberately:



There's only one combination of the three power leads that has normal wye-connected voltage (one phase x 1.732) with one phase reversed.  The other two combinations just have single phase voltage.

--
Chris

[jarrod9155]

Thanks Chris , 
         For testing that for me , so what you and flux are saying is that one coil could be backwards and that is causing that one phase to be diffrent  voltage . So the one way I see 100 volts is right and the other two legs that  show a rough 11 percent drop in voltage  is the cause of incorrect hook up of coil  witch causes the single phase to happen .  Well this morning I ordered some new wire for a new stator . I have had a problem with the inverter tracking the mppt in low to mid winds 8 to 12 mph winds stalling mostly this would explain alot  . On the top end this gene performed normal if not good .  Big thanks for the quick replys to help me figure this out

[Flux]

I have had another look at your figures and I agree with Chris, a reversed phase would bring you back to phase volts on the two connections.

Taking your 110v then phase would be 110 /1.7 or about 64v.

I have no idea how many coils you have in series per phase but if it is three then reversing a single coil in a phase will drop that arm to  about 22v.

You would then see a vector of 64 and 22 v between the other two.

If you have 4 coils per phase then one reversed coil will leave you with 2 coils in series for that arm.

This seems the most likely thing to me, it probably means you flipped a coil but made the connection correctly for the coil the right way up.

So easy to do. I normally check with a magnet and a moving coil meter. If each coil is correct the meter will flick the same way on approach . A reversed coil will flick the other way. Far better to check before you pot the thing.

In all probability you can repair the existing stator if you can remember where the interconnections are and get in with a Dremel tool or similar.

Flux

[jarrod9155]

I have 4 coils per phase and as for reapair I been wanting to redo it from the start the cast came out terible . But back to connection problem I know i connected the tail tol legs correctly but I could have fliped one upside down wasnt even aware this would make a change in phase as long as a coil end connects to coil begging and so on , would this be like winding one coil in a diffrent direction example only I know I didnt do that .

[Flux]

Yes flipping a coil over is exactly the same as swapping the start and finish.

It's probably not a bad idea to mark one face before you remove it from the winder then if all the marks are upwards you know nothing is flipped.

That sounds very feasible, you will have 4 coils in 2 phases and 2 coils in the opther.

The vector sum of the 4 coils will be the correct voltage at 1.73 times the phase volts. That is where you measured 110v.

The other two connections will be identical but you have a vector sum of a phase and a half. I expect some mathematician can do this quickly but it will certainly be somewhere between phase at 64v and the correct line at 110.

Flux

[Dave B]

Way to go Chris. It's always refreshing to have a builder such as yourself helping out by example. Comments after the fact pale in comparrison to hands on and photos. Thank you,  Dave B. 

I would agree with flux, and my best guess is that you accidentally welded one coil in backwards.

I build all my stators with six leads out.  I was working on a turbine this morning in my garage so I just tested this for you after reading this.  I connected it wye quickly with a piece of solder as a jumper and reversed one phase deliberately:

(Attachment Link)

There's only one combination of the three power leads that has normal wye-connected voltage (one phase x 1.732) with one phase reversed.  The other two combinations just have single phase voltage.

--
Chris

[scoraigwind]

We did that one time in France.  the last coil in the winding got flipped over during the stator casting process (while putting them in there).  We identified the coil using a current from a cordless drill battery and a magnet.   then we dug a hole in the casting and cut the wire connecting that coil so we could reverse its connections.  It as not that hard to do in the end and very interesting.

[jarrod9155]

Finding the problem buy chance  was luck in the end all the little problems I have had like stalling, vibration ,and shaking under load not tracking mppt well . The part that throw me off was that the mill would still make great power on the top end and like flux said under heavy load at high rpms the IR would drop and the coils would start cunducting or something like that ,any way thanks for all the help from the guys on the furom . I ordered 15 lb of some 17 awg wire instead of 19 I hope to get aleast 320 turns per coil with a little thicker wire and raise my cutin voltage .   

[nekit]

We did that one time in France.  the last coil in the winding got flipped over during the stator casting process (while putting them in there).  We identified the coil using a current from a cordless drill battery and a magnet.   then we dug a hole in the casting and cut the wire connecting that coil so we could reverse its connections.  It as not that hard to do in the end and very interesting.

Scoraig
Could you describe how to check the stator with a battery and a magnet.  I want to make sure I don't have any flipped coils.

Thanks

[electrondady1]

neket,
when a dc current is run through a coil it becomes an electro magnet with a north or south pole on its face
when a permanent magnet is places on it it will ether repel or attract the mag.
all the coils in a 3 phase alternator should be wound in the same direction so if the mag reverses position over one coil , that coil will have been laid out incorrectly.

[nekit]

Tested mine today and it's fine.  Thought it was, but wanted to make sure.

Very simple test. 
Hooked a 4amp 12Vdc Automotive battery charger to two of the three phase leads. 
Marked which coils were in which phase. 
Took a magnet marked N & S on each side.
Check each of the coils in each phase to be sure they pulling in the same direction.
Switch the leads to the other phase and checked it.

Hope this helps anyone else wanting to check their stator.

[jarrod9155]

A qiuck update i got the new stator  up and going now for the last two weeks and so far it is proven to help with vibration and inverter ability to matain steady output . With the 17 awg 275 turns 12 coils the resistance drop from 16ohms to 9.2 ohms between  legs and I am still mataining a high voltage like 100 volts Dc around 90 rpms at the rotor . Today was my new best record of 7 kw hours thats about 1,000 watts and hour s and peaks of 2400 watts before furling around 24 mph . The epoxy for casting was alot nicer to work with and alot stronger . I was able to fit 17 lbs of copper in there tight .