Fieldlines.com || Multiple phases

Front Page - [Homebrewed Electricity-- (wind) (solar) (hydro) (steam) (controls) (storage) (mechanical)] - Classifieds - Site News
Everything - Newbies - [Remote Living-- (housing) (heat) (light) (water)] - Reviews - Diaries - Our Products

Multiple phases
By scoraigwind, Section Homebrewed Electricity Posted on Wed Jun 18, 2003 at 04:05:39 PM MST
rectifier connections for multiphase stators

I am posting this as a new story so I can include this graphic. It's really a comment on an earlier thread. The one about three-phase.
You can have as many coils in the stator as you like. Just make them the right shape to match the magnet pitch and then pile them in.
Connect the tail like this.

Multiple phases | 18 comments (18 topical)

Re: Multiple phases (none / 0) (#1)
by Simon on Wed Jun 18, 2003 at 11:57:37 PM MST

Hi Hugh, Is there any added advantage to multiple phase machines in form of output etc? I see 2,3,5 phase machines? I have done 2 phase at most to switch series and parrallel depending on the voltage output. simon

Re: Multiple phases (none / 0) (#2)
by scoraigwind on Thu Jun 19, 2003 at 01:32:10 AM MST

Advantages

The output will be smoother. Single phase output makes poor use of the wires. The current pulses up and down and heats the wires much more than a steady current. The rms current from a single phase recitifer is about 40% higher than the average and losses can be as much as double the loss of a steady DC current. You also get vibration in the machine.
You can get more copper into the stator. Even without overlapping the coils you can stuff a lot more copper into the stator by making the coils the right shape and then cramming them in side by side. I find that 3 coils for every four magnets gives a nice three-phase output. This works best if the magnets are speced a little wider apart than Dan's ones. Then you can have a good sized hole in the centre of each coil.

If you do not have a multiple of four poles, then you have to use a different ratio. For example 14 magnets could have 10 or 11 coils. In the case of 10 coils you will only have 5 phases, but I'd still connect each coil to a different diode because it is nt easy to space the coils out perfectly evenly.
Hugh Piggott http://www.scoraigwind.co.uk
[ Parent ]

Re: Multiple phases (none / 0) (#3)
by Bach On on Thu Jun 19, 2003 at 08:50:45 AM MST

OK. I admit it. I'm confused (a perpetual state of mind, I'm afraid.) I've been to Ed's site and read his lucid explanation on 3 phase. But I still have a multitude of questions. Regular readers will know that this is generally the case.

I understand that there are numerous advantages to multiple phases over single phase generating. Hugh listed only a few of them here. The typical plan is to use three phase. It seems to be the standard in commercial power production.

I've just poured the fiberglass around my 12 coils. Each coil has 108 turns of 18 gauge wire. The alternator is going to have two rotors with 24 small magnets on each rotor. I have not done the connections yet since this is my first attempt at a fiberglass cast stator. My guess is that the stator will be just over a half-inch thick. It's about 12.5 inches in diameter. Sorry - No digital camera, so no pictures.

Some questions (e pluribus unum):

If I use diodes/rectifiers as Hugh has shown at the coils, isn't the current immediately turned into DC? DC power isn't in phases. Or do the diodes/rectifiers serve some other purpose here? My goal is to attempt 3 phase and send AC power via the connecting cables for conversion to DC closer to my battery bank (12 volt)? Perhaps this question reveals the depth of my confusion.
I want/need to test the output before erecting a full blown windmill (ahem). Can I still get 3 phase for testing without the diodes and/or rectifiers? In other words, can I try star and delta for testing without converting to DC?
With my 12 coils, I was planning 4 circuits with 3 coils in each circuit. Right? Then how do I connect the 4 circuits to combine for maximum output?

My aim is some usable power at low wind speeds. We have mostly light winds. I may not get a lot of power, but any will be good for charging batteries on my back-up lighting power supply system.

Maybe I should have started with single phase unit for my first alternator, but I'm convinced that 3 phase is better. I just have so many questions about it.

I'm not above doing "Monkey see; monkey do." But I really prefer to try to understand what I'm doing too - and "why?". It must be obvious to the informed that such is not the case at this point.

All advice will be very much appreciated.

Bach On
- - I'm not superman, but I am very dense! -

Re: Multiple phases (none / 0) (#4)
by Electric Ed on Thu Jun 19, 2003 at 10:50:27 AM MST

Answers to Questions-

If you want to transmit power over any distance, leave as 3 Phase AC, at the highest voltage you can generate, and transform/rectify at the load end.
To test star/delta arrangements, it must be done on the AC side of your rectifiers. Series/parallel tests can be done on either the AC or DC side.
For a 3 phase machine with 12 coils, use 4 coils per phase, so you have the option of series or parallel connections in each phase. (Sketch below) I will try to come up with the star and delta sketches later.

Isn't this fun,
Electric Ed

[ Parent ]

Re: Multiple phases (none / 0) (#5)
by Victor on Thu Jun 19, 2003 at 12:28:07 PM MST

Hmmm , 12 coils and 24 poles. You have built a single phase machine assuming the coils are evenly spaced.Polyphase needs to be phased.

[ Parent ]

Re: Multiple phases (none / 0) (#6)
by scoraigwind on Thu Jun 19, 2003 at 01:18:53 PM MST

1. the depth of confusion is quite normal. DC is the most efficient method to transmit DC power. There is no benefit to using 3-phase transmission to a rectifier rather than DC transmission. Benefits only arise when the 3-phase can feed three separate AC circuits or a transformer.

But! it does matter how many phases you have before the rectifier. The DC current from a single phase source is pulsating and it wastes much more power inthe wires than rectified 3-phase or other multiple phases.

Different phases deliver pulses at different times, giving you smoother DC and better quality/efficiency.

You will not get star or delta from your winding with 12 coils and 24 magnets. You have a single phase winding. I hope the coils are the right shape/size but I guess they will work, so long as you reverse every second coil. However they are all in phase with each other, in that ratio.
Connect them in series or parallel (after you have reversed every second coil).

Hugh Piggott http://www.scoraigwind.co.uk
[ Parent ]

Re: Multiple phases (none / 0) (#18)
by jubalearly on Mon Jun 23, 2003 at 04:24:45 PM MST

Q1. Yes, the power is converted to DC as soon as it goes thru the diodes. I like Hugh's arrangement, since using the bridges this way puts the diodes pairs in parallel. That doubles the current capacity and halves the voltage drop.
You wouldn't want to do it this way if you are going to remote locate the diodes because you would then be running many more wires (2 for each phase) a long way. The conventional way is 1 wire for each phase + 1 neutral, common, or ground depending on the connection - 4 wires for 3 phase. Then you need 3 diode pairs at the remote end, for the phase wires.

Russ

[ Parent ]

Back to school.... (none / 0) (#7)
by Bach On on Thu Jun 19, 2003 at 03:11:39 PM MST

Thanks for the clear answers. Hugh, I see now what you're saying about the phases. I can't have 3 phase with the 24 magnets 12 coil arrangement. All my coils would be in phase with one another. The coils are set in fiberglass and I ain't doing that over!

But, if I change the number of magnets - say to 16 or 20, I'd be able to make 3 phases with my 12 coils. Am I correct there?

Hugh, I'm finally seeing what you and others have been saying about using 3 coils for every four magnets. With that ratio, some coils will always be at different phases. I think I'd been focused only on the loss of power by not having as many magnets or coils. I was thinking in single phase. (I do understand that with single phase, I'd have to reverse the connections on every-other coil.

Ed, so I use 4 coils as what I'll call a "circuit". Depending on where I start, I pick every 3rd coil around the stator for each "circuit". Each circuit is a phase. Then I connect the 3 circuits up as either star or delta. Am I following that correctly? (All assuming I change the number of magnets I've got on my rotors.) Ed, you do such great drawings!!!

But I've got to say to Hugh - I've always heard that DC requires larger gauge wires for power transmission. Or is that an old wive's tale?

Thanks ever so much!

Bach On

P.S. If you guys think I have questions now, wait until I get started on blades.
- - I'm not superman, but I am very dense! -

Re: Back to school.... (none / 0) (#8)
by Electric Ed on Thu Jun 19, 2003 at 09:24:35 PM MST

[Quote]Ed, so I use 4 coils as what I'll call a "circuit". Depending on where I start, I pick every 3rd coil around the stator for each "circuit". Each circuit is a phase. Then I connect the 3 circuits up as either star or delta. Am I following that correctly?[/Quote]

Yup, You've got it.

If your coils are "side-by-side" (as opposed to lapped or "shingled"), use 16 magnets per rotor.

Ensure that the center-to-center magnet spacing is equal to the center-to-center "span" of the legs of a coil.

Electric Ed

[ Parent ]

single phase connections (none / 0) (#9)
by scoraigwind on Thu Jun 19, 2003 at 10:26:42 PM MST

If you have the same number of magnets as coils then you do have to reverse the wires on every second coil before you can connect them series or parallel.

However in the case of 24 magnets to 12 coils, you can connect them without reversing them because each coil will meet a north pole at the same moment.

If in doubt measure the voltages before finally connecting them. Try series voltage with and without reversing wires.
Hugh Piggott http://www.scoraigwind.co.uk
[ Parent ]

Connecting coils (none / 0) (#17)
by scoraigwind on Sun Jun 22, 2003 at 01:07:21 AM MST

You can find out how many phases you have by placing the stator onto the rotor in a diagram like the one Electric Ed just did.

Lable the magnets N,S,N,S etc. Look at the coils and how they lie on the magnets. Find a coil that lies directly centred on a N pole. That is coil zero. Count coils up from there one, two, three until you have another coil lying directly centred on a magnet pole. Note whether this is a N pole or a S pole.

The number you have counted to is the number of phases 'n'. It will be an odd number. If it is an even number then you have missed a pole half way.

If the second magnet pole is a N pole, then all the coils in one phase have the same polarity. You can connect them in series byy connecting the tails start-to-finish or in parallel start-to-start. Each coil in that phase will pass a N pole at the same time.

If the second pole is a S pole then you have to reverse every second coil's connections (start/finish tails). Then you can connect them as above. Reverse the connections of every second (even number) coil in the stator. This will also have the effect of reversing every second coil in each phase.

Going back to the AC/DC thing. DC is the most efficient way to transmit power at any given voltage. AC is only better because you can use a transformer to change the voltage. If there is no transformer, then the most efficient way to transmit power is with DC. Dc systems only have thicker wires because they are at lower voltage. That's the only reason. If you can use a transformer to change the voltage between the battery and the transmission line then AC will be a good choice for transmission. Otherwise DC is best.

I have to agree with Hannu that it is more practical to keep the diodes at the battery where you can replace them if they blow. But this does mean you have to AC for the long run transmission. In the case of single phase this makes no difference - the current will be pulsating in either case and the loss will be the same (high). In the case of 3 phase, you will need to have 3 wires, each as big as the size you would use for 2-wire transmission of DC. This will have the same loss as the DC wires but cost you more in copper.

If you have multiple phases it is a bit crazy to run all those wires and you really have to have the rectifier up there by the alternator.
Hugh Piggott http://www.scoraigwind.co.uk
[ Parent ]

Re: Multiple phases (none / 0) (#10)
by hvirtane on Fri Jun 20, 2003 at 03:15:03 AM MST

I think that it best to use three phases.
It is known that a three phase is reasonably smooth.
You can use star or delta wiring.

When you are charging batteries:
a) It is best to use high voltages for transmitting
the power any distance (little losses).
b) It is better to leave rectifiers in the end of
the the transmitting line, because you maybe
have to use a transformer as well...
if your voltage is really high...
And it is best to have as little as
possible technology on the top of the mast?
c) Please note: It is difficult to make good transformers
if your current is more than 60 Hz.

- Hannu

Re: Multiple phases (none / 0) (#11)
by wpowokal on Fri Jun 20, 2003 at 07:57:18 AM MST

Forgive me for throwing in a wild card but this discussion may be apropriate!

I have followed several discussionns on output wave form, and all seem to strive to achieve a sine wave. To me this would be desirable if the output was to be used as AC to a transformer etc. However, to simply feed a rectifier would not a theoretical square wave be more efficient?

Just askin!

regards allan
"Dream as if you'll live forever, live as if you'll die today." James Dean

Re: Multiple phases (none / 0) (#12)
by Chuck on Fri Jun 20, 2003 at 08:46:52 AM MST

An interesting point. The only issue being how do you make a square wave with an alternator ? Or generator for that matter....
chuck
[ Parent ]

Re: Multiple phases (none / 0) (#13)
by troy on Fri Jun 20, 2003 at 09:36:06 AM MST

Excellent discusion. Rotating machinery like alternators "want" to produce a sine wave. That's an inherent quality because of the relationship of the coils to the magnets. As the magnet approaches the coil, it starts to make a little voltage. The closer it gets the more voltage, until it peaks over the leg of the coil and then starts dropping again.

So yeah, virtually impossible to get a square wave out of a round (rotating) alternator. Now, how good the sine wave is, that's a whole other question that depends on matching coil pitch to magnet spacing, coil hole to magnet size, coil shape, coil thickness, etc, etc.

Best regards,

troy

[ Parent ]

Re: Multiple phases (none / 0) (#14)
by windstuffnow on Fri Jun 20, 2003 at 11:10:48 AM MST

The concern about Hz or sinewave seems a bit out of context when the unit is being used to charge batteries. It would be virtually impossible to match an electical appliance to the alternator requiring a specific Hz or perfect sinewave because the wind is always varying. I can see being concerned after the conversion through an inverter but why at the alternator? Efficiency is a concern but as long as all are lined up properly you should get a good wave. If you have "perfect" magnets and "perfect" coils and all the geometry is "perfect" then you'll get a "perfect" wave assuming all other interactions with the iron are also perfect. Since "perfect" is difficult to attain because of the shape of magnets, geometry of the designs, materials used, etc... doesn't it seem a bit overkill to strive for perfection with inferior components?

Just my thoughts...

Perfectionism is self limiting.

Have Fun
Ed

Have Fun! Windstuff Ed
[ Parent ]

Re: Multiple phases (none / 0) (#15)
by wpowokal on Sat Jun 21, 2003 at 06:10:57 AM MST

Me thinks Ed has hit the nail on the head, too much procrastination without getting down and dirty dose't make power.

However my thoughts were around maximum output for size, accepting resistive losses/heating.

We know that one of the big problems with cheep modified square wave inverters is a relative high DC component of the wave form. This affects inductive loads with extra heating, however resistive loads like heaters dont care.

OK so Troy's comments on the interaction between magnets and coils and thus the resultant wave form is reality. But accepting that magnets are a given, that leaves coil design as a variable..............

My point being the faster the voltage rise and the longer the voltage is high the better, having said that there may be some vibration introduced by this?. I probally should not have posted this until I had given the text books a hiding, but..... it has been on my mind and many minds make good mills.

regards Allan
"Dream as if you'll live forever, live as if you'll die today." James Dean
[ Parent ]

Re: Multiple phases (none / 0) (#16)
by laskey on Sun Jun 22, 2003 at 12:16:21 AM MST

3-phase AC is the thing you want to use. It's very smooth and has a tendency to be gentle on components. Put your hand against your face... push in slowly, and release slowly, and do it over and over again... that's a sine wave not so bad eh? Now for a Square wave... Smack your self in the face hold your hand there and then remove your hand hold it then smack your self... over and over again... Sucks eh? Now imagine your an electronic component living with that at varying speeds and intensity for years. Another thing to think about with square waves is RFI (radio frequency interference), those nice sharp rise times have a tendency to generate all kinds of RFI harmonics which will manifest itself all through you abode , but mostly as static noise on your radio equipment (including your TV).

3-phase AC is the thing you want to use. When You transmit through Star or Delta you only use each wire for 2/3rds of the time, so they don't heat up as much, and when you bridge rectify 3-phase AC you end up with a pretty low ripple DC which can be used to charge batteries (and run some loads)with out further modification.

Cya,
Chris

[ Parent ]

Multiple phases | 18 comments (18 topical)

Menu

� create account
� How to use the board
� FAQs
� search the board
� Google search the board

Login

Make a new account

Username:

Password:

Total Views

��101 Scoop users have viewed this posting.

Related Links

� magnet
� Also by scoraigwind

You must be a registered user to post here. It's easy and free, and the link is on the upper right side of your page.
All trademarks and copyrights on this page are owned by their respective companies. Postings are owned by the poster, but may be deleted or moved at the ADMIN's sole discretion. The Rest � 2009 Forcefield.
You can Email the board ADMIN here. PLEASE include the username you signed up with!