Author Topic: Automotive alternator to PM alternator  (Read 21313 times)

0 Members and 1 Guest are viewing this topic.

Smithson

  • Jr. Member
  • **
  • Posts: 75
Re: Automotive alternator to PM alternator
« Reply #27 on: July 31, 2014, 01:09:03 PM »
Sounds good.  This man has an interesting way to make the rotor.  I don't know if you have seen this.

http://www.windstuffnow.com/main/gm_alt_mod.htm

Arch

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #28 on: August 02, 2014, 10:45:27 AM »
The more I dig into this project the more interesting it becomes. I found this calculator for determining the gauss units at the coil of various magnet sizes and shapes, and was wondering if it is a valid tool to be using for selecting the most effective magnets to use:

http://www.adamsmagnetic.com/centerline_gauss_calculator_of_disc_or_ring_magnet.php

I was comparing .25" x .25" x 1"L magnets with .5" x .125" x 1"L using .160" for the distance of the magnet face to the center of the coil.

 W        T       L             N45       N52       
.25" x .125" x 1"         1632     1753 Gauss
.25" x .25" x 1"           2083     2238     "
.50" x .125 x 1"          2140     2299      "
.50" x .25" x 1"           2880     3094      "

Zubby and others like the whole area inside the coil filled with magnet, but it appears that mass, and not shape is the determining factor for field strength at the coil. Is this true, or am I missing something else that should also be evaluated?

Jeff

joestue

  • Hero Member
  • *****
  • Posts: 1352
Re: Automotive alternator to PM alternator
« Reply #29 on: August 02, 2014, 11:55:16 AM »
Basically neodymium magnets are like imaginary copper coils carrying tremendous amps. like on the order of one million amp turns per meter.
However the magnet is the same as air magnetically, (it also conducts electricity). because it is as air, it takes 1 million amp turns per meter to generate 1 tesla.

An iron core with copper wire wrapped around it only requires approximately 1/10,000 or 1/3000th as many amp turns to generate the same flux.
this is why a thousand turns carrying 2 amps can saturate your car's alternator with the stock claw pole.
however, the amps generated in the stator coils, oppose the amps flowing in the rotor.
there is also an air gap in the alternator. the rotor coil has to pass enough current to magnetize that air as well, and this takes a lot of magnetic mmf, or amp turns.

In most motors, we concentrate the flux in the teeth as high as is practical, for two reasons:
1 so that the flux lines do not travel through the copper coil, but rather around it.
2 so as to allow for more copper to be installed in the machine.

car alternators have relatively thin "back iron"
which means the back iron will saturate before the teeth do.
i'm not sure if this is by design but i wouldn't be surprised if it is.

in any case because the magnet spans nearly 3 teeth, the flux density through the air gap is about half the flux density in the thinnest part of the iron.

iron has non linear permeability and its practically impossible to get cheap motor laminations past 1.5T.
so you don't need real strong magnets.


with regard to the volume of magnet required, the amp turns flowing out of the machine demagnetize the magnets. the question is, how much.
these sorts of calculations used to be done by hand.. now we have computers.

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Automotive alternator to PM alternator
« Reply #30 on: August 02, 2014, 12:12:33 PM »
Lifting magnets are a different animal, Ignore all that lot for generators.

There are several factors that affect the flux in an alternator. The factors that affect the magnet are its Br value and its area.

Br is the remnant flux after magnetising in a closed magnetic circuit. Typical values for neo are in the 1.2T region depending on the grade. ferrite magnets are much lower ( 0.4 T region).

The total flux is Br x area (m^2) for flux in Wb.

No real alternator can be in a closed circuit, there has to be a mechanical gap for the thing to rotate. This introduces an air gap and air has a reluctance about 1000 times that of iron so most of the work the magnet does is in overcoming the air gap.

The effect of this is that the larger the air gap the thicker the magnet needs to be for the same gap flux. If you have a 1" thick neo magnet in an iron circuit with a 0.1" air gap you will get a bit less than Br, perhaps 1T.  Make the air gap 1" and you will be down to near 500mT.

With iron cored machines like your car alternator and a small air gap you can manage with quite thin magnets. For a small air gap magnets 1/8" thick will get you quite good flux and 1/4" thick will bring you to the point where you saturate the core teeth so thicker will gain you virtually nothing.

If you look at the axial air gap type of alternator you have very wide gaps and even with magnets 1/2" thick you will only get about 1/2 Br.

For your car alternator the useful magnet area will be the core length x magnet width so for a given core stack you can increase flux by using wider magnets but for a 3 phase machine, at any instant the current is carried by 2 phases so there is little point in making the magnet wider than 2/3 of a pole arc. This gives a good compromise with gap between magnets half magnet width, this does tie up with Zubbly's criteria of magnet width being that which will fit inside the coil hole.

If you use narrower magnets it still works but flux falls in relation to magnet area, the waveform also becomes strange but when rectified to dc that doesn't matter.

I hope that gives you a basic understanding. Choose your magnets on area and Br ( for neo that equates to the grade number, so N50 is stronger than N30). Then pick a thickness so that your air gap is small in relation to magnet thickness. Skewing rectangular magnets in a circular core means an increased air gap, that is one reason why Zubbly chose lots of small circular magnets but the mutual repulsion of multiple magnets for one pole is a disadvantage when fixing and I would use one magnet per pole for a small car alternator core.

Flux

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #31 on: August 14, 2014, 03:37:19 PM »
The preliminary coil testing is done. I used a round of some mystery steel to make the rotor, and slotted it at a 10 degree skew as Zubbly had suggested for a 36 slot rotor. I got volts! I used 12 of the 1" x .5" x .25" N52 magnets for the rotor. This is the max I can fit in the rotor. The results were not what I had hoped for, but about what I was expecting from what others have done:

RPM         VOLTS
300            4.3
400            5.8
500            7.4
600            9.0
700           10.5

These results are with the original 12 coils per phase OEM winding with 6 turns per coil. I am figuring another 3x turns (or more) should be adequate for my 12v needs at 400 RPM if I can manage to fit that many turns in. The only real disappointment was the cogging. It is pretty noisy and starts hard by hand turning the .590" dia shaft. I was going to use a 11.5 degree skew, but started with 10 degrees. I think I will have to increase the angle which means another rotor. Unfortunately more skew means more air gap, and experimenting. I will work on the cogging first as that is quite important for my purposes.

Jeff

joestue

  • Hero Member
  • *****
  • Posts: 1352
Re: Automotive alternator to PM alternator
« Reply #32 on: August 14, 2014, 06:42:56 PM »
10.5 volts at 700 rpm is more than i would have imagined possible with a stock winding. and yes that's one reason why the cogging is so high.

hiker

  • Hero Member
  • *****
  • Posts: 1493
  • BIG DOG
Re: Automotive alternator to PM alternator
« Reply #33 on: August 14, 2014, 07:58:34 PM »
if you only need 15-20 watts--go with some smaller mags or even ceramic--use some finer wire -more turns..cogging will not be a issue !
15-20 watts would be easy to obtain..!    i dont even bother to angle the mags and still get good results with smaller mags.........
WILD in ALASKA

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Automotive alternator to PM alternator
« Reply #34 on: August 15, 2014, 06:55:43 AM »
you don't say whether this is ac or dc volts but typically cut in at 12v dc is over 1000 rpm so you are way up on flux.

May be worth trying cutting the slots deeper to increase the air gap, this will reduce the cog.

Despite all the clever maths I don't think you can calculate the skew, approximately right helps a fair bit, you can get it to near zero by experiment but it is VERY critical. Easier if you have a blank core with laminations not welded, you can skew them for minimal cog before winding. With a welded core experiment is far more difficult.

You will have to wind with thinner wire to get in the turns you need but get the cogging right, see what volts you get, then you are well on the way. You have made a good start.

Flux

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #35 on: August 20, 2014, 07:52:24 PM »
Preliminary de-cogging efforts where not encouraging. The original 10 degree skew was not a full stator pole. While I could turn it by hand, it was difficult. A day later I needed a pair of pliers to turn it (?). I tried skewing angles from 11 to 15 degrees. The best seemed to be 14 degrees, but only marginally better than 15 degrees, if that. My magnets are exactly as wide at two stator teeth and the slot in between them. A 14 degree skew covers three stator teeth and two slots. 14 degrees seems to be a full stator pole skew which coincides with Zubbly's suggested skew. I tried increasing the air gap by .020" and noticed about an 8% drop in no load voltage, but no noticeable improvement in cogging. 

In light of these initial results, I am now thinking magnet skewing may not allow adequate cogging improvement for my my application. Possibly the thicker .25" thick magnets may not have been the best choice. I may try .125" thick magnets. I have read that crowning the stator teeth can help with cogging problems. It's a lot of work. Has anyone tried it?

Jeff

hiker

  • Hero Member
  • *****
  • Posts: 1493
  • BIG DOG
Re: Automotive alternator to PM alternator
« Reply #36 on: August 20, 2014, 10:03:37 PM »
ide go with the smaller mags....with those thick mags your losing output with all that skew..makes sence just to use less powerful mags and less skew=more output..  read somewhere on a old post--to large of a mag on this type of conversion it screws up the magnetic flux.
which again =less output..     just another thought--if your going to use it on a engine--use the stock field coil and rotor--and rewire the coils for higher voltage output--you can use some sort of varibale restor on the field coil to control the power output..all you need to do is hit the field coil with some voltage[small 9v. transitor batt. works] to get  things started-after that it goes on its own as long as you dont exceed the amperage output on the alt...worked on my pegen..............
WILD in ALASKA

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Automotive alternator to PM alternator
« Reply #37 on: August 21, 2014, 02:35:59 AM »
The core stack is very short on a car alternator, skewing works better with a much longer core so that you get a slot pitch skew with a lot smaller angle. I think you have confirmed a lot of my findings, some skew helps a lot but the skew for very low cog is extremely critical. I had a core pack relatively much longer than yours with the lams not welded, it was in a close fitting tube and I could skew it by putting a steel bar down one slot and twisting it. The thing was very critical and by the time I clamped it and welded it it was not as good as i had originally got it.

I suspect that the job would be far less critical with arc shaped magnets that present a cylindrical surface to the core. SWWP used arc magnets but instead of skewing they staggered the spacing, again I think it will be experimental, the basic maths won't get you close enough, there is too much flux fringing to be able to calculate things exactly.

I don't think crowning the teeth is practical in this case, it does help a bit with salient pole machines such as the F & P but the improvement is not great.

See if you can run it and see what you can get out of it, if you can get way more power than you need then you can widen the air gap and wind with more turns. For now just load it with the voltage that suits it and check the Watts, you can change the voltage later if it has plenty of power in reserve.

Weaker magnets will certainly reduce the cog but will reduce the output just as you will do with the present magnets and a large air gap, with only 8% loss of volts the gap flux is still high.

If you have the original claw rotor you will see that the pole tips are trapezoidal, this is a form of skew and the tips are thick near the disc and thin at the edge, this also makes the effective flux distribution tapered. This is not entirely done for cogging but I have noticed that the cogging on a fully excited car alternator is better than I would have expected.

Flux

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #38 on: August 21, 2014, 12:12:21 PM »
Ok, here is a quick attempt at a load test. Used a pair of 25 watt 12v tail light bulbs across the output of two legs only of the alternator:

RPM     No load VAC  Loaded VAC  Amps   Watts
200           2.5               1.9           1.45      2.8
300           3.9               3.2           1.80      5.7
400           5.3               4.5           2.12      9.5
500           6.8               5.8           2.40     13.9
600           8.2               7.1           2.66     18.9
700           9.6               8.4           2.90     24.3

What are your thoughts about getting reasonable output with a 3x-4x turn rewind provided an increased air gap will give a more reasonable cogging? Please keep in mind that my preferred output is 12vdc at 400-500 RPM. Thanks.

Jeff

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Automotive alternator to PM alternator
« Reply #39 on: August 21, 2014, 03:16:34 PM »
At the moment you can get about 12v dc at 800 rpm and the 3 phase rating ought to be a fair bit above the single phase, perhaps 40W with reasonable regulation.

As it stands you would need double the turns for 400 rpm I think you would get your 15W easily. I suspect that 3 times the present turns will be about as far as you can go, although if you can run at 500 rpm you have a better chance. This is a tiny alternator for much output at 400 rpm.

Regulating the output will let you push things a bit harder as the inherent regulation will be poor.

So much depends on how much you have to increase the gap to get low enough cogging.

Might be interesting to try the differential spacing instead of skew, you could probably play with the spacing easier than changing the skew angle.

Flux

gizmo

  • Sr. Member
  • ****
  • Posts: 379
  • Country: au
    • The Back Shed
Re: Automotive alternator to PM alternator
« Reply #40 on: August 21, 2014, 08:13:46 PM »
Didn't some of the Air-X windmills use a car alternator stator, with a clever magnet arrangement to reduce the cogging down to near nothing? Something along the line of uneven magnet spacing.

If there were, for example, 12 magnets, that if spaced evenly around the core would have a 1mm gap between each magnet. But instead, if the magnets are spaced  at 0.8mm apart, we end up with a gap in one place of about 2.5mm. I think there is a technical name for this type of arrangement but I cant think of it. I know there is a tiny decrease in power, but a massive reduction in cogging.

Glenn

 

smidy

  • Jr. Member
  • **
  • Posts: 77
  • Country: fi
Re: Automotive alternator to PM alternator
« Reply #41 on: August 21, 2014, 11:59:55 PM »
I did some diging in air 403 back in the days, they had car alternator similar stator, but it was 3 diffrent wire in the stator and some "clever" computer that chose how and when to use the diffrent windings, the magnets where sligtly skewed and even spaced around the rotor. yes it was rubbish...
Aland islands in the Baltic Sea

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Automotive alternator to PM alternator
« Reply #42 on: August 22, 2014, 07:13:11 AM »
I think the early ones used skew but like Glenn, I think later ones used the spacing method. The magnets were curved to suit the stator so things would be more predictable than rectangular magnets.

Basically skew aims to make the cogging torque zero over one tooth, the spacing method balances the clockwise and anticlockwise pull over the 36 teeth. With tooth saturation , flux fringing and non uniform air gap I doubt that calculation will be as good as guesswork.

Flux

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #43 on: August 23, 2014, 09:42:50 PM »
Timely discussion. I have a new rotor put together, and I am ready to machine the magnet pockets. This time I was intending to try the unbalanced offset method of de-cogging mentioned here:

http://www.otherpower.com/images/scimages/3538/decogging_tutorial_V1.pdf

With my 36 tooth stator, and 12 pole rotor, my offset angle by this method is only 0.83 degrees. This offset shouldn't create any noticeable unbalance in the 400-500 RPM range for my plans. If it does, it can be easily managed. Has anyone actually tried the offset method as presented in this article?

Jeff


Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Automotive alternator to PM alternator
« Reply #44 on: August 24, 2014, 01:47:33 PM »
Not tried it, but it was the method I called differential spacing and Glenn mentioned it as used by SWWP in some of their units.

I don't think the out of balance will bother you in the slightest at low speed. The idea of all the cog reducing techniques is to make the total reluctance  pull zero. Skew does this for each stator tooth lining up with a magnet edge. The spacing method does it over all the magnets so the clockwise and anticlock moments cancel over the total stator teeth.

Things like rounding pole tips or slot teeth only reduce the flux density on the tooth edges, it moderates the reluctance pull but doesn't get rid of it. I think the spacing method may work for you but I don't think the maths will be much more than a guide as in the skew case.

If you can stick your magnets on a round core with something strong enough to find the best position it would be better than trying to machine flats or slots until you have got it right. Might be a use for superglue ( not found one yet).

Flux

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #45 on: August 26, 2014, 11:49:17 AM »
Flux,
I tried the offset method as described in the link. Cogging is still present, but considerably reduced. It is low enough that I would consider it acceptable for my use. Voltage is > 0.1 vac of my first test voltage. I would still like to make some minor adjustments to the magnet locations to see what further improvements might be possible. The magnets are currently installed using only their attraction to the rotor. Risky, I know, but so far no problems up to 400 RPM with a close air gap. I would like to test at higher speed, and to move them around slightly to see what improvements can be attained with this de-cogging method. That would require super gluing the magnets temporarily to the rotor. When using super glue for machining purposes, I can knock a work piece loose using a sharp tap with a drift. Not sure if the super magnets would come loose similarly without breaking. Any experience removing temporarily secured super magnets?

Jeff

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #46 on: August 26, 2014, 02:06:27 PM »
After a little tinkering a little more, I think I have the cogging about as good and I can expect. It is pretty smooth, although I haven't been able to quiet it down much but that could be resonance from the lathe it is mounted on. I wish it had more poles. My 18 pole alternator was quieter but cogged a little more using skew with the same 1" long stator core. The offset (differential spacing) has was much easier in my opinion. The de-cogging and voltage output results turned out well using the math stated.

I guess now it is time to face the inevitable rewinding task and hunt up some winding materials. Hum...slot fill, circular mills....more to learn. Thanks for the help thus far.

Jeff

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #47 on: February 02, 2015, 09:52:36 PM »
No, I havenít abandoned this project. A few other important projects took priority, but I am back on it now. I wanted to review a few things just to be sure I am doing this correctly to get the most out of this automotive alternator stator. The stator is originally wired 12 pole using 12 overlapping coils in each phase winding. Each coil has 6 turns of #17 (.045Ē, 1.15 mm dia) wire making a total of 12 turns overlapping in each slot. My no load voltage phase to phase is 7.0 VAC at 500 RPM, and after connecting the 3 phase bridge rectifier the no load voltage is 9.4 VDC. I wasn't expecting that much difference, but that is what the VOM reads. I am assuming this information is as valid as a single test coil.

So, I presently have 12 coils x 6 turns each = 72 wire turns in each phase winding. If I rewind using the 6 coil virtual pole configuration (still donít understand virtual poles) and 12 magnets I will need 12 turns in each coil to duplicate the original winding output. All coils wound in the same direction. If I rewind with smaller wire, say #25 (.017Ē, .454 mm dia), with approximately 30 turns (or what fits), I should be able to get a no load voltage of about 20-23 VDC. I need a minimum of 15 VDC at load to adequately operate the switch mode external voltage regulator set to 13.2 VDC. My preferred load will be about 25-30 watts at 13.2 VDC. I will still use the offset method of locating the magnets to reduce the cogging level which seems acceptable at this point.

I would appreciate any comments regarding my understanding and my rewind plan. If correct, the only issue would be whether the stator slots would hold the 30 turns of #25 or larger wire. So, any suggestions on wire size and how to calculate wire size/slot fill would be appreciated. Thanks for all the previous information and suggestions.

Jeff

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #48 on: February 06, 2015, 09:29:46 AM »
If I am figuring this correctly, using the cross sectional area per turn of the different wire sizes for my calculation, I should be able to get at least 38 turns of #22 wire in the same slot area as the original 12 pole overlapping winding with 6 turns per coil of #17 wire. And possibly 48 turns of #23 wire. Based on 500 RPM  9.4 vdc no load voltage with the original 6 turn 12 pole overlapping winding, I should be able to easily reach my 20+ vdc at 500 RPM with a new 38 turn 6 pole winding using #22 wire. Hope that makes sense!

I would appreciate a confirmation and/or comments. Thanks.
Jeff

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Automotive alternator to PM alternator
« Reply #49 on: February 09, 2015, 02:45:57 AM »
It may be a good idea to load the existing winding with lamps or resistors to get 30W. This will give you a better idea how much the volts will drop below your nominal 9v open circuit. Keep the speed constant for the test.

For instance you got your 30W at 6v dc you would then you would need 3 times the turns to get 18v for your power supply.

For simplicity I would keep the original winding configuration, a full coil winding is not common on car alternators but as it is wound that way I would stick with it. Ideally changing to a half coil ( virtual pole) arrangement will get you the same result but if you change the slot distribution you will introduce new variables. Also if you keep the same arrangement you can safely assume the number of turns you can get to fit will be directly proportional to the wire area . It gets a bit easier with thinner wire and with a bit of care you may be able to get more wire in than the original but try to keep life easy for a first winding attempt.

Taking my example of needing 3 times the number of turns from above, you would use 18 turns per coil with 1/3 csa. That I think works out to 0.71mm or the nearest AWG equivalent.

Yes you are on the right lines but to avoid new variables I would keep to 12 coils per phase and reverse every other coil as in the present winding rather than using the half coil configuration.

Good luck you seem to be progressing well.

Flux

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #50 on: February 09, 2015, 08:43:12 AM »
Thanks Flux. I will stay with the original 12 overlapping coil per phase with the new winding. One further question....I am not familiar with your reference to "1/3 csa". Can you explain?

Jeff

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Automotive alternator to PM alternator
« Reply #51 on: February 09, 2015, 09:59:02 AM »
Sorry,csa means cross sectional area of the wire.

If you need to get 3 times the number of turns into a given space the wire needs to have 1/3 of the cross sectional area, not 1/3 of the wire diameter.


Flux

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #52 on: February 10, 2015, 06:48:33 PM »
Flux,
I tested the original winding, but before I did I increased the air gap by .015". I wasn't happy with the noise. It is now much quieter and smoother. I used Dinges off set method of decogging. The no-load open voltage with the increased air gap has dropped from the original 9.4vdc to 8.75vdc at 500 RPM. I then loaded the existing winding with 4 auto tail light lamps and measured 4.66A at 6.40vdc. That's 29.8 watts, and just where I want to be. Three times the original winding turns should give me what I need. Time to rewind.

Thanks.
Jeff

rustkolector

  • Jr. Member
  • **
  • Posts: 67
Re: Automotive alternator to PM alternator
« Reply #53 on: April 21, 2015, 09:01:44 PM »
Since I started this thread I thought I would report back with my results. This project was to build a slow speed (350-500 RPM) DC generator for a 1/8th scale working model of a 1906 twin cylinder gas engine. Design was for a 30 watt output at 500 RPM at 12vdc using a 93mm dia. 30 amp automotive style alternator stator. The original stator was 6 turns of #17 wire wound with 12 overlapping coils. I found someone that had experience rewinding small motors. Duplicating the 12 overlapping coils with 18 turns of #22 wire proved to difficult to manage the end turns with the shallow stator core. It was decided to wind it with 30 turns of #23 wire in a wave or serpentine winding which proved manageable. The rotor is 12 pole with a steel core using 1"x .5"x .25" neo magnets arranged using the offset method. Cogging is almost non existant. Upon testing, the no load DC voltage output was 25vdc. This was stepped down through a 3A switch mode DC to DC buck converter to 12vdc. The results were as follows"

RPM