comparing a single layer, 10 pole vs 8 pole, 6 coil machine.

[joestue]

Is anyone interested in the results if i run this test?

The way to make this comparison completely fair, is to use 40 magnets, 1.875x.875x3/8 thick, ferrite of course, because i already have 50 of them.

5 magnets per pole for the 8 pole machine and 4 magnets per pole for the 10 pole machine. i was going to have the 40 magnets touching at the inner diameter, so we're looking at 11.4" inner diameter and 15" outside diameter.

The coils remain exactly the same, except that the polarity is changed and the phase order is different.

The 10 pole machine has a .966 winding factor and the 8 pole has a .866 factor. As I understand it, most everyone on this forum is using the 4 magnet 3 coil ratio which is a .866 winding factor, which ..i know... doesn't entirely transfer over to non air core machines.

but basically i should expect to see more voltage and power out of the 10 magnet vs 8 magnet machine, from the exact same amount of magnets.

If this is true then it would make sense to do further testing.

I have about 70 pounds of 26 gauge wire, and friends who like 3 d printing so it would be rather easy to change the id to od ratio of the coils, as well as change the thickness, but this can be done rather easily with single layer test coils.

I have a lathe that can swing 14 inches without taking the bed gap out, so i might just shave off the inner edges of the magnets so i can fit them on a 14" disk.

I know a lot of you seem to have neodymium magnets that are spaced rather far apart.

It will be interesting to compare the output from both the 8 and 10 pole count magnet plates.. but with 1 magnet removed from each pole.. which almost works perfectly: it would be 30 magnets for the 10 pole count machine, and 32 magnets for the 8 pole machine. i would say 30 vs 32 is within the limits of normal harry home shop accuracy.


edit: just realized i will need 80 magnets, not 40. for the two magnet disk, single stator, standard design. (32$ on amazon for 50 of those things, so its negligible compared to my time).

[Adriaan Kragten]

You don't specify the winding for the 8-pole and the 10-pole generator. I assume that you use a 3-phase winding for both pole numbers. I also assume that you chose for the ideal angle in between the left and the right leg of a coil. This angle is 360 / 8 = 45° for an 8-pole generator and 360 / 10 = 36° for a 10-pole generator.

For an 8-pole generator you have two options. One is a 1-layer winding with six coils (see report KD 341 figure 5). The other is a 2-layer winding with twelve coils. All 24 possible positions for coil legs are used for a 2-layer winding. The 2-layer winding will give the largest power at a certain rotational speed because it has the largest copper density. However, a 2-layer winding has crossing coil heads and this makes the winding more difficult to lay.

For a 10-pole generator there is only one option for a 3-phase winding. This is a 3-layer winding with one layer for each phase and with five coils in each layer. So you have totally fifteen coils. But the magnetic flux through each coil is smaller than for the 8-pole generator as you have four in stead of five magnets for each pole. So I think that the 10-pole generator with a 3-layer winding will generate about the same power at a certain rotational speed as the 8-pole generator with a 2-layer winding. But a 3-layer winding has many crossing coil heads and laying this winding is even more difficult than a 2-layer winding.

There is an alternative 1-layer winding possible for a 10-pole generator. Each phase has three coils and the three coils of one phase are mounted close to each other. So the total space for one coil is 360 / 9 = 40°. However, this angle is too small to make the angle in between the two legs of one coil 36°. If this angle is taken 30°, the angle in between the two legs of adjacent coils is 10°.  The coils must be wound alternately left hand and right hand. The power is reduced because there is a phase shift in between the voltage generated in the left and the right leg of a coil. The power is also reduced because there is a phase shift in between the voltages generated in the three coils of a coil bundle of one phase. I think that therefore this alternative 1-layer winding with nine coils will generate much less power than the 3-layer winding with fifteen coils.

[joestue]

The whole point of doing the test, is that the exact same coils can be used for both configurations of magnets. You just flip the polarity of 3 coils, and the phase order also changes from acb to abc.

i've already done this test, but it wasn't valid for my intentions here, because i didn't have a way to make the magnets identical. i would need to source 24 magnets with the same volume as 30 of them. i made a 24 pole, 18 coil, 36 slot generator from a 56 c frame induction motor. the magnets were 2" long, 1/8" thick and 1/4" wide. i then flipped the polarity of half of the windings and added 6 more magnets for 30 poles. coincidentally it produces 120v at 400hz at 1800 rpm.

[MagnetJuice]

Is anyone interested in the results if i run this test?

I am interested in seeing the results. That sounds like an interesting test.

A few pictures would be nice also.

Ed

[Adriaan Kragten]

The whole point of doing the test, is that the exact same coils can be used for both configurations of magnets. You just flip the polarity of 3 coils, and the phase order also changes from acb to abc.

This can't be right. The optimum angle in between the legs of an 8-pole generator is 45° and for a 10-pole generator is 36°. So if you use the same coils for each generator, you can't have the optimum angle for both. The number of coils needed, is also different (see my previous post). It would be good if you show a picture of the coil pattern for both generators.

[MattM]

If he's not solid potting the coils into epoxy then he can use the same coils for both.

[Adriaan Kragten]

If he's not solid potting the coils into epoxy then he can use the same coils for both.

My comment has nothing to do with how the coils are connected in the stator sheet. It has to do with the optimum coil geometry and with the number of coils per phase.

[MattM]

Without epoxy in the way he can reshape the angles of the legs.  You assume he has optimal coil counts, matching magnets, and ideal shaping is critical to his test.

[brandnewb]

I have noticed that sometimes a few field tests are worth a thousand papers

Nah that is bollocks of course but for people like me that just can't grasp yet the fine details of theory just a few experiments (yes with pictures please) always does good with me

So I am interested to see results and pictures.