Magnets 'r us

The other way of lowering the cut in speed is to increase the field strength and/or the number of magnetic poles. With shunt field type generators and some alternators, it is possible to increase the power to the shunt coils, resulting in an increase in the magnetic field and a corresponding increase in output from the armature coils. There is a limit to how much these coils can take without rewinding. So there may not be a practical solution here. PM motors/generators could use a change of magnet type, from ceramic to rare earth for an increase in field strength (flux). Sizing becomes an issue here, as most rare earth magnets are quite thin ("), while ceramic magnets are fairly thick ( " and up). It becomes necessary to build platforms for the new magnets to get them close enough to the armature to maintain a decent air gap between the magnet and the armature. It may be worth it but I'm not sure it will get a decent cut in speed if you leave the number of poles the same.

The other method is to increase the number of magnetic poles. A doubling of poles (of the same flux) should result in a corresponding doubling of voltage at the same speed or the same voltage at the revolutions per minute. The following table assumes the same armature used for all the options and the same amount of flux for each pole. It also assumes that there won't be a problem with using an armature wound for 4 poles (there is ! Keep reading).

Volts

Amps

Watts

Revs/minute

Poles

12

50

600

1750

4

12

50

600

825

8

12

50

600

412

16

12

50

600

206

32

12

50

600

354

20



I won't get into motors with shunt fields. With PM motors/generators, to add poles it is necessary to remove the old magnets and replace them with the correct number of thinner (width only) magnets. If the magnets are less thick, then extensions of ferrous metal (Iron or steel) need to be used to keep the air gap close. In my situation, I chose to replace the four large poles with 20 smaller magnets. Looking on the web I found some 6" x " x 1" ceramic magnets at wondermagnet.com. Since they are located nearby, I drove over and picked some up. Six inches was way too long, so I bought 10 of them and cut them in half with a radial arm saw and a ceramic blade. I set the blade to only cut into the magnets 1/16 th of an inch and cut very slowly backwards. Most of the magnets broke cleanly in half while I was still cutting them. The few that broke at odd angles didn't bother me since I only needed 2 " of length to match the 2 " of armature surface. I didn't think the difference would be a problem.

Lining up the magnets inside the motor housing, I measured and found a large air gap. The original magnets were " thick and were curved to fit the inner surface of the housing. These new magnets were " thick and were flat. I had hoped the lack of curvature would offset the difference in thickness, but I was wrong. There was only about a 1/32 " gain over the " thickness. This was remedied by cutting matching pieces of 3/16" x " x 3" weldable steel bar and epoxying them to one face of each magnet.




I used slow setting (30 minute) epoxy to glue the steel bars to the magnets, after roughing the surface with emery cloth. It was then time to place the magnets and check the air gap.




Paper was used to space the magnets evenly before gluing. As can be seen in the photo above, There is no way to fit another pair of magnets into this housing. From a physical standpoint, 20 was a good choice. The original 4 poles were placed so that the poles alternated every 90. The brushes and armature windings were set up accordingly. So in order to have a chance of this working without rewinding the armature, there had to be a similar arrangement with the new magnets. This requires a number of magnets divisible by 4. I originally was thinking of using 18 poles since there were 37 armature slots and commutator segments. This would have probably lead to a bit more power for a given rpm, but would also have a stronger "cogging" effect resulting in poor startup. For better or worse, I glued in 20 magnets, making sure to alternate the poles. From a physical standpoint, 20 was a good choice. The original 4 poles were placed so that the poles alternated every 90, with north being 180 from the other north. The brushes and armature windings were set up accordingly. So in order to have a chance of this working without rewinding the armature, there had to be a similar arrangement with the new magnets. This requires a number of magnets divisible by 4. I originally was thinking of using 18 poles, but the 90 angle wasn't there. As it turned out, many of my thoughts on this subject and about not rewinding the armature were wrong.

Now just to confuse the issue, books on generator design refer to magnetic degrees. 360 magnetic degrees is from north to north or south to south. 180 magnetic degrees is from north to the nearest south or south to it's nearest north. The coils in a generator should span 180 magnetic degrees. In other words, from the center of one pole to the center of it's nearest opposite pole. See this diagram for clarification. When there were four poles, each coil spanned 90 physical degrees in order to span 180 magnetic degrees. With 20 poles, each coil would span 18 physical degrees. In the new configuration, the original four pole windings actually end up spanning 900 magnetic degrees, or 5 changes of pole type. When I put the 20 pole casing together with the four pole armature, I was only able to get .5 volts by hand spinning the shaft and speed didn't change the reading. Trying to run it as a motor from a 12 volt battery resulted in no movement at all and a loud hum. This was not a good sign. It is apparent that merely increasing the number of poles is not sufficient and that the armature needs to be rewound to match the new pole configuration.

Next - windings