Final touches for the generator.

Balancing the armature

The bearings and fan blade are installed onto the newly varnished armature. The coil ends have been secured to the commutator segments and it is ready for testing, However, the armature is slightly out of round and not particularly well balanced at this point.


Testing

I installed the bearings on the armature and mounted it on my quickly assembled stand. Spinning it numerous times revealed that one side of the armature was heavier than the rest. It would consistently stop with this area on the bottom. At times it would stop in another position and then slowly roll backwards to it's favorite position, the heavy side down. Furthermore, tapping on the stand would cause a motionless armature to rotate towards the heavier side.

To counteract this imbalance, I cut and reshaped pieces of lead so I could wedge them into the gaps between the coils. I then placed them on the light side of the armature and spun it numerous times, refining my placement of the weights until I was unable to determine a heavy side. I then epoxied the weights into their proper positions.


The arrows point to places where lead has been inserted and epoxied into place for balance. With any luck, the balance is good enough to keep vibrations to a minimum and allow longer life for the bearings.

Grinding those magnets

In installing the armature, I discovered a slight bit of rubbing between the armature and the iron magnet extenders. I knew that it was a very small air gap and so was not particularly surprised. I looked around my workshop and found an old 7" abrasive disk that had worn down to a 6 1/8" diameter. Mounting it on a " mandrel gave me a means of reaming out the inside of the casing and shaping the magnets to a circular inner face. With the armature outside diameter being 6", the 6 1/8" disk would give me an air gap of 1/16" which is still close but shouldn't rub.

I mounted the disk on the drill press, secured the shell securely and spent the next 3 hours slowly reaming out the magnets. I set the drill press to 700 rpms and proceeded very slowly to avoid heating the magnets beyond "warm to the touch". This was to keep the epoxy solid and to keep the magnets strong. In retrospect, it was wise to put the iron spacers on the inside rather than between the magnets and the shell. I would not have wanted to be grinding directly on the magnets.




View of the grinding process. Note that the case is not entirely round, nor are the thicknesses of glue, magnets and iron bar all uniform. This results in uneven results when grinding. Some bars are shaped across their entire face, others are untouched. It should result in a more uniform air gap however.




A view of the finished product of grinding. Note that the left and right (of the photo) bars have been fully ground across their faces. The upper bars are only partially ground. Before assembly the iron filings which fill the gaps were removed from between the magnets.



The generator shown assembled for testing. The threaded bar holding the ends together is not original. The original motor had two long bolts that connected the ends inside the case, passing between the pole pieces. With the new magnet arrangement, there is no longer any space between the poles. Luckily, the face plate is oversized and has tapped holes I was able to use for the new exterior straps. Note the small 12 volt battery in the background. The original motor turned very fast when connected to this battery, in fact it jumped when first connected (bolt that thing down first !). When rewound, connecting to this battery resulted in 160 rpm clockwise and ~240 rpm counter clockwise. Results are shown in a table below.


Method

Rpms

voltage

Stage of completion

12 volt battery

~1700

12

Original state as from junkyard

Hand spinning of shaft

0 to 160

< 1

Original state as from junkyard

12 volt battery

Would not turn

12

20 magnets, original armature

Hand spinning of shaft

0 to 160 ?

0.5

20 magnets, original armature

12 volt battery

160 clockwise

12 (2 amps)

20 magnets, rewound armature

12 volt battery reversed polarity

~240 counter clockwise

12 (6 amps)

20 magnets, rewound armature

Hand spinning of shaft

0 to 160 ?

up to 14

20 magnets, rewound armature



Tuning the brushes

Because this is a generator with a commutator, the positioning of the brushes in relation to the poles is critical. This isn't the case with alternators, but this is what I have and it needs to be adjusted properly. The only problem is that I don't have a clue how to do it. The following is the process I used to determine the best placement for the brushes.

Below is the very simple setup I used for hand turning the shaft. I used a 9" pulley to give me enough leverage to give it a good spin. At each setting (shown next photo) I gave the pulley a strong clockwise spin and observed the bounce of the voltmeter needle. I rather arbitrarily labeled one of the leads from the generator + and the other - for testing purposes. I connected the "+" lead to the "+" lead of the voltmeter and likewise with the "-" lead.


Once the generator was assembled, it was necessary to adjust the placement of the brushes in relation to the magnets. This was done by turning the rear plate (which the brushes are attached to) in relation to the shell casing (which houses the magnets). This yeilded some very interesting results. The back plate of the generator is shown below with markings on the tape relating to the voltage I was able to generate by hand turning the shaft. I tested turning both clockwise and counter clockwise. The voltages for clockwise are shown on the top of the tape. The counter clockwise readings are shown on the bottom of the tape. The polarity of the leads was kept the same for this test.


It should be noted that when in the 14 volt position, it was relatively easy to turn the shaft clockwise and difficult to turn it counter clockwise. As the adjustment was moved further to the right it became more difficult to go clockwise, but became easier to turn the shaft counter clockwise. The results are shown in table form below.

Position from left

Clockwise voltage

Counter Clockwise voltage

Clockwise torque required

1

9

- ?

Some resistance

2

14

- ?

Easiest

3

9

2

Some resitance

4

5

5

More resistance

5

- ?

7

Hard

6

- ?

9

Very difficult

7

- ?

7

Hard

From the standpoint of being spun by a relatively small force, like the one my 7' three blade rotor will produce in my light wind area, it would appear that the 14 volt setting would be the most likely to give decent results. Still, I needed more information.

Motoring

I hooked up an ammeter, switch and battery to the generator in series to find out more. I loosened the straps holding the rear plate (and the brushes) so I could turn the brushes while motoring the generator. I read the ammeter (which had a scale of 1 to 10amps) while motoring and turning the rear plate. The voltage is coming from the battery which we will assume is 12 volts, although the voltmeter says 13.5VDC when I test just the battery. I discovered that there is a range of position (shown around the 14V mark on the tape) where the speed when motoring appears fairly constant near 160 rpm. Within this range, the current draw and noise does change but only slightly. It appears to draw the least amperage (2 amps) and run quietest at the left side of the range.

At the right side of the range it draws 3 to 3 amps and sounds coarse. It gets louder, slower and draws more current as the plate is moved to the right out of the range. Moving the plate further left results in a rapid increase in speed and current draw up to about 240 rpm (6 amps). Turning even further than this pegs the 10 amp ammeter and causes the speed to slow, stop and then reverse.

There is a slight to amp bump that occured throughout the test which was smallest when the draw was 2 amps.

Testing with a fan and the final placement of the brushes

OK, so there've been all sorts of tests, but what really can be used to determine the best placement of the brushes. That would be the place where the most power is generated with the least amount of power input. The only way to determine that is to put a measurable load on it and spin it. Since I was limited to hand spinning at low speed (under 100 rpm) I had to find a really light load that I could measure visually. I found it in a spare wafer fan. Wafer or muffin fans are the small flat fans used in computer cases. They usually are rated for 12 volts. The one I had showed a rating of 12 volts .2 amps. I don't know what speed that was rated for and I really don't care.

I hooked up the fan and spun the generator. The fan didn't spin so I reversed the leads and spun the generator again. This time the fan spun quickly, even with just a partial turn. I had found my final test. I knew where the most efficient motoring position was, and I knew the easiest spinning position. They were the same place, which was where it was set now. I moved the brushes each direction to where more amperage was required and spun again and again, watching the movement of the fan each time. Each time it took more force to move the fan less than it had at the original position. This confirmed that the most efficient motoring position was also the most efficient generating position. It was also the quietest operating position in both modes. So now I know that this is the most efficient placement for the brushes. I don't know how much voltage I can get at 300 rpm, but I think I have it adjusted optimally for generating.

Testing the generator and final adjustments