This article is not exactly talking about the same thing I asked about but close enough to be relevant and very helpful.
Starter Generator Assembly
In this article we are going to look at the starter generator assembly, how it is made, how it works, how to test it and how the voltage regulator controls it. This starter generator unit combines two items in a single unit, giving it it's unique name. There are four basic combinations, short frame [about 6 1/8" field housing] low torque and the high torque long frame [about 7 ¼" field housing]. The other option is the direction of rotation. Cub Cadets and some other applications put the drive pulley on the PTO end of the engine instead of the flywheel end. These units rotate in the opposite direction. These can be reversed by installing new field coils in the unit. The short frame units were typically installed on the 8 and 10 HP vertical shaft engines and the long frame units were usually on the cast iron block horizontal shaft engines. The longer frame allows for a longer stack of laminations on the armature and longer pole pieces and field coils, resulting in increased torque and generator output.
The starter consists of an armature that is connected in series with a field coil that is wound of very heavy wire. Probably about 10 turns of 1/16 X ¼ inch copper buss. As can be seen in the following sketch the voltage applied to the A or Armature terminal goes through the field coil to the brush to the armature. A generate field coil is also connected to this brush. This coil is wound of many turns of fine wire. The other end of this coil is connected to the [F] or Field terminal. The [F] terminal is normally grounded through the voltage regulator when the unit is not generating. When a high current source such as a battery is connected to the [A] terminal the current flowing through the series [starter] field provided a very strong magnetic field on the pole piece to attract the field created in the armature. [See previous article on motor and generator principles]. Since this is a series field and it has VERY low resistance, it is capable of carrying very heavy currents and the more current that flows through it the more magnetism it creates. The more load on the armature [i.e. the harder the engine it is trying to start turns over] the more magnetism is created in the field because the motor draws more current.
To illustrate how a little extra resistance, maybe just in a connection, can seriously affect the performance of a starter circuit, lets examine a hypothetical case and see what happens. Remember the power formula P = EI and 746 watts = 1 horsepower. If the engine being started turns over hard it requires more horsepower to turn it. Since the battery voltage is fixed [12V] and it goes down with load, the limiting factor is the available current and the efficiency of the starter motor. The amount of current is controlled by the TOTAL resistance in the circuit and the current available from the source. [battery]. To illustrate how critical resistance in these circuits is, lets look at an example of what happens. Remember OHMS LAW: E=IR. Lets pretend our starter motor is 100 percent efficient [not possible] and we need 1½ horsepower to crank the engine to starting speed. At 746 Watts per HP the power requirements will be 1119 watts. Lets also pretend that the battery voltage will not drop any from a constant 12 volts under the load [not possible] and we will pretend that the cables and connections have no resistance [not possible].
P = EI, I = P/E. Therefore, the current it will require will be [P] Watts 1119 divided by [E] Volts 12 = about 93.25 amps. To be able to draw this current we must have a circuit with less than __?____ Ohms. To find out, let's use the E = IR, therefore R = E/I. So, R = [E] 12 Volts / [I] 93.25 Amps = .13 Ohms. This is assuming in our hypothetical case that the 12 volts was right at the starter. Now lets pretend that our battery cables had .13 ohms resistance also. If you add this to the resistance we calculated that the starter must have to provide 1.5HP (.13, or 13 one hundredths of an ohm) plus .13 more in the cable, then the voltage getting to the starter would be 6 volts. Half (or 50%) of the battery's voltage would be lost in the cable. Putting 12V into this circuit with .26 Ohms total resistance would result in a maximum current of:
E = IR, therefore I = E/R = 12V/.26 ohms = or about 46.15 amps.
Since 6 Volts would be lost in the resistance of the cable and only 6 Volts would be at the starter, if we take the P = EI and calculate the power we get 6 Volts X 46.25 Amps = 277.5 watts or about .375 horsepower. When you double the resistance in this series circuit, the power the starter would deliver was reduced to ¼ of what it would produce originally.
All of this calculation is not something you would do in troubleshooting an existing system, but it illustrates in these high current circuits how a tiny little bit of extra resistance, can result in serious losses in power and excessive voltage drops. If one understands these principles then it is easy to take a voltmeter and identify high resistance areas and trouble shoot your starting system. It is the concept one must understand. Applying the principles of what we just discussed to Simple Troubleshooting will be addressed in one of the next articles. Now, refer back to the figure above and let's see how the generator part of the assembly functions.
Now that we have our engine started, the 12 volts from the starter solenoid is removed from the A terminal and the [F] terminal is grounded by the voltage regulator. The pole pieces that concentrate the field coil magnetic field retain some residual magnetism. As the armature turns, the windings in the armature cut the magnetic field of the pole pieces and this induces voltage in the armature. This voltage is connected to the brushes from the commutator bars. These bars are nothing more than a rotary switch. As the armature turns the bars that are connected to the windings that are cutting the flux at the pole pieces rotate out of the field, the bar connected to the next winding connects to the brush and on and on it goes. As the induced voltage is connected from the commutator to the brush this voltage flows to the generate field winding. This is shunt connected [connected across the armature to the field terminal, where it gets its ground]. The resistance of this winding is normally about 8 to 12 Ohms, such that the maximum current that will flow through it is about 1 to 1½ amps. This current in this coil creates the magnetic field in the generate pole piece. As this field increases the output of the armature increases, feeding more voltage to the field coil, reinforcing the magnetic field, and resulting in more output. When the output gets above 12 volts the regulator connects the armature to the battery and the current flows to the battery charging it. When the voltage gets to about 14 volts the regulator starts to interrupt the ground on the [F] terminal. This controls the magnetic field at the pole piece which controls the amount of voltage induced in the armature. The regulator action caused by the vibrating contacts in the regulator happens at a very high rate, nearly 300 times a second. Please see the section on regulators for more details on how the regulator does this.
The same thing happens when starting. When the voltage from the [A] terminal gets to the brush it connects to the bar connected to the coil that is just approaching the pole piece and the intense magnetic field there. As it approaches the center of the pole piece where it would want to stop, the brush switches to the next bar [segment] and this magnetizes the coil just approaching. So, it continues indefinitely, constantly pulling and when it just about gets there, it is disconnected and the coil preceding it is connected. The brushes don’t really run on the commutator, but the current is conducted through a microscopically thin (just a few atoms thick) cloud of gas that is created when the electrons flow to the brushes.
There are several common problems associated with these units. The amount of current they can generate reliably is limited by the fact that they have only one generate field coil. Regular generators for cars and farm tractors have a fan at the front and the front and rear frames are open to allow cooling air to be pulled through. Since generating electricity creates heat due to resistive losses, the closed design also imposes limits on the amount of energy they can generate. If the commutator gets out of round, the brushes "bounce" and cause the output to be intermittent. One won’t be able to see this as the armature is typically turning about 10,000 RPM when the engine is running at governed speed. This problem gets accelerated if the rear bearing is getting worn. The rear bushing can be replaced and is available in 3 undersizes in .002 intervals. The undersize bushings are not "cut" for the oil wick and need to be before installing. Another common failure is the voltage regulator not regulating properly and the unit running "Full Fielded" and the generate field coil burning out. The field coils are available in complete sets and the generator field separately. The armatures are available remanufactured. The quality varies from manufacturer to manufacturer. We have been buying from the same company for about 18 years.
When repairing one of these units the commutator on the armature should be turned in a lathe. After turning, the armature should be undercut. Undercutting takes the mica separators below the surface of the commutator bars. In generators this keeps the mica from shaving the brushes as the copper wears down. The brushes in these units have some copper added to them to reduce the resistance and handle the starting currents. Regular generator brushes out of a car or tractor generator won’t work. The starter won’t have the torque and they will be very short lived. One other thing, regulators from farm tractors or cars, even though they look EXACTLY the same, are not compatible with these units as the current sections are calibrated for 20, 25 or 30 amps of current and would burn the field coils out of these units. Check the front end [drive end] to see that the bearing has not egged the hole out where the bearing seats. Often the bearing will have up to .010 clearance around the outside race in spots usually on the side the belt pulls the armature to. The best fix is to replace the drive end frame, but sometimes they may be saved with some very high grade Loctite, not the regular stuff.
1. With the engine running more than half throttle, measure the voltage at the A terminal, it should be 13 to 14.5 volts. If it is 12.5 or less, ground the F terminal, if the voltage goes up to 13 to 14.5, most likely the problem is the regulator. Other possibilities are the wire from the F terminal to the voltage regulator, and the ground on the regulator not being good. If the voltage at the A terminal is 14.5 to 17 volts when you ground the F terminal, the regulator cut-out section is probably not connecting the A terminal to the battery. If the unit is measuring 13 to 14.5 volts at the A [without an external ground applied to the F terminal] terminal, the voltage measured at the battery should be within .1 or.2 volts of the voltage measured at the A terminal, if it is the system is working correctly.
2. If the unit fails the above tests, disconnect it and remove it. Using an ohmmeter, measure the resistance between the F terminal and the A terminal. A normal reading would be 7 to 15 ohms. If it measures less than that remove the end frame and pull the armature assemble out and look at the field coil connected to the F terminal. If the insulation is scorched, or it looks as if it has been overheated it is probably shorted.
3. Inspect the brushes for wear or the leads being loose in the carbon. Look at the sides of the brushes to be sure the sides haven't worn into the indentations in the holders causing the brushes to "Hang Up".
4. With the armature out, use the ohmmeter to check from the copper commutator bars to the armature laminations or the shaft. It should show no resistance, infinity. If the reading is 0 ohms you have a shorted armature. If it is 40 or more ohms, it probably has carbon dust between the commutator bars and the frame or shaft. Blow it out all around the commutator and windings. Be warned this dust is black and messy, so do it where it won’t make a mess. Blow the field housing out at this time also. If these two tests are OK, the armature needs to be "GROWLED." This checks for shorted turns, and shorted bars in the armature. This requires a growler to do this test. Any good starter shop will have one. If the field coil needs to be replaced, one will likely need a pole shoe screw remover. This is a clamp device with a screwdriver bit on a wrench that clamps the bit in the head of the screw to remove the pole shoe that holds in the field coil. Sometimes they will come out with an impact wrench, but when the pole shoe is reinstalled one needs to be sure they it is installed perfectly straight or it will hit on the armature if the curve is not exactly parallel with the armature. At this point for most people, this would be a "take it to the shop" job.
5. Check the bushing in the end plate for looseness. If the bushing needs to be replaced one needs a blind bushing puller which costs $400.00, so this is another "shop job."
Check that there is 12 volts to the A terminal of the starter generator unit when starting. If not, measure step by step using the principles outlined in the FORK LIFT troubleshooting article. this will explain how to isolate the external problems. The internal problems usually are: short brushes, worn rear bushing with armature dragging on the pole shoe, shorted armature, or the start field shorted to the field housing. Low cranking torque may be caused by someone putting regular generator brushes in, because they fit. Most likely it is a high resistance problem in the power source to the unit.
One little, off the subject point, about undercutting. In regular car and truck starters, you don’t normally undercut 12 volt starters, because the brushes have so much copper in them to keep the resistance down that it gets in the grooves between the bars and shorts them out, they are also hard enough that they wear off the mica. On 24 volt starters you need to undercut because they add some carbon to the copper in the brushes to keep the brushes from tearing up the commutators. They are softer and the mica gets them, also with the extra carbon the shorting problem is lessened. In many big truck and heavy equipment starters, the only difference between the 12 volt and 24 volt versions is the undercut armature, the material the brushes are made of and using the correct [12 or 24 volt] solenoid. The windings are different in the 12 and 24 volt solenoids, plus the contacts in the 24 volt ones are usually silver, or silver brazed on the copper faces. Neat Huh!
Dealing with the problems of resistance that go with low voltage, high current systems is why the automobile companies are going to 42 volt systems. The automotive industry has recently settled on 42 Volts as the best option to deal with the increased demands for power in the automobile. In our starter trade publications they are saying that in 3 years some of the car companies will have 42 volt systems in the showrooms and in 5 years all will be there. This means they will be able to use significantly smaller wire and cables and to power many more electrical accessories. 12 Volts will soon be as obsolete as 6 Volts is now.