Lets have a look at some simple (poorly drawn) schematics of what we are going to play with.
First here is a simple diagram of the business end of the alternator. It shows the stator coils in star form, with three outputs U V and W and the neutral point as SP (star point?)
It is a simple matter of a stator with a rectifier, either 6 or 8 diodes in the diode blocks. The rectifying power diodes are either pressed in or soldered into the heat sink. There are two of them, as one heat sink will be the batt- output and grounded to the casing, and one heatsink will be floating above the casing (electrically) and becomes the take off point for the big B+ battery connection.
By necessity, the diodes are of two types. One group on one heatsink is cathode to casing of the diode (for the B+) and one group is anode to the diode casing and so this becomes the B- heatsink (grounded)......... So if you have to replace a diode or two, you need to know which heat sink it is from to get either a cathode to case or anode to case replacement from another alternator.
The diodes press out ok, but you need to unsolder the stator from the diode blocks to get them out where you can terrorise them properly. Support the aluminium heat sink as close to the diode casing as possible, so you can drift or press it out without deforming the aluminium.
In the case for the soldered type (my favourite), then a good heat gun will heat the heat sink up easliy to solder temp, and it falls out/off, and then simply sit the replacement in and add more solder.
So thats the diode block. The six diode ones are the same setup as the ones we use for rectifying the output of a three phase alternator wind generator.
The 8 diode ones are essentially the same, but they have 2 extra diodes in the same orientation as the six units, but their leads go to the star point, instead of one of the phase outputs as do the other 6 diodes.
This is for a star/delts switch. ....... If the voltage drop in the stator over a two phase group (to make up a star phase) gets excessive, then the star point starts to conduct, and this turns the stator into a delta setup. It does two things, firstly it lowers the resistance of the stator winding in half (resistance ...... but also due to phase addition, two legs provide extra current to help the third leg... we can get twice the current), but more importantly, it lowers the AMP TURNS in the stator. This allows the magnetic field of the rotor to better concentrate the flux into the coils of the stator...... short explanation:
These alternators are the same as the air coil alternators in their schematic..... but they follow entirely different rules for their max power.
For the axial we usually build, we are concerned with the stator resistance. The emf in the coils less the battery voltage x the current in the coils will dictate the output. There is very little armature reactance to worry about.
With iron cored alts like these, it is the same thing for low power, but as we increase the current in the stator, we set up a very powerful back MMF (magnetomotive force). This is in opposition to the magnetising field of the rotor.... so for a set rotor current (magnetising force), the back MMF will fight it more and more as the current in the stator increases. (the current in the stator produces it's own field, amplified by the core of the stator, and reflecting at the rotor teeth.... we have no concentrator in the axials, and large air gaps, so it has little effect)
This is easily seen if when we build a motor genset like we are going to....... if we make it as simple as possible, and just wire the brushes directly to earth and b+ but with a 5w globe in series as the brushes, we will have a set current in the rotor. We may see 5 amps at 1500rpm into a flat battery load...... if we increase the rpm we expect the current to rise..... but it wont. We can rev up to 12000 rpm, and there will be barely any change at all. This is because we expected to increase the EMF with revs, but instead of this happening, the back MMF cancelled any further current in the stator, so the magnetising force was unable to further influence the stator..... we flat line.
We get over this to some extent by using a regulator instead of a constant current source like the globe. Now as we want more power past the 5A, we can simply pour more voltage into the rotor. It will draw more current which in turn gives us more MMF to aim at the stator.... so we have effectively driven the constant current point higher up the amp scale.
When the rotor has the full 12v across it, we are at max current from I=E/R... so if rotor resistance is say 6 ohms, and voltage is 12v then current max will be 12/6=2amps. If we have a lower resistance rotor the same formula applies. In a small auto alt, 5A seems to be around the maximum current draw, and so our upper current limit seems to be about 35-55 amps. If we drove the rotor at a higher voltage, say 24v, then this may increase a bit, but the rotor may suffer overheat in this instance. This can be fixed with either a fixed upper pulse rate in the regulator (volume control thing) to keep to 5A, or a resistor in the line to the rotor to keep the current in the design region...... or even another rotor coil outside the alternator. As was seen in the previous post, the rotor presses apart, and so if using it for 24v operation only, then possibly the best thing is to simply rewind the rotor with more turns of thinner wire.
The "pot assisted" regulator will alleviate most of the 24v rotor problem.
They design these things to run at or near saturation. So if 5A is saturated, then driving it harder will do little if any good. In this case, we use the pot to drive the regulator to drive the alternator at 24v to near saturation and current limit, and no more. The rotor current should still be around design. A normal 24v reg will overdrive in a large bank and stationery situation, as it will turn on fully into a flat bank, and it won't know the extra rotor current is doing no good.
So EMF is not the problem, we can get heaps of voltage out of these things, but current will be limited at some point.
It is unwise to run the alternator unloaded, as if you rev it up a bit, the high emf may destroy the main diodes. This is probably a cause of most diode faults in car alts.... unless you decide to replace the battery back to front.... that will be a destructive excercise as well.
Here is what we will see in an internal regulated alternator. There are a further three diodes of low amperage that connect to the U V and W outputs like the main diode block.....3-6 amp diodes.
They are hard wired to the regulators positive input. They provide the excitation current for the rotor.
To get it started at low rpm, the small indicator light on the dashboard of the car will light up when the key is turned. This is because the light is connected to the B+ of the key switch, and the regulator tries to ground the rotor negative slipring to ground. The current going through the light to the rotor positive slip ring. The light sees this as low impedance load (short?)..... and so it thinks it is between b+ and ground.... it lights up.
It is a very small current, but enough to make some field in the rotor.
As the car engine starts up, the potential in the stator starts to rise under the influence of the revs and the small field in the rotor. It is below "cut in for the batteries" and so as the field builds, the small diodes direct all the emf into the rotor and it quickly causes the stator to increase output.
As this voltage in the stator rises to battery potential, we can see that the point the light is connected to in the alternator is now at around the same potential as the b+ from the battery, and as such no more current flows in this circuit.... the light extinguishes and the system is running normally. From this point on, the three diodes provide the excitation current for the rotor, and it is self sustaining.
We wont avail ourselves of this, and the three diodes are not useful to us at all.. but for the sake of completeness I described it's operation.
Note: if when testing a car alternator, if you short this point to b+, and the revs are above idle, these little diodes will try to charge the battery as well..... and they will die.. Thats why we won't use them here, they would be effectively wired to b+, and try to charge the battery, rather than just run the rotor field requirements. They need a highish impedance to b+ (small globe on the dash) so they cant try and play with the battery directly.
We will connect the regulator + directly to B+ AND the brush+ for the rotor when we hook the battery to the charger, and that will provide the necessary drive for the rotor b+. Then we need only pulse the b- rotor brush to ground, and we can control the output...... and we can do it with this little circuit.
This circuit as is, will replace any internal regulator that has failed.... it is that simple.
To take control of the voltage points, we need only put a potentiometer in series with the voltage divider formed with the 1k and 1.5k divider network on the input of the zener.
If we place a 5k pot in series with the 1.5k resistor to ground, we effectively give the 1k resistor a bigger say in the divided voltage that the zener sees. If we trim the pot to zero ohms, the circuit will regulate to 14.4v or so, if we turn the pot to higher resistance, the divided point will increase voltage, even though battery voltage remains constant, so we fool it into thinking the battery is more highly charged than it really is. With 5k in series, we can get the voltage down enough to charge a 6v battery , at fully off, it will charge a 12v battery.
The big problem with these motor generators is starting the charge on a flat battery. The regulator will naturally turn on fully to charge the flat battery up as fast as possible. This usually loads the motor up very severly.
If we have a "volume" control on the front, we simply turn it to zero, start the motor and load it as we please until it gets charged up a bit, and then turn it to full for unattended regulation to 14v or so.
I have to end this at this point... things to do....
Comments and corrections welcome
...............oztules