Bosch Regulator REF3639A

Regulator striptease!

Top cover just levers off after removing the tape:

^ The two rivets at base of coil are an electrical connection and may have some resistance due to corrosion, on my spada’s reg I had to solder a wire direct between this contact and D+, thus bypassing this riveted connection.

^ I think the green tubular thing is a RF (radio) noise suppressor choke (viz, a simple small coil with a low inductance, here wound onto a tubular former) ~ warning: at least one connecting wire is very thin and delicate ~ all the output current to DF goes through this…

^ Square metal thing with two white wires is ack-shirley a high-power resistor. ‘R1’ in following schematic.

Internal schematic:

Connecting DF to ground (D-) when switch goes to contact ‘2’ seems very odd, as this appears to short-circuit the rotor field winding for no obvious effect. However when the D+ source current to the rotor coil is switched off, its magnetic field collapses, which induces a current in the coil in the same direction, but which then presents a negative Voltage at DF (aka, back EMF). Shunting this output to ground (D-) keeps the current flowing at the same sort of Amps as was going in. Due to inductive reactance, the coil’s current does not start or stop instantaneously. Therefore, (ideally) the regulator keeps switching the field winding on and off (contacts ‘1’ and ‘2’) to maintain the rotor’s current at some mean average value, which is directly determined by how strong a magnetic field is required to maintain the alternator’s output at whatever load. When the load changes, the field rotor’s mean average current is increased or decreased to compensate.

I should mention that electronic regulators are much better at doing this as they can switch very much faster than the electro-mechanical switch, typically hundreds of times a second, which makes for a much smoother and steadier rotor field current waveform. Also I was surprised by the 1.5V difference between the low Voltage threshold level (12.5V) and the high Voltage level (14V). My electronic regulator has a difference so small as be almost impossible to deduce, I figure 0.1V or maybe even less.

FYI, while the electronic regulator feeds from D+ to DF using a transistor switch, the back EMF shunt to D- is done by a reverse-biased power diode, which becomes forward conductive when the field winding goes negative during the ‘off’ part of the cycle.