How does this alternator regulator work?

[ELS122]

I reverse engineered an alternator regulator circuit from a soviet car (VAZ), and I just can't wrap my head around how it works... that also means that I can't draw an easy to understand schematic for it 

here's the schematic:


I could only read a few values of parts because they're all hidden bellow the riveted-in board, (the 100 ohm resistor was especially confusing, read "100R ... K"
and the only transistor I knew what it is was the K3107, the others I have no idea if their NPN, PNP, or Fets. they're in a TO220 like package
and the pinouts of the other transistors I have no idea if they're correct too,  I just used a random soviet TO220 package transistor's pinout.
the transistor who's base is connected in the schematic to the 67 terminal is mounted to the case, the rest have no heatsink at all.

the terminal labeled 67 is what connects to the alternator rotor coil, and the 15 is what connects to the battery as power and as the sense voltage.
there's only 3 connections, 67, 15, and the ground which connects trough the case.

if you don't know how alternator regulators work, they power a rotor coil in the alternator with a DC voltage, which then produces a magnetic field and makes the alternate output current, then when the sense voltage gets above around 14 volts it turns off, or reduces the voltage to the rotor coil.

I actually modified it some time ago to increase the regulation voltage to around 15 volts for reasons, and I did this by paralleling a 2.7k resistor across the 1k resistor which already has another resistor paralleled. (I did not put the modification in the schematic)

[ELS122]

After some thinking... I think I finally got the schematic right, although it still seems a bit off...



the KT3107 turns out to be a pnp, my bad... (and I left out the reverse polarity protection diodes there, which btw seem useless, the alternator has a 3 phase bridge rectifier built in so it wont output any negative voltage, besides 2 tiny diodes wouldn't hold a chance against it anyway)

but what a weird way to make a regulator... but kinda typical of soviet circuits imo.
I still don't get why reducing the resistance on the top part of the sense divider INCREASED the regulating voltage... maybe I made another error in the circuit somewhere, but the divider part I checked like 10 times, and I know for sure it has to be correct.

what's the point of the 3.3k in series with the 100nF feedback path? and also the other 100nF from base to emitter, why not connect it to ground, why would AC negative feedback be desired? instead of just filtering any noise to ground.

[SeanB]

Capacitors are there to make the response slow, as you are dealing with a massive inductive load, and want slow changes in current. The capacitors slow down the turn on and turn off, the resistor in series limits the rate of change of the high current in the output stage.

How it works is that the alternator warning light provides a small current to the 15 terminal, so the power transistor gets turned on hard, and the rotor gets a small current from the bulb, so it can bootstrap the alternator to provide the excitation voltage on the separate 3 small diodes in the pack parallel with the 3 large diodes connected to the positive output terminal. As the alternator output rises the high current diodes start to feed current into the battery, and the positive side of the small diodes follows this, but isolated from the battery voltage, but tracking it. PNP transistor is then going to start to have base bias removed from it by the zener diode and forward biased diode, with the 3 resistors on the left providing a sample of the voltage to it. In turn as the PNP transistor starts to turn off, slowed down by the 100n capacitor between base and emitter, it starts to turn off the NPN darlington output stage, the base being pulled down to ground. The series capacitor and resistor provide AC feedback so the PNP turnoff is not affected by the noise on the rails much, so that current in the rotor drops, and then it stabilises to around 14V4, with the rotor essentially running with zero current in it, just enough to keep the bridge rectifier conducting slightly to keep the regulator powered. The diodes on the terminals are there because the rotor has a lot of inductance, so the diodes provide a path to clamp it so the inductance of the rotor and wiring does not damage the regulator.

Normally the power diodes in the alternator are designed to break down at around 20-40V across them, for at least the 3 in the positive side, so that if the battery is disconnected, with engine running, the alternator does not attempt to charge into an open circuit, which can easily reach 100V plus on the 12V nominal vehicle wiring. Instead it is clamped roughly to around 40 to 60V, limiting the damage.

[hid3]

ELS122, that's some interesting info you provided. Which particular regulator from VAZ did you draw the schema for?

I am trying either to make or find an existing regulator to replace electromagnetic relay based regulator in ZAZ. Rotor's coil resistance is 7 Ohms.
Also, I'd like to increase output voltage on Terminal 67 up to 15-15.5 Volts in order to increase rotor's excitation as much as possible (within safe limits).
The reason for that is that this alternator due to it's nature hardly provides any charge below 1500 engine RPM...

Thanks