Hi,

I have to make a testing installation that, on short, will output a current [A] vs. rotational speed [RMP] graph for automotive alternators.

In order to do that I need a shunt regulator that keeps the voltage between alternator terminals at a lower value than that stabilized by the alternator voltage regulator. In this way I can avoid the influence of the alternator voltage regulator.

So I designed my shunt regulator to avoid the voltage of the alternator to increase above ~13 Volts and to withstand up to 150 [A]. In the attachment you have the schematic I used for the shunt regulator.

Now the problem is that the schematic works fine until 5-6 A. Beyond 5-6 amps the oscilloscope show me some "holes" in which the voltage drops to 5-6 volts. If I try to further increase the voltage across the shunt regulator, the oscilloscope show me that the voltage settles to 5-6 V.

If you have any suggestions about how I can make the shunt regulator work fine at high currents please do not hesitate to advise me.

 

Other than D12 appearing to be backwards, your circuit looks reasonable. If you can post a scope photo that might help in getting a diagnosis.

 

Indeed, D2 was drawn backwards, but in the real board it was connected in the direct polarity. Anyway, I reattached the correct version of the schematic.

Now I removed the old pulley from the motor and I intend to reclace it with a smaller one, because at small rotational speed the motor doesn't have enough torq. After that I will make some pictures for you.

 

Please explain the purpose of C3: it takes D12 & D2 out of the bias equation.

 

Please explain the purpose of C3: it takes D12 & D2 out of the bias equation.

It was a drawing mistake. Check again the schematic in the attachment. C3 was intended to lower the chance of self oscillations.

 

Since the circuit should not be on at 5 or 6 volts it sounds like maybe the zener is in backwards so it is on at a much lower voltage.

 

ronv's point is very well taken. You should be able to test this with any current limited power source, even a wall wart with a series resistor.

Take some voltage measurements around Q1, that should tell you the story of what is actually going on, or post them here for more eyeballs to check.

BTW, it is better to post new schematics rather then edit to old ones. It preserves the context of other posts in the thread. I would just touch the orginal posts to note changes are shown lower down.

Good luck!

 

ronv probably missed that I already said that below 5-6amps my shunt regulator works well (it mantains 13,2 V across the alternator terminals).

 

Hi,

Now the problem is that the schematic works fine until 5-6 A. Beyond 5-6 amps the oscilloscope show me some "holes" in which the voltage drops to 5-6 volts. If I try to further increase the voltage across the shunt regulator, the oscilloscope show me that the voltage settles to 5-6 V.

I was talking about this statement. Maybe you meant the current settles at 5 or 6 amps? If so maybe that is all your system can put out. How big is the motor driving the alternator?

 

No, the 5-6 volts is only the limit beyond which the voltage across the alternator, from time to time, drops to 5-6V.

I have an 1500VA/3000RPM motor driven by a frequency convertor which let me to adjust the motor speed as I please. The motor and alternator pulleys have different diameters: the alternator spins 2.7X faster than the motor. The frequency converter protection disconect the motor power when the motor reach a speed of 550-600 RPM. I've made some calculations and I find that:
- the motor is capable of around 5 Nm of torque ar 3000RPM;
- the power given by the alternator when the motor has 550RPM + some mechanical energy loses would force the motor to create a torque of around ... 5 Nm;
- from the above observations I can say that at 550-600 RPM the motor reach its maximal torque.

That why now I intend to reduce the pulley ration to 1:1. After that maybe I will try to use a powerfull motor, but in the first place I need to see the shunt regulator working.

Now, I don't thing that the small torque of the motor is the problem. Even if the alternator couldn't give more than 5-6A, the voltage shouldn't drop below the set voltage of the shunt regulator (13.2V).

 

At 600 motor RPM your alternator is currently running at 1620 RPM?

That's way too low. A car alternator has approx 3:1 gearing from the crankshaft and won't make any decent amount of power until the engine is about 1500 RPM, which is 4500 RPM at the alternator.

And for peak power (150A!) out of the alternator, the engine RPM may need to be >2200, so alternator is at >6600 RPM.

 

Lets try something different.

When from time to time the voltage is 5 or 6 volts, what is the voltage at:

1- Base of Q1

2- Base of T2

3- Base of T3

4- Emitter of T3

 

I'm sorry for the image quality. I made several movies starting at 0 RMP until the oscillations are visible:

Voltge accross the alternator terminals
https://www.youtube.com/watch?v=E-jMiLmJY9Q

Voltage on the BD140 base terminal
https://www.youtube.com/watch?v=iseLjsya35w

Voltage on the BD140 collector terminal
https://www.youtube.com/watch?v=adFpqzF7h_4

Voltage on the T2 emitter terminal
https://www.youtube.com/watch?v=LUcetFbls38

Voltage across the alternator voltage (with 10nF connected in parallel with the zenner diode)
https://www.youtube.com/watch?v=eVf06_yhQB0

Now I think about addition of an low pass filter set at 1-2 Hz at most.

Do you have any suggestions ?

 

Yes, I see (I think). It oscillates. When the voltage gets high enough to turn it on the alternator voltage drops and it turns off because the voltage across the zener drops as well. Add a big cap - say 1000Ufd. from the base of the first transistor to ground. It will still pulse a little on the ripple but it should be stable.

 

Do you mean a capacitor connected between the base of BD140 and the ground ? But in this case, when I first connect voltage accross the shunt regulator, that capacitor will drive the output stage to saturation.

Now I think about tying to insert a low pass filter between BD140 collector and the base of the first TIP35. The LPF will have 2 filtration cells, each one made with a 5ohms resistor and a 1000uF capacitor. With this values the filter will cut almost all signals that have more that a few tens of Hz.

Anyway, it you have any other ideas, feel free to share it with me.

 

Yes, Ok. I thought you ramped up the alternator like in the videos. If you put a cap on the collector of T2 you should probably add some resistance in series so the transistor is not turning on into the capacitor. R6 could probably be bigger to accommodate it.

 

Finally it works ! It still has some small peaks but theire are too small and rare to influence the precision of the measurement.

So, THE_RB, my alternator (rated at 110 amps) gives me more that 200A before reaching 2000 RPM. I found that by measuring 80mV on a shunt of 70mV/200A.

Now I think that the 110amps rating refers to a long term work, because I don't think that my alternators will gime me more tham 200amps for long periods of time without damaging the coils or the diodes.

 

Do I hear a high power stereo?

 

Since you have my curiousity what voltage are you at when you are pulling 200 amps from a 12 volt alternator that was rated at 110 amps?

Also are you powering the rotor at the normal 12 volts or just letting the alternator do what it can at what ever voltage you are pulling it down too?

 

...
So, THE_RB, my alternator (rated at 110 amps) gives me more that 200A before reaching 2000 RPM. I found that by measuring 80mV on a shunt of 70mV/200A.
...

Not when supplying 14.5v it doesn't.

You can't just connect the alternator output to a short circuit to measure it's output current (power).

They are designed to give peak power with engine RPM >2000 RPM, and there is 1:3 pulley gearing (so alternator peak power is generally designed for >5000-6000 RPM.

There's no way they are going to use windings that generate 200A 14.5v (2900W !!) at 2000 RPM for something designed to be running in the 5000-6000 RPM range all its working life.