Greetings,

I am trying to design and build a 6 Volt Positive ground voltage regulator for my 1949 Allis-Chalmers Model C tractor which is equipped with a 3rd-brush type generator (dynamo). This 3rd brush generator is inherently self-limited regarding current (amperage) output, so Voltage limiting is the only regulation required.

The generator is of the “A” circuit type meaning the field coil is internally connected to the generator output and the field current is controlled by externally grounding via a resistor or alternatively bypassing the field resistor, thereby “full fielding” the unit for maximum output.

This ‘regulating’ was accomplished on these old tractors by manually selecting the field circuit via the three-position light switch; pushed all the way in is Low Charge of ~3 Amps, middle position is High Charge of ~10 Amps, and all the way out is High Charge plus the headlights are On. A very primitive system indeed, but it functioned satisfactorily so long as the operator was sufficiently attentive.

There are commercially available units that do exactly what I want, but they cost $Dollars, and as an impecunious retiree who likes to tinker, I prefer to spend my time rather than $Dollars. Also, since my light switch has failed, this is a good time to upgrade.

My design goals are as follows:
1) Reliability. This is by far the most important criterion.
2) Simplicity, following the KISS philosophy with low parts count.
3) Readily available discrete components, no custom PCB.
4) Small form factor to fit inside the original Cutout Relay box. This may not be practical however.
5) Retain the 6-Volt Positive Ground feature.

I desire your help in selecting components and improving my circuit. This is just a starting point that I did on my outdated Multi-sim student version from when I went back to college some 15 years past, and it has a very limited selection of components. The circuit did function in simulation but I had to substitute a resistor in place of the field coil, and I had to remove the cutout diode D3.

 

I designed one for a generator field that covered 24v to 240v generator field voltages :

 

Controlling the field is the more efficient way of regulating an output voltage. A shunt regulator, in the form of a zener diode on a massive heat sink, is a cheap and less expensive way to regulate an output. The flaw is that, on those motorcycles, the connection to the shunt diode can overheat and open circuit. This usually happens while driving fast at night. The result is usually that the lights fail because of excess voltage. (The tale of the linear regulator fix, and the customer satisfaction, is not rated for this website)

 

Controlling the field is the more efficient way of regulating an output voltage. A shunt regulator, in the form of a zener diode on a massive heat sink, is a cheap and less expensive way to regulate an output. The flaw is that, on those motorcycles, the connection to the shunt diode can overheat and open circuit. This usually happens while driving fast at night. The result is usually that the lights fail because of excess voltage. (The tale of the linear regulator fix, and the customer satisfaction, is not rated for this website)

Yes, I am definitely planning to control the field as shown on my rather primitive schematic. I have ordered some components from Mouser which should be here in a week or two so I can build a prototype.

Is there any benefit in adding a capacitor to the base of the transistor?

 

Not on the base of transistor Q1 (2N2222), since it is switching. BUT a capacitor on the GATE of the OTHER Q1(HEX-10) could slow the on/off and reduce any inductive transient a bit. BUT there is a trade-off between time in the linear region between cutoff and saturation versus heating. That means some analysis is needed. The circuit in post #1 has a diode D2 bypassing the field coil Almost a direct shorted circuit

 

Not on the base of transistor Q1 (2N2222), since it is switching. BUT a capacitor on the GATE of the OTHER Q1(HEX-10) could slow the on/off and reduce any inductive transient a bit. BUT there is a trade-off between time in the linear region between cutoff and saturation versus heating. That means some analysis is needed.

Thanks for the response MisterBill2, I believe you are referring to the schematic from MaxHeadRoom rather than my schematic which is in post #1.

I have ordered some components from Mouser which should be here in a week or two so it will be a while before I can build a prototype and begin testing. I will post my findings and hope there isn't too much smoke

 

You are correct!! That comment is about the circuit from Max, shown in post #2. Certainly I should have mentioned that. My error.