I own an older boat with in inboard engine. It used a magneto charging setup on the harmonic balancer to develop 12vdc used throughout the boat. Originally, the design used a seriously heavy duty resistor and power transistor that 'clamped' the magneto output at 14v as the engine rpm increased. This unit was water cooled from the engine's coolant system and frequently failed by becoming an open circuit and allowing the voltage to rise to 40 or 50 volts! Most owners of these systems simply toss the whole charging system overboard and install a belt driven alternator.

I rather like the concept of magnets inside the harmonic balancer rotating to produce charge on the engine mounted stator. This takes up little valuable engine bay space and it appeals because of its design efficiency. The problem is how to regulate the center tapped AC stator coil? Unlike an alternator, the flux is developed by magnets so I can't control the field. A mosfet power transistor is one idea, but I'm not sure how to design a circuit such that the input can vary from 12 to 60vac while maintaining a steady 13.8vdc output all while being able to supply up to 40 or 50 amps into a changing load? This system must charge the battery--a group 31 size--and run onboard electronics--radio, lights, fans, etc safely. Ideas?

 

Welcome to AAC!
Sounds like you need a hefty bridge rectifier plus a buck-boost converter. Finding one to handle 50A might be difficult/expensive, but perhaps you could use several smalller ones (such as these) to drive individual loads?

 

Such a generator is similar to those used on most motorcycles, which always use a shunt regulator to control the voltage since a series regulator will experience a very high voltage when you try to limit the current (as you experienced even under load).
It likely can go higher than 40-50V at high RPM since the output voltage of such a generator is proportional to speed.

Such a setup is much less efficient than a normal alternator, where the field can be varied to control the voltage and my suggestion would be to go to that for the high current you need.

If you still want to try with the built-in alternator you might look for a motorcycle regulator used on some of the larger types, such as the Harley Hogs. You just have to make sure it's a shunt-type. Of course that may cost more than adding an alternator.
Do you know if the output from the windings is 3-phase? (How many winding wires are there?)

 

It might have been better if it was three phase; its a center tapped set of coils series wound around the circle, just like a lawn mower, or perhaps as you say, a motorcycle--three wire output, one wire to ground, the other two to a set of diodes and then on the the shunt regulator. It is indeed inefficient to shunt the excess--that's why they water cool the thing.

I didn't know anyone else built this sort of thing--I had chocked the design up to Mercruiser's mechanical engineers doing the best they could at the time. The regulator gets so hot, it heats up the side of the block where its mounted. That's why I was convinced there had to be a better way. If they mass produce these things for motorcycles, then perhaps this issue is something that is begging for a solution beyond just my need. Why couldn't a mosfet or maybe an ibgt be used for this purpose? I'm thinking maybe of a capacitor for a bit of storage, then just series feeding a mosfet to the load. Would this work as long as I kept a small shunt load on the mosfet?

 

You'd waste the same amount of power (generate heat) as you are now trying to throttle the voltage by way of a series mosfet or igbt.
They would have to be on big heatsink if operated outside of fully "on" region.
Whatever you choose to do you will ultimately come to understand why so many people switch over to an alternator.
Cost and trouble leading to pulling your hair out and the nashing of teeth...
It is a journey of re-discovery for some.

 

They do use MOSFETs for the motorcycle regulators but they still dissipate a fair amount of heat, especially when you are shunting several tens of amps, probably nearly a watt per amp.
Here's a article I wrote on a paper design for such a motorcycle regulator which may help with your understanding.

 

It might have been better if it was three phase; its a center tapped set of coils series wound around the circle, just like a lawn mower, or perhaps as you say, a motorcycle--three wire output, one wire to ground, the other two to a set of diodes and then on the the shunt regulator. ................

If it has three wires and you can lift the ground wire, then you may be able to operate it as a three-phase circuit.

 

The requirements at that design time was small.

To supply the power required today........install alternator.

 

The requirements at that design time was small.

To supply the power required today........install alternator.

Actually of course, with a shunt regulator the larger the load current, the less power dissipated in the regulator.

 

I don't know your environment, but let's say it's nighttime (lights), pouring down rain (pumps, wipers), channel limited with heavy traffic (radio/radar), and you have a 30 min. no-wake zone to transverse.

Will your battery be fully charged at the end of the 30 minutes?

What if your out in high water? How long can battery supply all needs at half throttle?

It's best to plan ahead when going into foreign territory.

Redundancy and tools are necessary on boats. Murphy lives there.

Having no power out on the water is no fun.

 

There are a number of ways to work with your system. The simplest I can think of is to use SCR's in place of the two diodes and use a basic phase angle firing voltage regulator system.

It's basically an SCR based rectifier circuit for AC to DC conversion with a feedback loop that keeps its DC output voltage stable regardless of the variations of the input voltage and frequency or output load demand.

For your system it would probably be the most efficient and easiest to work with.