Voltage Regulator for 12V 20A Permanent Magnet alternator

Thread Starter

teliocide

Joined Sep 26, 2013
112
Greetings,
I am in the process of making a very very small camping generator. I am aiming for around 250 watts.
I intend to drive a small 12V 20A Permanent Magnet alternator (see pik) with a Honda GX 25 engine.
I need a voltage regulator. So I hope some one can provide me with a circuit.
Also there will be at times, the need to power a 240 volt AC inverter, so I may need to smooth the supply.
I am hoping I can gain some assistance/guidance with the best way to do this with and without an in circuit Battery.11 Alternator.jpg
 

LowQCab

Joined Nov 6, 2012
2,170
250 -Watts is not very likely ..........
You have plainly stated that the Alternator is Rated for 12-Volts @ 20-Amps, 12 X 20 = 240-Watts.

Its rating is at some particular RPM and Temperature,
if the RPM drops, or it starts getting excessively Hot, the maximum Output will drop.

A performance Graph will be necessary to determine just how much Power you might expect.

This Alternator is designed to be a Battery-Charger Power-Source for
a Tractor, or a Stationary-Farm-Machine of some sort,
it is not necessarily designed for continuous maximum-Output.

For the Power levels You are wanting,
You would be much better off using a generic Automotive-style Alternator,
most of them will provide the Power-Levels You want,
and are much easier to Regulate,
most of them have a built-in Voltage-Regulator.

You must also take into consideration the Horsepower required for the Power-Levels You expect.
The ~1-Horsepower Honda Engine You mentioned is not going to have enough Power
to produce the Wattage You have requested.
It is also not designed for continuous maximum-Output.

2 Complete, Direct-Drive Generators, as You have described, "might" do what You want,
but the Regulation and Filtering Circuitry is going to take-up more space than you'd like.

Honda makes some very nice, and quiet, Generators.
.
.
.
 

Thread Starter

teliocide

Joined Sep 26, 2013
112
Your numbers are correct but as I was assuming (being a battery charging alternator) the operation voltage would be more like 13.8 volts dc . Thus 250 watts is arithmetically achievable.
The actual capacity required to run a small Waeco fridge (50 watts) and a few LED lights would rarely exceed 100 watts.
I have not been able to find the RPM vs Volts for the Alternator, if not I will measure it.
The GX 25 max output (746 watts) is around 7000 rpm but it will rev out to more than 8000 rpm
At 3000 rpm the output is around 350 watt. Which is around what most 4 cycle generators run at.
It will not be direct drive but with a v-belt at the "correct"ratio.
The numbers appear correct enough to produce the meager amount of power required.
Yes, I do understand that real world efficiencies must be considered. What motor power actually goes into the Alternator and how much of this power is lost in the Alternator.

I have 5 Honda inverter generators I scored from the company I worked with several years ago.
Motors purred beautifully but all of them have dead inverter boards. They live under cm of dust these days.
The motors are going to the local Scouts for Go-karts I believe, according to the Mrs.

To fully understand the limitations of a project, sometimes a prototype must be built from which the required corrections become obvious.
 

LowQCab

Joined Nov 6, 2012
2,170
Loses, Loses, Loses, they are everywhere.
A "V-Belt" Drive is particularly "lossy", a Rubber-Cushioned Direct-Drive would be much more efficient.

If you're dead set on seeing how much You can get out of this setup,
I would recommend a Buck-Switching-Controller with an external FET to handle the Current.
It's not going to be simple.
You will need a very low-loss Switched-FET-Bridge-Rectifier,
a big stack of High-Current Bulk-Storage-Capacitors,
a Voltage-Regulator for the Controller,
a Heat-Sink for the FET,
a custom High-Current-Inductor,
another big stack of High-Current Bulk-Storage-Capacitors.

This Circuitry might as well directly supply 120-Volts-AC with a large "Push-Pull" Toroidal-Transformer
because every time You change Voltages, there will be additional losses.
Do You really need ~12-Volts-DC ?
.
.
.
 

Thread Starter

teliocide

Joined Sep 26, 2013
112
Unfortunately everything is 12V as everything is mounted in a vehicle.
I am puzzled as why you feel so much smoothing is required given All runs well directly from the Vans 12 volt system.
There are different interpretations of efficiency, mine is far more tailored to cost than absolute perfection. Not necessarily by choice.
Gates suggest V-belts are 95% and Serpentine Belts 98% and synchronous belts 99% efficient.
Not really that terrible.
 

crutschow

Joined Mar 14, 2008
30,102
A permanent magnet generator can put out a very high open-circuit voltage, thus they are difficult to control with a series regulator.
That's why motorcycles with permanent magnet generators typically use a lossy shunt regulator, which simply shunts the excess voltage/current to ground (and I know from experience that they get really hot when doing so).
 

LowQCab

Joined Nov 6, 2012
2,170
The required Electronics to make this work will not be "cheap".
Spend your Money on a bigger Engine that's more likely to last for more than ~50-hours,
then go to a Junkyard and pull an Alternator of of a GM Truck,
no complex Custom-Regulator required.
A 5-Horsepower Engine will provide somewhere around ~100-Amps while maintaining ~14-Volts.
The most difficult part of this plan is installing an ~8" Serpentine-Belt-Pulley on the Engine.
.
.
.
 

Ian0

Joined Aug 7, 2020
5,504
Buy an LPG fridge. It might not be as efficient as a compressor fridge, but driven by a 20% efficient generator it will be!
 

Thread Starter

teliocide

Joined Sep 26, 2013
112
A permanent magnet generator can put out a very high open-circuit voltage, thus they are difficult to control with a series regulator.
That's why motorcycles with permanent magnet generators typically use a lossy shunt regulator, which simply shunts the excess voltage/current to ground (and I know from experience that they get really hot when doing so).

Back in the 1970s, on bikes like a Suzuki TS 250, we made up shunt regulators to eliminate the battery.
Just a zener diode and a 2n3055 bolted to a chunk of aluminium. Any volts above the desired voltage was turned into heat.
 

Thread Starter

teliocide

Joined Sep 26, 2013
112
The required Electronics to make this work will not be "cheap".
Spend your Money on a bigger Engine that's more likely to last for more than ~50-hours,
then go to a Junkyard and pull an Alternator of of a GM Truck,
no complex Custom-Regulator required.
A 5-Horsepower Engine will provide somewhere around ~100-Amps while maintaining ~14-Volts.
The most difficult part of this plan is installing an ~8" Serpentine-Belt-Pulley on the Engine.
.
.
.
The primary objective of this project was Small and light.
It seems to me very silly to drag around a 2KW genset for 100 watts.
 

Ian0

Joined Aug 7, 2020
5,504
Back in the 1970s, on bikes like a Suzuki TS 250, we made up shunt regulators to eliminate the battery.
Just a zener diode and a 2n3055 bolted to a chunk of aluminium. Any volts above the desired voltage was turned into heat.
I’ve seen a stud-mounted power zener (on a Lambretta) - even simpler!
 
PM generator circuits are usually a bridge rectifer comprised of two diodes and two SCRs, for the regulator. You'll lose about 3 volts in it. This is a problem for 6V mopeds with lighting going too dim at idle.
Attached circuits are old and still used today even in ultralight airplanes. These are low cost and could be improved in many ways.

OP I don't see any extra cooling with your generator, no fan so it might not make rated output for a long time.

Kubota_Schematic.jpg150cc GY6 voltage regulator schematic.jpgRotax-Ducati_Rectifier-Regulator.jpg
 

Dodgydave

Joined Jun 22, 2012
10,361
Most of these alternators use a shunt method of regulating the voltage by shorting out the alternator output, here is one for a 3 phase type it can be used for single phase as well, ..

Regulator2BRectifier.png
 
I looks like the same 2 diode-drop losses, but I hate shunt regulators basically shorting out the generator is a waste of energy - you get a hot generator, more fuel used - back to 1970 lol.
 

Ian0

Joined Aug 7, 2020
5,504
Look at bridgeless power factor correction circuits. You should be able to adapt the concept to produce a DC bus voltage higher than the peak voltage of the generator, with no diode losses. You could probably use the series inductance of the alternator as the inductor in the boost circuit.
 

LowQCab

Joined Nov 6, 2012
2,170
The Alternator in question will make an excellent ~10-Amp Battery-Charger,
that is precisely what it was designed to be, and to do.

It's not designed for continuous Maximum-Output for hours on end,
and neither is his proposed Weed-Eater-Engine,
( it's a very nice, high-quality, Weed-Eater-Engine,
but it's still a Weed-Eater-Engine,
and it won't last very long doing Generator-Duty ),
( Check the Oil-Level at least every hour, or else ......... ).
.
.
.
 

crutschow

Joined Mar 14, 2008
30,102
As a shunt-regulator example, below is the LTspice simulation of the second circuit by prairiemystic in post #13, with a few modifications to get it working with the part models I have.
I also substituted the low-cost TL431 adjustable voltage regulator in place of the Zener to get more accurate regulation.
The output voltage is adjusted by selecting a value for R2 that gives 2.5V at the TL431 control input for the output voltage you want..

I assumed some alternator characteristics for the simulation, but need the actual alternator winding resistance to properly design the circuit.
Do you have any specs for that alternator other than its voltage and current?

1656714517775.png
 
Last edited:
Top