# coupling a 3phase alternator to a Diesel car engine

Status
Not open for further replies.

#### Hero666

##### New Member
I would like to couple a 3phase alternator to an old working Diesel car engine to be used as standby power supply for my house.
Would any kind soul guide me designing a Governor (be it mechanical or electromechanical?). I like to have the simplest possible design.
Thanx in advance to all who are willing to help.

#### spec

##### Well-Known Member
Hi Hero,

The motor control system that you want should be quite simple to do electronically (famous last words). Off the top of my head, here is a suggested approach:

The main components you will need,
(1) DC power source, a stabilized 12V would be ideal
(2) Slotted opto sensor
(3) OPA2192 dual advanced operational amplifier
(4) TIP41C (NPN) or TIP42C (PNP) medium power transistor
(5) Solenoid powerful enough to operate the engine throttle
(6) Potentiometer

Circuit function
(1) A disc with a hole in it rotates at engine/generator speed
(2) The slotted optocoupler is placed so that the disc rotates in the optocoupler slot. The optocoupler thus produces a pulse for every revolution of the engine/generator
(3) One of the opamps is configured as a pulse counting integrator that inputs pulses from the slotted opto sensor and outputs a DC voltage proportional to engine/generator RPM
(4) The second opamp is configured as a precision amplifier that compares the Voltage output from the pulse counting integrator and a voltage that you set with the potentiometer.
(5) The output from the precision amplifier connects to the base of a medium power transistor that drives the solenoid, which in turn controls the engine throttle.

In this way the engine/generator RPM will be stabilized.

The above is a simple approach, which you asked for, but there are much more sophisticated approaches. The Rolls Royce approach would be to use an Xtal controlled frequency reference and a phase sensitive detector. While this may sound complex. in practice it would not necessarily be.

Another Rolls Royce approach would be to use an Arduino controller to sense the pulses from the opto sensor and drive a stepper motor that controlled the engine throttle. This would be, by far, the simplest and cheapest hardware implementation, but a bit of simple program code would be required. This is pretty much how automobile engine tick-over RPM is stabilized.

The above are just some thoughts.

spec

Last edited:

#### JimB

##### Super Moderator
A thought!

If you used an engine from a car which had a cruise control feature, you could use that to set the engine speed to the required 3000 RPM (or whatever).

JimB

#### crutschow

##### Well-Known Member
The output from the precision amplifier connects to the base of a medium power transistor that drives the solenoid, which in turn controls the engine throttle.
I'm confused.
Are you thinking of some type of linear solenoid, since most solenoids are basically non-linear on/off?

#### crutschow

##### Well-Known Member
If you used an engine from a car which had a cruise control feature, you could use that to set the engine speed to the required 3000 RPM (or whatever).
Or get a stand-alone cruise control from an older junk car (I think the newer types use the engine computer which wouldn't work for this).
But they can't use one that uses vacuum for the throttle control, since a diesel has no appreciable manifold vacuum, it would need to have an electric throttle actuator.

Edit: One difficulty. Cruise controls are designed for the vehicle to be brought to the desired speed and then set.
This application requires the speed to be preset so the actuator control would need to be modified or have added circuitry.

Last edited:

#### spec

##### Well-Known Member
I'm confused.
Are you thinking of some type of linear solenoid, since most solenoids are basically non-linear on/off?
Good point Cruts- I should have explained better.

I had in mind a solenoid attached to the throttle with a spring pulling the throttle to the tick-over stop. The servo loop then puts sufficient current, at any one time, into the solenoid to open the throttle enough to maintain the correct engine RPM.

spec

#### crutschow

##### Well-Known Member
I had in mind a solenoid attached to the throttle with a spring pulling the throttle to the tick-over stop. The servo loop then puts sufficient current, at any one time, into the solenoid to open the throttle enough to maintain the correct engine RPM.
Okay.
But my point was that standard solenoids have a very non-linear force vs. current relationship (the force increases significantly as the solenoid gets near the closed position) so I think a standard solenoid won't work for that. .
You would need one designed to give a linear force versus current, independent of position, i.e. a linear servo.

#### crutschow

##### Well-Known Member
I would like to couple a 3phase alternator to an old working Diesel car engine
What type of 3-phase alternator?
If you are thinking of a automotive type alternator, those won't work because of the AC high frequency the output.
If not, why do you want 3-phase? Do you have some 3-phase loads?

#### spec

##### Well-Known Member
Okay.
But my point was that standard solenoids have a very non-linear force vs. current relationship (the force increases significantly as the solenoid gets near the closed position) so I think a standard solenoid won't work for that. .
You would need one designed to give a linear force versus current, independent of position, i.e. a linear servo.
The solenoid is inside the servo loop, so its transfer function is not fundamentally important. But even so, a solenoid with a reasonably constant force for about a 20mm travel would not be difficult, especially at the very low force required to operate an engine throttle. And as you imply, it would be no big deal to make a suitable solenoid.

But if push comes to shove, just use a little motor with a rack and pinion, or similar.

The method of moving the throttle it is not a show stopper.

spec

#### debe

##### Active Member
Most small engine govenors these days on generators use a stepper motor & small gearbox to adjust the speed on petrol & diesel engines.

#### tcmtech

##### Banned
The simplest to work with would be to get a self governing injection pump like most heavy duty/ industrial application diesel engines use. Odds are your vehicle engines injection pump is simply one that was assembled without a governor mechanism being that how most vehicular application injection pumps are set up.

Without knowing what engine you have and what injection pump system it uses we are just guessing here.

#### crutschow

##### Well-Known Member
The solenoid is inside the servo loop, so its transfer function is not fundamentally important.
It is if it's very non-linear since that can make it difficult to stabilize the loop.
Perhaps if the solenoid only has to move over a portion of its range, it may be linear enough to not be a problem.

But as you say, there are other alternatives, such as a stepper motor, that may work better.

Last edited:

#### Les Jones

##### Well-Known Member
A model servo would probably be powerful enough to move the throttle butterfly. It would probably be better to modify it's internal electronic to either just drive the motor or so the position was controlled by voltage rather than pulse width modulation. I have used one of these model servos to control a butterfly valve that controls the airflow to my metal casting furnace. (Pictures here.) Many years ago I built a generator with a 5KW alternator coupled to a Fiat 500 petrol engine. The speed sensing on that was just a frequency to voltage converter from the ignition pulses. The voltage from that was compared with the speed adjustment voltage from a potentiometer using an op amp and the output amplified to drive a small geared motor. I had quite a lot of trouble getting the control loop to be stable. I think now using a PID algorithm would be a better approach.

Les.

#### tcmtech

##### Banned
A model servo would probably be powerful enough to move the throttle butterfly.
Pretty sure diesels don't have one.

#### Les Jones

##### Well-Known Member
Hi tcmtech,
I have just had a look on the internet and you are correct. It had not occurred to me that if the air flow was restricted the air would not reach a high enough pressure and would therfore not reach a high enough temperature due to adiabatic heating to ignite the fuel. If the engine was not very old then I would imagine speed control would be an electrical signal to the engine control unit. If it was before the time of electronic control I would imagine there would be some kind of mechanical control of the injector pump.

Les.

#### tcmtech

##### Banned
If it was me I would either get a old belt drive mechanical governor (Hoof or Pierce brand governor units) off an old industrial engine from a farm and heavy equipment scrap yard ($25 -$75) or a off the shelf electronic control unit and throttle actuator ($75 -$100 from eBay or Amazon) and use that.

That is assuming the injection pump that the engine uses now doesn't have a built in governor or the ability for one to be added or a whole different injection pump unit with the proper governor can't be had for similar costs.

#### dr pepper

##### Well-Known Member
If you are using an older mechanical injected engine like a pug xud9 or citroen dw8 then depending on what actual rpm you want, there is a maximum rpm governor built into the pump, you might be able to adjust it down a little, they are usually set to about 5 krpm.

#### spec

##### Well-Known Member
It is if it's very non-linear since that can make it difficult to stabilize the loop.
Perhaps if the solenoid only has to move over a portion of its range, it may be linear enough to not be a problem.

But as you say, there are other alternatives, such as a stepper motor, that my work better.
Hmm, I have had a look at some off-the-shelf solenoids, and like you say their force, varies greatly with extension.

spec

#### tomizett

##### Active Member
a pug xud9 or citroen dw8
You're not an LDV owner by any chance?

<edit>
On a more on-topic note, is it also necessary to regulate the voltage output by controlling the field current, or will this take care of itself if the RPM is correctly regulated?

Status
Not open for further replies.