Wanted help with electronic generator govenor

[RobSmith]

Hi All,

I am building a wood-gas powered off grid generator set up.
I will want the little car engine I am using to be self governing.
The ignition is currently via contact points and coil but I may soon transitorise this with a kit I built years ago. That uses a Hall effect sensor and I could pickup that as a triggering signal.

My plan is to take this signal (which would be 50hz at 1500rpm or 33.33hz at 1000rpm) and use it to then drive a small servo to open and close the throttle a small amount. That might possibly be a radio controlled car servo.

My thought is that I might use a Frequency to voltage IC to create a voltage related to speed. I was then thinking I might use a 555 PWM circuit to control the servo using the voltage related to speed to control the pulse width.

Has anyone done this or a similar thing and does anyone have a circuit diagram or advice on the best components to use and how I might achieve this?

Rob

 

[RobSmith]

Hi all,

I have been hunting around for variations of the 555 pwm circuit I was hoping to use to control the engine throttle.

The best one I can find is this one... http://sarconastic.tripod.com/servodriver.html

The on/off timings are controlled by the resistances above and below the discharge connection on the 555 timer.

My original plan was to use a frequency to voltage converter to alter the pwm on to off ratio but I don't think I can do that now. Can anyone suggest a better way to create and control the pwm signal?

Rob

 

[duke37]

This looks like a difficult project.

One thought, if you start a timer from the ignition with a pulse length of 20ms, then if the next ignition pulse comes late, the later it is, the more the throttle needs to open. Using set/reset ICs would give a PWM directly. I am sure someone could give a computer solution to this.

The 4046 phase lock loop may be of use.

Stability of frequency and overshoot will need consideration.

 

[RobSmith]

Hi,

I am not using it like that. The ignition will be just using a fixed timing every 180 degrees. That will be timed by me physically bolting the trigger wheel in the correct place. The 180 degree timing will be achieved by having two equally spaced trigger points.

It is the engine speed that will be controlled by throttle control..
The PWM setup is to produce a 20ms repeating signal. On that signal is a square wave On/Off. That has a variable pulse width between 1 and 2 ms.
On the throttle on the engine I will have a radio controlled car servo. The servo just uses this signal. A 1.5ms On/off each 20ms will be sensed by the servo circuitry to adjust to a central position. An On/Off of 1 or 2 ms per 20ms will get the servo to drive to either extreme.
I have been thinking about the circuit and I think if I used two 555 timers that would be better and simpler to implement.

One 555 timer would produce a trigger timing every 20ms. That would trigger the second 555 timer to start its On/Off output timing. I can adjust that output timing easier on just the one timer than using the circuit I found on the web that just uses the one 555 timer for both the 20ms timing and the pulse width too.

I will make my scribbles clearer and post a diagram when I have figured something out.

Rob

 

[VAT]

This sounds very interesting... would you be able to put the mechanics and the specs of what you want onto the diagram as well?

 

[debe]

You could use a vehicle cruise control, they will run of a hall switch quite happily.

 

[KrisBlueNZ]

Your frequency-to-voltage converter idea sounds like a good start.

You will have a desired RPM for your motor, and this will produce a corresponding desired voltage at the output of the F-V converter.

Then you need to generate a control signal to the servo, and you need to "close the loop" by controlling the servo driver from the F-V converter output. Or actually, from a "difference" signal that you get by comparing the F-V converter output voltage against the desired voltage, which is known as a "setpoint". This may be fixed, controlled by a trimpot (set by a screwdriver), or controlled by a potentiometer with a knob, or controlled by a signal from another source.

Your description of the control signal needed by the servo does not seem clear to me. Can you give the model number of the servo controller and/or a link to the data sheet?

There are various ways of closing this control loop. A common one is PID (proportional integral derivative) - see http://en.wikipedia.org/wiki/PID_controller. The control loop has to be "tuned" so that it gives an appropriate amount of correction at an appropriate rate to properly handle the delay between changing the throttle position and seeing a corresponding change in the engine speed, so that the control loop is stable. If it's not stable, it may take an unnecessarily long time to reach the right speed ("overdamped"), or it may "hunt", i.e. oscillate above and below the right speed ("underdamped"). Some experimentation will probably be needed here, and the loop control parameters may need to change depending on the actual target speed. This is about the limit of my knowledge and experience on this subject.

 

[duke37]

The servo data that I saw says that the servo takes in a series of pulses of different width to determine the servo action. This seems to be a complicated method which is used so that another seven servos can be used on the same channel.

The servo may just take the average voltage as its signal, if so, it could be directly connected to the PID circuit.

 

[RobSmith]

Hi,

I have been figuring things out a little.

I have made an adapter up to fit the trigger wheel onto a little lawn mower engine. I now have an AC signal to use to trigger things. It is a 60 tooth trigger so at 1500rpm produces a 1500hz signal.

I am using an LM2907N as the frequency to voltage controller.
See http://docs-europe.electrocomponents.com/webdocs/0b13/0900766b80b13a41.pdf

Looking at the data sheet showing the maximum frequency (Page 7) I think I will need to make a different trigger wheel with 15 teeth to get the signal frequency down to 325 to 400 hz. That is fine as The current trigger wheel is not great anyway.

On page 16 there is shown a method to change the gain and minimum output voltage.
I want my minimum output to be 4v to feed into the control pin on a 555 timer.

Using the output calculations on the front page I have..

Vcc 10v
R1 61538 ohm
C1 0.02uF
Frequency 325 to 400 hz
Output 4v to 4.92v

The diagram on page 16 mentions the altering the gain but does not show how to calculate this. I would like to change the gain so at 400 hz the output is 6v but cant't see how you do this.



The other problem I have to solve is trying to figure out how long the pulse width would be on a 555 timer in a monostable setup when the control voltage on pin 5 is varied.
I shall be feeding this 555 timer with a trigger signal at 50Hz.
I want it to produce a 1ms high output with the control voltage at 4v and a 2ms high output with the control voltage set to 6v. I might have that reversed but it wouldn't mattter which way it works I would just mount the servo the other way round. I can't find any calculations for this too.


Any help or pointers would be greatly appreciated.

Rob

 

[duke37]

I do not have time to go through all the details but,

1. Vo=Vcc*fin*C1*R1*K . K is given in the data and is approximately 1.
400Hz 6V then C1*R1 = .0015
If C1= 100nF then R1= 15k
or C1=10nF, R1=150k

2. Fmax=I2/(C*Vcc)

From equation 2, chose a value of C1 to get the frequency coverage you want. Drop C1 to raise the maximum frequency. I do not know what the absolute maximum is but 100kHz should be possible, well above what you want.

 

[RobSmith]

I do not have time to go through all the details but,

1. Vo=Vcc*fin*C1*R1*K . K is given in the data and is approximately 1.
400Hz 6V then C1*R1 = .0015
If C1= 100nF then R1= 15k
or C1=10nF, R1=150k

2. Fmax=I2/(C*Vcc)

From equation 2, chose a value of C1 to get the frequency coverage you want. Drop C1 to raise the maximum frequency. I do not know what the absolute maximum is but 100kHz should be possible, well above what you want.

Thanks for the reply. I went over my spreadsheet and found I had made a cockup in one of the equations.

I am going for: C1=6.8nF and R1 = 45250 ohm.
That gives me a 4v output at 1300rpm (also 1300hz).

Looking at the graphs I think for Vcc = 10v the value of I2 = 170uA.
Using that the equation Fmax=I2/(C*Vcc) = 2500 which is plenty. If I choose 10nF for C1 then Fmax comes out at 1700Hz which is maybe too close to my working range which might get close to that.

This is the crank trigger. I have added two more teeth to it. It had 58 and a gap so the engine ECU could work out the crank's position. I am just wanting a continual signal so splodged a big dollop of weld in the gap and then machined and filed in the two 'missing' teeth. It now has a full complement of 60 teeth.

Rob

 

[RobSmith]

Your frequency-to-voltage converter idea sounds like a good start.

You will have a desired RPM for your motor, and this will produce a corresponding desired voltage at the output of the F-V converter.

Then you need to generate a control signal to the servo, and you need to "close the loop" by controlling the servo driver from the F-V converter output. Or actually, from a "difference" signal that you get by comparing the F-V converter output voltage against the desired voltage, which is known as a "setpoint". This may be fixed, controlled by a trimpot (set by a screwdriver), or controlled by a potentiometer with a knob, or controlled by a signal from another source.

Your description of the control signal needed by the servo does not seem clear to me. Can you give the model number of the servo controller and/or a link to the data sheet?

There are various ways of closing this control loop. A common one is PID (proportional integral derivative) - see http://en.wikipedia.org/wiki/PID_controller. The control loop has to be "tuned" so that it gives an appropriate amount of correction at an appropriate rate to properly handle the delay between changing the throttle position and seeing a corresponding change in the engine speed, so that the control loop is stable. If it's not stable, it may take an unnecessarily long time to reach the right speed ("overdamped"), or it may "hunt", i.e. oscillate above and below the right speed ("underdamped"). Some experimentation will probably be needed here, and the loop control parameters may need to change depending on the actual target speed. This is about the limit of my knowledge and experience on this subject.

Hi,
The Servo is a Futaba S3004.
It has a working voltage of 4.8v to 6v and a signal voltage of 3v to 5v.

The signal is a square wave pulse. Each pulse is delivered at 50 hz (every 20ms).
The width of the pulse is what the servo uses as its information it needs to know where the output shaft should be rotated to. A middle position is when the pulse is about 1.5ms. Rotation to one extreme is caused by narrowing the pulse to 1ms and rotation to the other extreme is caused by widening the pulse to 2ms.

My plan is to use one 555 timer to supply a constant 50hz 'falling edge' signal to a second 555 timer that will control the pulse width. I am wanting to control the pulse width with the voltage output from the frequency to voltage converter. I am struggling with the maths or information how to figure that out.

Rob

 

[duke37]

Nice job on the timing wheel.

As I suggested earlier, perhaps the servo just uses the average voltage, if so, you do not need any pulse generating circuit. Take a servo and slowly increase the input voltage and note the effect.

A DC motor could be used to open and close the throttle with its supply controlled from the DC signal.

You should have secondary speed measurement to kill the ignition if the speed goes over the top.

 

[RobSmith]

Nice job on the timing wheel.

As I suggested earlier, perhaps the servo just uses the average voltage, if so, you do not need any pulse generating circuit. Take a servo and slowly increase the input voltage and note the effect.

A DC motor could be used to open and close the throttle with its supply controlled from the DC signal.

You should have secondary speed measurement to kill the ignition if the speed goes over the top.

Hi,

The servo arrived in the post today so I wil try the voltage thing.

I will have a rev limiter. That will use a second frequency to voltage converter. In the data sheet they have given me the circuit diagram for that. It will cut the power to the ignition at about 1700rpm. The generator I will be driving should not be taken beyond 1800rpm.

Edit: I was also emailed the information about the servo. That email said it was all confidential etc so I will not post that.. It does say its input is a pwm signal as I describe but I will try the average voltage thing as it will not harm it.

Rob