Sully's Micro Cogen Project - Needs critique and input from you experts.

[artificer]

Concerning the throttle control, I haven't seen one yet that acts like the honda's eco mode.  I plan on building on into my system that I posted about.  Measure the current going to the inverter, and choose a throttle position.  I plan on using an Arduino micro controller to do the job.  It has built in PWM output, so I can use a simple solenoid, and get variable positions.  Add a rpm pickup, and you have feedback, but the feedback isn't required.

Now to get the time to build it, and program the controller.

Another option is to use a PWM chip (part numbers are at work), to control the solenoid.  The chip I've looked at will output from 10% to 90%, and is controlled with a voltage.  Use a current to voltage converter circuit, feed the voltage into the chip, and let the chip power the solenoid.  More current gives more throttle.  Set the gain, add some filtering, and its an analog control circuit.

Yet another option is to get a DC linear actuator.  Measure the current with a shunt.  Use that voltage, and compare it to a voltage produced by a potentiometer connected to the linear actuator (position sensor), and use a simple comparitor.  Feed the comparitor output to the motor drive/relay.  If the current sensing voltage is higher than the position voltage, the motor turns on, and adds throttle.  If its lower, it reverses it.  Make sure you have a dead band in the middle to prevent hunting, and it should work.

Hopefully someone can pipe in with something similar thats off the shelf already.

Michael

Edit:  take a look at the Woodward 828 Digital Control.  Analog inputs, so you could use a current sensing transformer with a resistor as an input.  Probably expensive, but they're made for marine use.

[EBI-WPO]

Artificer,
Most available solenoids don't react linearly to variable input. They are unpredictable in other than end positions. Linear actuators/motors and solenoids are different animals. You would also need some type of interface driver (such as FET) between your PWM device and any coil driven "work" device to handle the current demands of that device.

Terry

[BruceM]

I think Arificer's on the right track with his preferred approach.  With an Arduino or Pic or Picaxe you could cope fairly easily with the extreme non-linearity of the solenoid, and also the non-linearity of the throttle. By using a table driven compensation, you can just take a WAG at the initial values, then "tune" it experimentally with your debugger software.  

As Arificer suggests, this could be done in analog circuitry, too, but with more development time and circuitry.  Creating non-linear curves (multiple linear approximations, actually) can be done readily with diodes and zener diodes in an op amp amplifier feedback loop. With a wireless breadboard wired into the running setup and a handful of zeiners of different values and some trim pots for setting the gain of each zeiner's feedback, you COULD do it. (You'd want to opto-isolate the Mosfet drive circuit.) This is a valid approach for someone very handy with op amps for linear circuits and a strong disinclination to software.  

It was enjoyable for me to read your the gamut of ways to "skin this cat", Artificer.  Bravo.
Best Wishes,
Bruce

[artificer]

If I'm not mistaken most of the Woodward solenoids are capable of position control, not just on/off.  Here's a video of a listeroid controlled by a Woodward electronic governor.  Sully could use a similar solenoid and DIY circuit, or if the budget allowed, just use the production/off the shelf controller.  The controller has speed feedback, so the closed loop control takes care of all of the unpredicatbility.

One way to get around the problems of PWM and unpredictability is to use several solenoids in full stop mode to get several run speeds.  None energized is low rpm.  Energize another and its high idle.  Another gets you mid-range, and another gets full speed.  You can use off the shelf components like current sensing relays to control the setup.  Use as many solenoids as you want set points.  The solenoids will be cheaper, since you only have two positions, so that should counter the cost of needing more.  I'm not sure if this is better than just going with a heavy duty servo control.  Do they make marine duty servos that are affordable?  I'd be a bit hesitant to use hobby servos on a commercial vessel.  At least with the DIY circuitry, you could pot the unit to keep the nasty environment out.

The chip I mentioned (I'll look it up when I get to work) can directly drive up to 2.5 amps.  I have no clue how much the Woodward solenoids draw, so not sure if they are adequate.

Thinking about it this morning, I'm not sure how appropriate a solenoid control system would be.  If it was part of the Woodward system, and its designed in, thats one thing.  But the engines will already have governors, so you don't need high speed control.  You just want to change the rpm setting based on the current draw.  Slower is better in a way, since you'll moderate the speed changes due to transients.

I'm interested in this topic because its central to the power unit I'm designing.  For me, since I'll be using the Arduino controller, PWM to the solenoid only requires a driver and the solenoid.  If the solenoid ends up being too expensive, I'll go with a large RC servo.  The Arduino has a library of servo functions, and can drive them directly.  I guess its time to research hobby priced servos, and see if they can handle hundreds of hours of use, if not thousands.

Michael

[BruceM]

Nice setup with the Woodward Governor.  Thanks for that link, Micheal.  I'm interested in your take on the servos out there.  

I'm interested in homebrew electronic governors as I've long thought about getting rid of my barely adequate (Listeroid) stock governor and replacing it with a servo and my own electronics with rpm sensing from the existing flywheel spoke sensor. This would  facilitate a 80% propane mixing scheme by second servo as well.  

There are serious safety issues that have to be addressed for my scheme.  My pneumatic rack closer actuator could overpower even a jammed servo, and of course there's the pneumatic valve lifter if even that failed. (I already have microcontroller monitoring of rpm and auto shut down if out of range.)  I would use a solenoid on the propane line to failsafe that servo metering.

For Sully's situation, a salt water marine environment adds another whole layer of design concerns that I'm largely clueless about.

[Lloyd]

Here you go off the shelf Linear Actuators and PWM Controllers. http://www.precisiongovernors.com/applications/index.cfm?application=13&type=actuator I poste about these in my Perky/cat\Volvo thread.

quote:SZ Linear Actuators

The model SZ linear actuator is designed for direct mounting on the engine in place of the fuel shutoff solenoid.  This integral type actuator is linkage free and provides precise speed control.  The Linear Ball-Bearing design ensures smooth rapid operation of the actuator providing a stable operating speed of the engine when combined with one of PG's controllers.  This combined with low operating currents of less than 1.5 amps result in an actuator with consistent performance over a long operating life. PG's linear actuators are designed to respond in milliseconds to changes in the current from a pulse-width modulated signal generated by a PG controller.  A magnetic pick-up signal transmitted to the controller provides for closed loop speed control.  The actuator and controller maintain engine speed adjusting the fuel requested to obtain preset engine speed settings.  This eliminates the droop inherent to mechanical governors.

Lloyd

[artificer]

BruceM: I'm not knowlegable enough with servos to give any advice.  About the only think I know is to make sure it has enough torque for the job, and feed it a proper pulse width for position control.  They have some really cool RC items out there.  Since they're for toys, the prices aren't bad.  Under $100 for a servo with 12lbs of pull?  Not bad.  I'm interested in the Arduino system because of the shields/daughter boards they have.  Since its a hobby level controller, its cheap.
Arduin board $30
Servo  $50
Proximity sensor for RPM  $20  (automationdirect.com)

One safety item you might want to look at is the anti-shudder intake valve on VW TDI's.  Block off the egr opening (or use it for propane injection), and find a non-vacuum method to actuate it.  On the TDI's, if the key is off, the butterfly is closed.  No air, no running, and the seal looks fairly tight.  Kinda expensive new, but you could probably find one used.  (why do I keep sounding like a cheap SOB?  )

Lloyd:  SWEET!  Thats exactly the type of unit I was thinking about.  Made for the job, quick acting, and with the controller its an easy solution.  I've been trying to read the entire forum... but its taking awhile.  Apparently I haven't gotten to you thread yet.  Thanks

Michael

[BruceM]

Nice, very high quality linear pulse actuators, Llyod.  Just what Sully needs for his marine application.

My Listeroid IP rack requires very little force to move if the governor linkage is removed- less than a pound.  That's why I was thinking "disposable" RC servos. (Replace every ?1000 hrs?) I think a decent servo with ball bearings on the output shaft is under $30. at TowerHobbies.com.  The current requirements for the pulsed actuators, plus the EMI, is too much for my application.

[EBI-WPO]

BruceM,
I don't have any problem with the circuitry involved, in fact am very well versed on it, but a solenoid is not a servo. A slug of iron in a coil is not variable with any repeatable results. Servos, linear motors, motors with gear train, etc will respond like he describes, but not many plain old solenoids. A typical solenoid takes x current to move from one position to the other (end to end), but can remain there with a 1/10 of that amount of current, with quite a load against it.
Otherwise his description of simple servo control is text book.

The Woodward ACTUATOR in the video is NOT a simple solenoid. (it appears to be a Dyna 2000 or 2500)

Seperating apples from oranges, thats all,
Terry

[BruceM]

Thanks for sorting that out, Terry.  I've seen pulsed solenoids used in a few linear applications but didn't realize there were special designs for doing so and I've never used one. 

How are the solenoids designed for pulsed linear use different than the typical coil and slug? Dual air coils???

 

[artificer]

+1 Thanks, Terry

In my mind I was thinking of the higher end solenoids, not the type you find on the washing machine door.  I think that if you use a return spring of significant force (1/4 of total pull?), you can get in the ball park with just applying a PWM signal.  If all you want is idle/solenoid off, fast idle, faster, and wide open, then it not a terribly precise requirement.  Getting a unit designed for the purpose, however, is the right way to go.

Michael

[BruceM]

I did a little research online and while it is possible to use a conventional solenoid as a linear actuator moving against a spring,  the PWM frequencies must be very low to avoid inductive heating. One would have to experiment to insure acceptable heat levels.   Position will not be linear with respect to PWM duty cycle, as the magnetic force increases as the slug becomes fully enveloped in the coil. Other issues for a dynamic/linear use of a solenoid would be wear/durability, since constant movement was not part of the original design requirements. Many of the PWM/linear solenoids have linear bearings and such to address that issue. 

Older designs of solenoid valves designed for PWM use are often limited to a 30Hz maximum PWM frequency.  The modern "linear" designs (used in some newer electronic controlled transmissions) are compatible with high frequency (300KHz) drives, with much smaller inductances.

[thomasonw]

On throttle control, I am looking to reuse an automobile 'Cruise Control' throttle unit in my DC generator control project.  Small DC actuator connected to a pull cable (needs external return spring) and built in limit switches.  Plus they have done some level of 'weather protection'.  Got mine from a junk yard for under $20.

Will still need to close the loop by monitoring RPMs and deciding needed RPM vs. load being placed on the motor, but for the actual throttle control I have high hopes for these auto throttle cruise control units.

-al-

[Westcliffe01]

If you read US Patent 7209020 it explains in detail how the latest "Variable force solenoids" work.  The ones that adjust the hydraulic pressure in your transmission or the phase angle of your camshaft if you have that kind of engine.  They have eliminated the need for position feedback, but at the same time still have to use a certain strategy when operating the solenoid to get repeatable results because of the hysteresis of the materials in the magnetic circuit.

For a boat, which has to function as a self sufficient "island" (especially with paying guests) one needs reliable solutions, not disposable parts...   Generally, getting access to work on things is hardly a comfortable job. Logistics can be very troubling indeed and make the actual cost of a part seem like a secondary consideration.  Ideally, one would try to minimize reliance on belts of any kind, lest you find yourself changing it during a force 10 gale...  The other thing that comes to mind is redundancy.   The recent misery on the cruise ship is a reminder why one does not want single points of failure.

If you need custom 3 phase PM motors, you should look at the "kit motors" sold by Parker  http://divapps.parker.com/divapps/emn/pdf/Frameless_Motor_Brochure.pdf  The power to weight/size ratio of these motors is superb.  Yes, you need to put them in your own housing and supply a bushing and shaft, but most of that is straight forward lathe work.  Putting the rotor into the stator needs a fixture since the magnetic forces involved are very high.  But besides that hassle, these are very competent motors with good speed potential or high torque in the bigger diameters and one can "stack" these cores on a single shaft to multiply torque if needed.

[Lloyd]

If you read US Patent 7209020 it explains in detail how the latest "Variable force solenoids" work.  The ones that adjust the hydraulic pressure in your transmission or the phase angle of your camshaft if you have that kind of engine.  They have eliminated the need for position feedback, but at the same time still have to use a certain strategy when operating the solenoid to get repeatable results because of the hysteresis of the materials in the magnetic circuit.

For a boat, which has to function as a self sufficient "island" (especially with paying guests) one needs reliable solutions, not disposable parts...   Generally, getting access to work on things is hardly a comfortable job. Logistics can be very troubling indeed and make the actual cost of a part seem like a secondary consideration.  Ideally, one would try to minimize reliance on belts of any kind, lest you find yourself changing it during a force 10 gale...  The other thing that comes to mind is redundancy.   The recent misery on the cruise ship is a reminder why one does not want single points of failure.

If you need custom 3 phase PM motors, you should look at the "kit motors" sold by Parker  http://divapps.parker.com/divapps/emn/pdf/Frameless_Motor_Brochure.pdf  The power to weight/size ratio of these motors is superb.  Yes, you need to put them in your own housing and supply a bushing and shaft, but most of that is straight forward lathe work.  Putting the rotor into the stator needs a fixture since the magnetic forces involved are very high.  But besides that hassle, these are very competent motors with good speed potential or high torque in the bigger diameters and one can "stack" these cores on a single shaft to multiply torque if needed.

Look at Precision Governors, the link I posted are the same . They have a retrofit, or a specific fit.

All of them can be retrofitted with a little tool and die.. Much easier then building housings. Plus they are cheep considering they're off the shelf

Lloyd