The
Washing Machine Motor
Project
Ever since I saw one of the newer washing machine motors, I've wanted to have one to fiddle with.  
These are low speed, high torque three-phase motors and should be good for something other than washing clothes..
--------------------------------------------------
29 June 2014:
My boss (Wife) decided that she needed a new clothes washer and that was fine with me because I wanted to experiment with the permanent magnet direct drive motor from it.  After dismantling the washer, I got a bunch of bolts, hardware and bearings.  When the new washer was delivered, the guys who were going to pick up the old one got a kick out of the fact that I'd been cheating the junk man.  Here's the motor.

Permanent magnet motor, controller and shaft.
These motors are driven from a three phase variable frequency drive that is in the controller shown in the photo above.  I thought I could use the power driver chip on the controller to try to run the motor on 12 Volts to crank one of my small engines but it  looks like it requires a microcontroller.  I'm not conversant with writing code for microcontrollers so,unless I can figure out a way of driving the IGBTs in the power driver chip; the chip and data sheet will live to possibly see another day.  The rest of the controller is minus some components and is in the trash can.  The shaft and bearings are to support the rotor (magnet ring) and I saved them in case I needed them.  I removed the speed sensor (attached to the motor stator) and put it aside for the time being.

I think the motor would make a pretty good alternator because each of the three windings are about 16 Ohms resistance.  It won't produce a lot of current but should make enough current at 24 Volts to drive something else I cheated the junk man with.

Drivetrain from the scooter.
This little jewel is the driveline from one of those little three-wheeled personal electric scooters.  The motor and gears check out fine.  The motor is rated for 24 Volts, originally run by a couple of 12 Volt gel cells.  I may try to see if the alternator has enough guts to power the scooter motor.  If that works, I'm thinking of scrounging for a couple of rear mountain bike wheels and tires.  I can chain drive them from the drivetrain, getting enough reduction to make the speed compatible with the scooter.

I'm thinking of using Engine Number Four to direct drive the alternator because it idles down well and seems to make a reasonable amount of power.  Also, because it has the electronic governor, I can easily connect the speed control to the original foot feed from the Algore Edition Green
Hoyt-Clagwell tractor.  I've got some thinking to do on the method of controlling the drive motor but, depending on how the motor works as an alternator  it may be something like switching each of the three phases in and out to suit driving conditions, engine RPM and foot feed positon.  I'll get to this project after finishing The Upside Down Engine.
-----------------------------------------------------------------------------------------------
6 July 2014:
Since I need a test frame for the motor to discover its' characteristics as a variable frequency motor and as a DC generator, I made a frame using some scrap lumber.  The shaft is the original splined shaft from the washing machine.

Motor test frame.
I know it looks awful but it will do the job.  No sense in going to the trouble to mount it on Engine Number Four and find that it won't going to work.

Just before leaving the shop, I added a lead to the common connection to the three sets of coils. When applying 12 Volts between the common and, alternately, to the red, blue and yellow motor wires, the motor will cog the distance of one pole.  Since there are 36 poles (coils) in the stator, it will take 36 sequences of connecting the 12 volts to the red, blue and yellow wires.  If I want to make it motor at 100 RPM, I will need to switch the three poles at 3600 sequences per minute.  A circuit to do this shouldn't prove to be too big of a problem using a three stage phase shift oscillator working as a ring counter.  The approximate sinusoidal waveforms can be buffered up to drive three power transistors, each driving a winding.

I'll give that idea a shot tomorrow and let you know how it -turns- out.  Forgive the pun.  Sometimes, I can't help it!
-----------------------------------------------------------------------------------------------
7 July 2014:
I fiddled a bit with the motor.  First, I tried to use a very simple driver to make it motor but I will need a bridge driver to do that and, unless I can figure out a simple way of using the driver I.C. that came with the washer, I may have to abandon using it as a cranking motor.

Then, I decided to see how it would work as a DC generator.  Here's a drawing of the motor stator and a couple of ways to configure it for rectifiers.

Several ways of hooking up the stator.
The upper left drawing shows the simplified schematic of the stator as it came out of the washing machine.  Note that each of the three phase windings consists of 12 poles in series.  One end of each of the phases are connected together.

In the upper right drawing, I show the "Common" lead that gives a connection to the common ends of the coils.

In the lower left drawing, I show the simplest method of connecting diodes to half-wave rectify the AC voltage that comes out of the stator into DC.

In the lower right drawing, I show a typical 3-phase full-wave rectifier connection using six diodes.  There are differences in the two rectification methods.  In the simplest method, each winding is half-wave rectified.  All else being equal, the half-wave hookup gives a lower output voltage than the 3-phase full-wave rectifier.

I hooked up the test rig for half-wave and added a capacitor to the cathodes of the diodes.  When I gave the rotor a fairly easy spin, the voltage quicklu rose to over 100 Volts with no load.  I then hooked a 32 Volt, 2.4 Amp light bulb to the output.  It was a lot harder to spin and with some muscle applied, I couild light the bulb moderately bright.

Then, I hooked the output of the rectifier to the transaxle motor.  The motor is rated for 24 VDC and an Ohmmeter gave about 8 Ohms.  This means that the motor should draw about 3 Amps at full load (more or less).  When I gave the fixture a spin, the transaxle motor turned the wheels with only moderate force on the rotor.  This gives me hope that it will work as a generator for the transaxle.

Of the several things that need to be worked-out, controlling the Voltage to the load looks like it is the most critical.  The "generator" makes power at very low speeds.  I think that if I tried to drive it at, say, 500 RPM, the unloaded output would be several hundred Volts, much too high to work with.  Of course, under load, the Voltage would be lower but still much too high for the motor and a controller to drop that Voltage down to a safe voltage for the motor would require some pretty muscular transistors.

One way to make the "generator" output more closely match that of the motor, would be to re-connect the windings for each phase.  If the 12 poles of each phase were to be
reconnected series/parallel, the Voltage would be lowered and the current capacity would be raised.  In order to find out just what the re-configuration would have to be, I'd have to spin the rotor at a speed close to the engine speed and then measure the Voltage at a fixed load.  The re-connection possibilities would be to parallel all twelve coils in each phase to get minimum Voltage and maximum current.  The next highest Voltage connection would be to paralled six sets of two coils and series connect them.  Next highest would be four sets of three coils, series connected.  Next would be three sets of four coils, series connected.  Finally, two sets of six coils could be series connected.  Doing something like this would require some doping-out.

What I'd like to do is to get the generator to produce Voltage from about 100 RPM to about 600 RPM.  I could use a feedback loop to the engine governor speed control to reduce the engine speed when load decreases to a point at which the generator is producing about 30 Volts.

More head scratching is in order.
-----------------------------------------------------------------------------------------------
11 July 2014:
Today, I machined the original washing machine shaft for a wooden flat belt pulley I had sitting around and mounted the "generator" and a variable speed DC motor to the bench for testing as a generator.

Test setup.
I first hooked the "generator" up with only one of the phases hooked-up and no rectification.  With this triple single phase plan, I only need one of the phases for testing.  Adding identical phases will increase the current only. I then used a 150 Watt, 125 Volt light bulb as a load and spun it up.  If rectified, theoretically, the voltage will be 0.707 times the AC voltage.

Here's what I got:
                                           Calculated
    RPM   -   AC Voltage   -   DC Voltage   -   DC Current
      375              102                  72                  1.225A


I want to run the engine at about 600 RPM for full load operation so that means that the AC voltage as above will be around 163 Volts or about 115 VDC.Since I will be running a 24 Volt motor, I need to reduce the Voltage by four times.  This will give about 28.75 Volts out of the rectifier.  To reduce the Voltage, I will need to re-connect the 12 series connected each pole (phase) to three sets of four paralleled coils in series.  Of course, connecting the half wave rectified outputs of the other two phases will leave the Voltage the same but will increase the current by four times. This should work.

Then, I hooked up the "generator" as in the lower right hand schematic (full wave 3 phase rectifier).  Here's what I got:
     RPM   -   DC Voltage   -   DC Current
      135              70                    1A

As you can see, using a three phase full wave rectifier substantially reduced the RPM for my load which was limited to 1 Amp due to the rectifiers I used and belt slippage at higher loads.  Theoretically, if the RPM were to be raised to 600, the DC Voltage would rise to 311 Volts.  To get that down to around 24 Volts, all of the coils for each phase would have to be paralleled, giving 25.9 Volts DC at 600 RPM.

Since I like the idea of a full wave rectifier from three phase, and the fact that it will be simpler to re-connect the coils in each phase to parallel, that's that I'm thinking of doing.

Now, I haven't thought this out fully but I think it will work close to how I calculated.  What I will do, however, is to experimentally connect a lead-out from one coil of each phase, spin it up and see if there's any agreement with my cypherin'.
-----------------------------------------------------------------------------------------------
12 July 2014:
Today, I isolated one of the coils in each phase for testing purposes.
 
Stator with one of the coils in each phase isolated.
I hooked-up the single coil per phase to the rectifier and spun-up the generator.  Here's what I got:    RPM    -   DC Voltage   -   DC Current
     580               30                  0.785A

Now that is more like what I need.  The actual Voltage is a little higher at a slightly lower RPM but we're in the ballpark.

I then rewired the entire stator so each phase consists of all 12 coils in parallel.  This should give something like the same voltage at something like 9.4 Amps, plenty to run the motor in the transaxle.  Now, if only Engine Number Four can make enough power to produce about half of that.
Here's the stator with all twelve coils in each phase paralleled.
The glue I'm using to hold down the paralleling rings is drying and tomorrow, I'll stick the generator back together and see how it does.  You may note that there are some radial cracks in the stator frame.  I don't know if I'm going to reinforce them or not.
-----------------------------------------------------------------------------------------------
13 July 2014:
Well, my cypherin' worked out.  After assembling the generator to the test fixture, I spun it up with various loads.  Here are the results sfter I cobbled-up a couple of large bridge rectifiers to be able to make some significant current.  All of these measurements were done with various parallel sets of light bulbs.  Here's what I got:
    RPM    -   DC Voltage   -   DC Current    -    Watts
     149               9.1                  1.00A               9.1W
     350             24.0                  1.072              25.7
     600             43.8                  0.758              33.2
     350             24.0                  1.585              38.0
     440             30.0                  1.95                58.5
     366             24.0                  2.50                60.0
     395             24.0                  4.20              100.8
     491             30.0                  4.80              144.0

That's not too shabby although it looks like I won't be running the engine at near 600 RPM, even if it can make it because of the amount of power to make about 150W.

Then, I hooked up the transaxle and ran it.  There wasn't a way I could load it up and measure the RPM at the same time but I made up a brake, locked one axle and fiddled the voltage and load until the motor was drawing about 5 Amps at 24 Volts and the generator was handling it fine.  As a matter of fact, when I finished testing it, I used the 'ol finger test to check the temperature of the coils and they were just barely warmer than room temperature.  Note that the transxle has a differential so the RPM was measured by locking one axle and then halving the RPM measured at the other axle.

Generator and transaxle under test.
Here's what I got with the motor hooked-up:
 Generator   Transaxle
    RPM    -    RPM   -   DC Voltage   -   DC Current    -    Watts
     340            200             24.0                  1.10A             26.4W

Since I now know what the unloaded RPM of the axles is, I can start figuring wheel size and ratios.  I'm thinking of using fairly large diameter wheels so I will have to reduce the speed by chain or cog belt drive.  My target speed is going to be around 5MPH or about 440 feet per minute at full speed.

Just taking a look at what wheels would be cheap, I thought of the "toy tire" spares.  Since I have one in my trailer, I measured it.  The outside diameter is about 23.5", which makes the circumference 6.152 feet.  Dividing 440 feet per minute by 6.152 feet circumference gives 71.52 RPM.  That means that I will need a reduction of around 2.8:1 to get 5 MPH.  Somewhere between 2.25:1 and 2.75:1 should work.

Now, I gues it's time to start on the CAD drawing of the newest Butt Buggy.  Here's what I've got so far.  Sorta like a slingshot dragster as I see it now.

Here's the CAD so far.
The thing at the top of the drawing is an overhead (plan) view of the transaxle.
-----------------------------------------------------------------------------------------------
17 July 2014:
For the last couple of days, I've been working on the mounting for the generator on the engine.  First, I had to remove the engine from it's skid and move it to the bench.  Then, I decided how I was going to set it up with the generator.

Mounting bolts in engine base.
To properly get to the bottom plate for layout, drilling and tapping, I had to remove the plate and set it in the mill.  The five screws are 10-32s.
   
First two mounting plates.                                                    Flywheel in place.                                             Stator temporatily mounted.
Next, three plates were made from 1/2" steel.  The plan was to mount the stator to the frame made up of the two plates you see in the left hand photo.  A third plate (just visible through the center hole of the stator) was bored to clear the rotor shaft and allow solid mounting of the stator.

I'm almost ready to start measuring for the rotor shaft.  
I still have to make a steel ring about 5/8" thick to act as a hard spacer between the stator and the mounting plate.  Once that's done, I can get a measurement to determine the length the rotor shaft needs to be cut to.

I will make a circular plate that will dowel into the center of the flywheel.  It will be bored to accept the end of the washing machine main shaft that has a spline that matches that of the rotor.  I hope to get a really tight press fit of the shaft to the plate.  If not, I will do some electrical gluing.  Just to make sure it has no runout, I will put the shaft in the lathe and take a face cut on the mounting plate.  It will bolt to the flywheel  with the two bolts shown.

I've been careful laying out the location of the stator but, before installing the rotor, I will indicate the stator using the rotor shaft as a mounting for the indicator.  If necessary, I can always hog-out the stator mounting holes to be able to bump it into center.
-----------------------------------------------------------------------------------------------
19 July 2014:
In the last couple of days, I measured, measured and measured again to come up with the mounting for the stator and the stub shaft for the rotor.  BZZZZZZT!  WRONG!

Yesterday, I shoulda stayed out of the shop because no matter how many times I measured for the vertical stator mounting plate, I got it about 0.070" off.

Mount is low!
As if that wasn't enough, I made the stub shaft for the rotor.  Again, I measured it three times and came up with an accurate dimension to get the right length to cut the original shaft.  BZZZZZZT!  WRONG AGAIN!
 
Facing mouting plate perpendicular to shaft.                                                                      Rotor shaft mounted to flywheel.
I don't know how that happened.  The stub shaft is over an inch too short!  It's not too much of a disaster except for the waste of time cutting the shaft and pressing it into the plate.  I have the other end of the washing machine main shaft that has an identical spline on it.  THIS time, I'll be even more careful!

Too danged short!
I can compensate for the stator mounting plate by making the stator spacer with oversize mounting holes for adjustment. Tomorrow, I'll get it right.  (says here in the small print!)
-----------------------------------------------------------------------------------------------
20 July 2014:
I got the previous day's screw-ups fixed and have everything mounted up and aligned..

"Correcting" stator spacer.
The stator spacer has oversized holes for mounting to the plate so it can be bumped to the center of the shaft before bolting the stator on.
 New rotor shaft in place.                                                                                       Rotor test fitted.
After pressing the previous too-short rotor shaft out of the mounting plate, a new shaft was made to the correct length.  After a bit of fiddling, I got the rotor to run clean without rubbing on the stator poles.  There's very little clearance between the magnets and the stator poles.  If I have trouble with interference when it's out in the dirty world, I think the output will be high enough that I could sacrifice some by putting the stator on a fixture in the mill and turning a few thousandths off of the pole pieces.  This will reduce the output but may be something to contemplate.

Ready to test.
Now that I've got the generator and engine coupled, I need to get the rectifier diodes and figure out how I'm going to test run it.  I may just re-mount it on the old wood frame.  That's probably the best bet because if I were to go to the trouble of building the butt-buggy and the generator didn't work as expected, I'd have a lot of work gone to waste.  I'll have to think on it.
-----------------------------------------------------------------------------------------------
27 July 2014:
The diodes have arrived and I've been thinking slowly about the control protocols for the engine and generator.  One of the things I've gotten done is the improved tachometer sensor interruptor wheel.  The original one had only eight slots (holes) in it which made the governor kind of "nervous" at low speeds.  We now have a sixteen slot wheel.  Whoopee-doo!  We are now upgraded from 8 bits to 16 bits.
 
         The sixteen slot tach wheel.                                                               The tach re-mounted on the engine.
I've retained the optical interruptor to sense the wheel.  This has an LED that shines across a slot onto a phototransistor.  When the light path is interrupted, the transistor turns off.

I'm now working on the timer.  It is going to be driven from the tach end of the camshaft which you can see to the left of the tach in the right-hand photo above.  I retained the magnet wheel but have changed to a smaller magnet to facilitate the electronic spark advance. With the electronic advance, the positioning of the magnet detector is fixed.  The electronics will switch the ignition from "fire on break" (retarded) to "fire on make" (advanced).  The absolute timing will be set by rotating the magnet wheel on the shaft.  The amount of timing change from retarded to advanced will be set by positioning the sensor closer (wider range) or farther away (narrower range) from where the magnet passes.  I'm not sure how this is going to work out as the magnet collects iron fuzz over time.  I've wanted to try something like this for a while and this is the time.

The ignition circuit is presently designed to work in conjunction with the tachometer electronics to change the timing when the engine speed passes an adjustable point.

The thought process is working out how to arrange feedback to the governor circuit from the generator rectifier in order to override the governor when generator voltage rises too high, somewhere around 30 volts.  I don't want to burn out the motor.  I'd say this shouldn't be too big a deal but if I did, it would backfire on me and end-up being a bag of worms.  So I won't
-----------------------------------------------------------------------------------------------
29 July 2014:
Yesterday I got the ignition sensor worked out.

Ignition Sensor.
As planned, I've got the sensor (a magnetic reed switch) attached to the bracket shown above.  The mounting hole to the engine is slotted so the sensor can be moved closer or farther away from the magnet. The closer the sensor is to the plane of the magnet, the greater the change in timing there is from "break" and "make" of the switch.  Right now, I've got it adjusted so at the "break" or retarded magnet position, the timing is about 10 degrees after TDC.  At the "make" or advanced position, the timing is about 20 degrees before TDC.  As I recall from earlier fiddling with the engine, it does all right with that much lead.

The working schematic. (Revised 2 Autust 2014).
Yesterday and today, I worked on the semi-final schematic for the battery charging circuit, the voltage regulator, the tach, the advance/retard and spark circuits.  I think it will work but have no doubt that I will have to change some things to get it right.  If there's enough interest and you want me to describe how the circuit works, I'll do it, but only if there is enough interest.  If you've got any electronic circuit experience, you should be able to work it out yourself.  Of course, I welcome comments.
 
Partially completed wire-wrap board.
At this time, I haven't included any motor control circuitry because it may not be needed.  Idling the engine should lower the output voltage enough to keep the speed down.  I'll probably put a switch on the "foot feed" so when the pedal is all the way up, the motor is out of the circuit.  Pressing down a little connects the motor at idle.  When "floored", the engine will attempt to speed to the governed limit but motor load should hold the speed down until the motor speeds-up and the motor current decreases.  When the motor load decreases, the engine will be able to speed-up until the voltage out of the generator reaches about 30 volts, at which point, my circuit will override the governor to slow the engine enough to keep the output below 30 volts.

At least, that's how I think it should work.  I'm going to do some testing of the engine/generator/controls at whatever loads I can apply and see how it works before building "The Mighty Hoyt-Clagwell Washbuggy".
-----------------------------------------------------------------------------------------------
2 August 2014:
The wire wrap board is done and tested.  The schematic above has been revised to reflect the few corrections and changes made during the tests.
 
Wire wrap board finished and tested.
For the time being, I will put the circuit back in the original Engine Number Four control box.  A nicer housing will be made for The Mighty Hoyt-Clagwell Wash Buggy.

Next step is to wire the edge connector to everything and mount it back on the engine.
-----------------------------------------------------------------------------------------------
3 August 2014:
The control box is almost done.  Next step is to mount it on the engine stand and connect the ignition sensor, tach sensor and throttle servo.

Control box almost finished.
Once the control box is mounted-up and wired, I'm going to see if I can make an electric starter for it like the ones on Delco 750 light plants. They use a pivoting DC motor with a friction wheel.  When the starter circuit is made, the starter motor pivots and is thrown into the flywheel by the inertia of the rotating mass of the motor. Once the starter is turned off, the motor rocks back so the friction wheel doesn't rub on the flywheel. I've got a motor off of a deceased 12 volt DC wash-down pump that I will make a friction wheel for and see if it will crank the engine.  Electric start would be a nice thing to have on The Mighty Hoyt-Clagwell Wash Buggy.
-----------------------------------------------------------------------------------------------
8 August 2014:
The control box is wired up to the engine and, after chasing my tail with the governor and throttle actuator circuit, I realized that the circuit would be flakey until it saw an input from the speed sensor.  A day wasted.

Starter motor with friction wheel ready for mounting hardware.
I made a hub and then made a rubber friction wheel out of something from the junk box.  Holding it against the flywheel by hand, it cranked the engine fine so I will make a sliding mount with a switch that will start the motor when the starter is hard against the flywheel rim.
-----------------------------------------------------------------------------------------------
9 August 2014:
I'm still working on the starter.  Because the motor base is flimsy and rusted, in order to make what I think will be a smooth slider, I ended-up having to do a lot of milling and turning to make the parts.
  
Bottom of motor base plate.                                                                     Top of motor base plate.
The motor base plate was milled with 3/8" slots for the shanks of the "nuts".  The motor base plate was also milled with 3/4" slots to fit the nuts.  The base was milled and the nuts were made so the shanks of the nuts would be proud of the top of the base plate by about 0.020" to allow clearance for the nuts and motor to slide.  The "nuts were made of 1" bar stock and, after turning, facing, drilling and tapping, were set in the mill and 0.125" was removed from two opposite sides so they would fit into the slots. 
 
Motor set on plates on top of base.                                                       Top motor plates on motor.
With the motor base plate plate right side up, there are two 1/8" thick spacers (not drilled yet) that fetch-up against the shoulders of the nuts.  Then the motor is laid on the plates and two more plates are mounted on top. These plates will be drilled to clear the 1/4-20 studs that will be bottomed-out in the nuts and will extend through the plates and have nuts to clamp the motor in position.  I fugure that, once the plate is mounted on the engine frame with the nuts in place, it can be bolted down.  Then, the spacers and motor are set on the mount, the motor is lined up and the studs are tightened down.

I'll arrange a switch that will close when the motor is firmly shoved into the flywheel.
-----------------------------------------------------------------------------------------------
10 August 2014:
It's slowly getting there.  The starter's on and working.  I mounted a big spring to hold the motor back away from the flywheel and a switch to turn the motor on when it's engaged in the flywheel.  It works like a champ and cranks the engine at a decent speed.  The method of engaging the starter motor is makeshift at this time and consists of a piece of 1/2-13 threaded rod that sits in a hole drilled in the engine skid.  Shoving the top of the rod toward the engine moves the starter into engagement with the flywheel.

Starter and gas tank in place.
With the gas tank re-positioned, it was time to see if it would run.  It took a few tries to get the timing set to where it would run.  This is because when I removed the flywheel, I replaced it in a different position and the timing marks don't line-up.  After I figured out what was happening, it started right up.

Then.....................................
KERPOW!!
After getting it running and before I could put a meter on the generator output, one of the electrolytic capacitors exploded on the control board.  DOH!!

The capacitor that blew is the one directly filtering the generastor output before the input of the first voltage regulator which sets the voltage for the battery charger.  Since it puts out about 14 volts, the 16 volt cap on the output is fine.   I wasn't thinking when I selected the input cap, though.  It is only rated at 16 volts and the generator can put out about three times that voltage when it's unloaded.

Tomorrow, I'll take the board out and replace the cap with one rated at 50 volts. That should do it.  What I probably should do is to put a couple of the 32 volt light bulbs in series across the rectifier to kind of dampen it's enthusiasm.
-----------------------------------------------------------------------------------------------
12 August 2014:
Today has been spent first replacing the capacitor with a 100 Volt version.  THEN, the fun began.  First, I had a lot of ignition issues and finally found the culprit.  It was R38.  At 33K, it was allowing a "dwell" or coil "on time" of 20 milliseconds, about 20 times too long.  This caused the Hexfet transistor driving the coil to fail due to the peak current being present longer than the transistor specification allows.  The IRF840 is an 8 Amp transistor that can withstand up to 32 Amps for very short intervals.  Since the coil resistance is 0.8 Ohms, the peak current will be 17.5 Amps.  Twenty milliseconds at 17.5 Amps is enough to short the transistor.  I really hate it when that happens.

Debugging mode.
Once I got radiator on and the engine would run for a while, I discovered that the output of the generator rectifier is over 40 Volts at idle with a two Amp load.  This is way too high!  I may have to re-rig the stator so I can use a half-wave rectifier as shown in the earlier figure as a "simple rectifier".  That should lower the voltage enough to be useable.  I'll think on it some.

Just think.  If I'd left it hooked up as it came out of the washing machine, the output would have been around 500 Volts at idle!  Now, that's some serious smoke potential!

While I had it running, I checked-out the throttle control and, although it worked, I changed the opamp from a 1458 to an LM319.  The LM319 can swing closer to the rails and offers better control.  More tweaking will be required to get it working the way I want.

I also checked out the automatic spark advance a bit and I think it's working but it needs to be tweaked-in.  Right now, the timing's close enough for the engine to run but it could run better.  

The battery charging circuit appears to be working but, because I cranked on the engine so much, the battery was pretty low and the charging circuit had a ways to go to get the battery happy.
The last thing I checked was the voltage driven engine speed limiter.  It does work but, at the high output of the generator, it will barely get above idle with the control set all the way up. 
-----------------------------------------------------------------------------------------------
13 August 2014:
Today I brought out the common terminal of the generator stator and changed the rectifier to half-wave.  This will work.  After fiddling with the timing and mixture and adjusting the Voltage limit, I connected four 26 Volt, 2.7 Amp light bulbs to the output of the rectifier for a total of 10.8 Amps.  The engine picked up the load fine and, running with the governor control to maximum (1,000 RPM), the Voltage limiter pulled down the RPM to 950 at 25 Volts.  Removing the load causes the voltage to rise and the engine speed to fall to around 400 RPM.

Running with a 10.8 Amp load.
When I added two more bulbs, the engine went to almost 1,000 RPM and the voltage dropped to around 18 Volts.  The total calculated load at 18 Volts was about 14.25 Amps.  This shows that the engine can drive the generator with power to spare.  The voltage drop is due to the resistance of the coils in the stator.

After starting and stopping a number of times and running at various loads for a couple of hours, the battery Voltage was back to about 13 Volts.  I think that if I hadn't done so much cranking and diddling around, the battery would have charged to nearly 14 Volts.  Since the charging circuit only puts out about an amp (at the best of times), it will take some time to fully charge the battery.  

I had to stop the test run due to the CO level in the shop rising to nearly 100 PPM.  This is not the alarm limit (200 PPM) but I don't want to take chances.  I guess that the CO is coming from the crankcase breather because I have a prety tight outside exhaust system so I can run engines during the summer without fear of heat stroke.

While the engine was running, I had some thoughts.  Maybe I need to just allow the generator voltage to control the engine speed then use a PWM controller
(which I have on the shelf) for regulating the speed of the propulsion motor.  A switch on the foot feed will bypass the Voltage control and force the engine to idle.  Remember, my ideas are subject to change without notice.
-----------------------------------------------------------------------------------------------
14 August 2014:
While I think about how I want the controls to behave, I'm working on The Wash-Buggy design.

The Wash-Buggy design so far.
I've still got a lot of design work to do, including selecting the wheels and steering arrangement plus the motor mounts, the dash and all of the dimensioning.

After I get the CAD mostly done, I'll look at places like Harbor Freight for the front wheels and a junkyard for the rear wheels.  I may (if the price is right) get a rear axle out of a small front wheel drive junker with the wheels, brakes, etc.  I can then cut and splice the parts into my design.  It would be nice to find something with mechanical emergency brakes.  That way, I'll have really REALLY good brakes on two wheels.  The transaxle has a built-in brake on the motor but I really don't think it's designed for anything but a parking brake.
-----------------------------------------------------------------------------------------------
17 August 2014:
Today, after thinking about it, I modified the controller so the engine speed is primarily controlled by the generator voltage and what used to be the foot control has been changed to a fixed maximum RPM control.  The schematic is below:

17 August 14 revision.
As you can see, I've added an R-C network to the "Max RPM" setting to get rid of potential jitter.  Also, notice that I've added another stabilizing factor for the Voltage control.  D12 and C15 will average out any distortion and noise in the output of the generator caused by the motor commutator.  This will give better control of the Voltage.  I've also added a calibration chart for the tachometer.
-----------------------------------------------------------------------------------------------
19 August 2014:
I dug out the PWM motor controller I got with the motor I used on the Algore Edition Green Hybrid Hoyt-Clagwell and hooked it up.  It'll do fine.

Testing the PWM motor controller.
Then, I came to the realization that I don't have any extra space for The Wash-Buggy.  Gotta get rid of something to make room.  Either that or forget the new Wash-Buggy design and again modify the 2009 Hoyt Clagwell tractor.  Presently, it has the 30-60 engine on it.  I suppose if I can't sell it, I could just remove the 30-60, shorten the chassis and put the original Algore Edition motor back on it and call it The Former Algore Edition Green Hybrid Hoyt-Clagwell.

So now, if anyone is just burning with desire to own a one-of-a-kind butt-buggy with the homebrew 30-60 engine on it, give me a holler.  Click on "The Hart-Parr 30-60 Semi Replica" for the engine and "The Hoyt-Clagwell 30-60" for the tractor it powers.  We can do business.
-----------------------------------------------------------------------------------------------
24 August 2014:
During the test runs, the flywheel and rotor of the generator have been getting more and more runout.  Today, I tore down the engine and am working on repairing the crank.  You can see what I've been doing by going to Engine Number Four, Page Two.
-----------------------------------------------------------------------------------------------
8 September 2014:
We now return here for the application of the engine/alternator to The Hoyt-Clagwell 30-60 chassis which will be re-powered and changed back to the electric drive that it originally had.  This time, batteries will not be used as the propulsion motor will be driven directly from the rectified alternator output.  A lot of work will be avoided because I saved all of the major electric drive components.

I think we can assume that the tractor will be modified enough that we can change it's name to "The Washbuggy".  After installing the engine/alternator and hooking everything up, if it runs well and seems to be a keeper, I'll shorten the frame to fit the present locomotion scheme.  Shortened, it will look a bit "compact" so I think I could call it "sorta green".

In the last few days, I've removed the 30-60 engine from the tractor and put it in hibernation (sitting on a creeper, shoved out of the way).  That engine will eventually be put on a skid for showing.  It will do what it does best - make sounds like a small Hart-Parr 30-60 and do nothing.

 
 
      Chassis with engine/alternator temporarily in place.               Original drive motor and foot-feed back on tractor.  
The engine/alternator will sit as far back on the chassis as I can manage it.  After wrestling it around, I think I have it in about the position I like.  I'm not sure I will have to put a housing around the rotor because a test-sit (on the seat) has my feet out of the way of it far enough that I don't think there will be a problem.  I will re-install the brake switch to disable the motor drive when the brake is on.

I've also added a microswitch to the foot-feed that is adjusted so it closes when the pedal is all the way in the idle position.  I intend to wire this switch through a dash switch to the control board to override the governor control to force the mixer to go to the idle stop.  The dash switch will be there to disable this "feature" when I want to move out immediately after mashing the feed and not have to wait for the engine to pick up speed.  The particular motor controller I have cuts-out at below about 20 Volts and when the engine is at the idle stop, the Voltage is only about 15 so there would be a pause while the Voltage builds-up.  I think I'll call it The Exclusive Hoyt-Clagwell Getaway feature.

Motor and starter mounts.
As usual, the motor mounts come from a scrap pile somewhere.  They are 2" X 4" box tubing.  I think they will be strong to keep the engine from falling off because the wall thickness is 1/4"!  As you can see in the photo above, the tubes are roughly positioned as they will be on the frame.  They will go crossways to the frame rails.  The angle iron on the left-hand side will be welded to the tubes and will mount the starter.

The little radiator that is on the engine now seems to be just adequate to handle the heat load of the engine but, if I think the tractor needs more class, I may keep the stack and move the little radiator there and save the big radiator for something else.  Of course, if the engine is found to run hot under driving conditions, I could always put the big radiator back on.
-----------------------------------------------------------------------------------------------
9 September 2014:
Much of the day was spent on the mill, cutting a tunnel for the oil drain pipe and squaring-up the motor mount.

Partially finished motor mount.
I think the tubing that I got came from a shop back in Kentucky.  One end of each piece of tubing had been chomped-off with a cutoff wheel and the pieces were probably thrown out due to their being bent-up.  It took a lot of time to get the top and bottom surfaces trued-up after welding them together.  Tomorrow, I hope to get the holes drilled and get a coat of paint on it.
-----------------------------------------------------------------------------------------------
10 September 2014:
I thought I'd have a photo of the engine/alternator bolted to the chassis but, at the last minute, I found an interference between the heads of the bolts holding part of the stator frame and the motor mount.  Close, but no cigar.

Almost but not quite a fit.
Tomorrow, I'll unship the whole thing and mill a slot in the mount to clear the bolt heads.
-----------------------------------------------------------------------------------------------
11 September 2014:
Well, the engine's finally mounted to the frame and the gas tank and starter are back on.  I'm considering hooking up the radiator in the stack instead of using the little one on the engine.  I'd have to do some fabrication to make a mount for it and it just seems to be too much trouble at this time. I might even route the exhaust to the bottom of the stack to see how it sounds.
Once I get that done, I'll be ready to start on the dash panel and wiring it up.
-----------------------------------------------------------------------------------------------
14 September 2014:
It's getting close to testing time.  For the last couple of days, I've been working to move the electronics from the temporary wooden box on the test stand to a purpose-built control panel.  

In rat's nest mode.
All I have yet to do is to finish wiring the chassis and check for smoke leaks when I power it up.  After some preleminary electrical testing, it's off to the official Hoyt-Clagwell Test Track for trial runs.  This could be in the next couple of days or so.
-----------------------------------------------------------------------------------------------
16 September 2014:
It didn't go quite as quickly as I planned.  Only had one smoke event.  The ignition coil driver decided to leak.  After winkling out the cause of this and making some trivial wiring changes, I cranked it up to see if everything worked.  The only things I haven't checked yet are the brake safety switch (disables the motor drive when the brake is on) and the battery charging voltage.  

One thing I have yet to correct is the motor direction.  After exhaustively making sure I had the motor polarity correct, it wasn't.  No biggie.  I just have to reverse the motor leads.  I think it will be better to have six speed ranges forward and two ranges backward rather than the other way around.

The last thing I did before giving it a quick crank was route the exhaust to the stack.  Wow!  What a sound!

Here's how it looked when I quit for the day.  Tomorrow, I do the motor polarity thing and adjust the settings and go for a ride. .........  I hope.

The Washbuggy, nearly ready for the first road test.
Eventually, I'll make a cover for the control panel (dash) to protect it from industrial espionage agents.  You can never be too careful with trade secrets.  I know the secrets of Hoyt-Clagwell & Company are safe with you!
-----------------------------------------------------------------------------------------------
17 September 2014:
The motor polarity change has been made, calibration done and road testing has begun.  The drive won't give you wiplash from acceleration but it goes better than with any of the previous engine/drivetrain conbinations.

Things were going along pretty well on a flat street and I decided to try a slight hill.  As soon as I turned the corner, the tachometer quit which, of course made the engine try to race but because the ignition control works off of the tach signal, it went to retard and that kept the engine under 1,000 RPM.  The mixer is handy so I grabbed the throttle while I got off to see what was the matter.  I figured the tach sensor was not working.

As soon as I got near the tach sensor, it started working so I decided to head back to the shop.  As soon as I got back on, it quit again so I nursed The Wash Bugy home hanging on to the throttle shaft.  After I turned the corner, the tach started working again and worked all right until I turned into the driveway then it quit again.  It acted like the sensor was intermittent so I adjusted the interruptor wheel position, thinking that would fix it.  When I first started the engine, it acted fine but when I stepped away, it tried to race again.

I found the problem to be the sun.  The tach uses an infrared LED and a phototransistor facing each other.  The interruptor, mounted on the crankshaft, has slots in it which interrupt the light beam.  When the sun shines on the sensor in the right position, it blinds the transistor so the tach quits working.  Below, you can see my temporary fix.

Temporary shield for tach sensor and muffler.
I also quickly tired of the engine sound when it was routed to the stack (the thumping was giving me a headache) so I substituted one of my larger home-made mufflers.  Now, the loudest thing heard is the whine of the alternator.

I noted that, if running in a higher gear (requiring more motor current), if I mashed quickly down on the foot feed, the speed controller acted like it was in some kind of limiting mode, only allowing about half-speed then slowing to a stop.  Ordinarily, if this was the case, backing off of the foot feed would correct the problem and you could gently step down and get full power back.  Not the case with this controller.  The only way to get full power back is to go to idle (the low Voltage shutting down the controller) then slowly press down on the feed.  With the pedal all the way up, full idle to the throttle stop is forced.  Moving the pedal down a little defeats this forced idle and the engine picks up speed to provide adequate Voltage to the controller.  Slowly pressing further starts the motor through the controller.  Doing this would allow full speed operation.

My theory (not tested today) is that the controller, being designed for battery power, doesn't like the ripple from the alternator rectifier and is getting upset by it.  Once shut down and re-started, it's okay until pushing too hard causes the ripple to become great enough to destabilize the circuit.  What I plan to do in the next session is to add an electrolytic capacitor to the output of the rectifier.  I've got a biggie of several thousand microfarads at 50 Volts.  This should serve to at least prove the theory.  More capacitance may be required for heavy loads.  We'll see.
-----------------------------------------------------------------------------------------------
18 September 2014:
The permanent cover for the tachometer sensor is done and I've added a 41,000 microfarad @ 60V electrolytic capacitor to the output of the alternator regulator.  I also changed the exhaust to the system that was on the tractor when it was powered by The Homebrew Hvid and The McVickerish engines.  It's not as quiet now but it's permanent.

Testing proved that adding the capacitor stabilized the motor controller.  The tractor now runs a little faster and accelerates a bit better.  I'm a bit concerned that the alternator output can barely reach 22 Volts when the drive motor is pulling top gear.  The engine speed is set to max out at 1,000 RPM and it almost constantly runs at that speed when driving.  1,000 RPM is, to me, a little fast for that plastic rotor!

What I might do is add the three diodes to the rectifier to make it full wave again.  This will increase the Voltage output by a factor of about 1.5 (maybe too much) and could allow the engine to run slower and provide 24 Volts to the drive motor.  I'll think on it.

All in all, I'm getting happier with the performance of The Mighty Washbuggy.  It will just take a little more happiness for me to remove the stack and shorten the frame about two feet to give it that sporty look.
-----------------------------------------------------------------------------------------------
19 September 2014:
Today, the rectifier got converted back to full wave and a few other tweaks were done.  Once I got it running, several things happened.  The battery charging regulator shorted, taking out the 16 Volt electrolytic in the output (KERPOW!!) then taking out the 2N3055 charger driver.  As I figured, the voltage from the alternator is really high, and can get to over 70 Volts if the RPM is allowed to climb.

In any case, I did try a short road test.  Where yesterday, it acted like the clutch was really slipping, today, it acted like it was stuck in high gear.  In order to keep the alternator output to a reasonable level (under 40V), the engine was barely over idle.  When accelerating, the engine would try to speed-up but was running too far below the torque peak to be able to pick up.

After replacing the blown components on the circuit board, I had an idea that I'll try tomorrow.  That is to simply disconnect one of the alternator windings.  That ought to lower the Voltage and may allow the "gear ratio" to be correct for the engine.  If the alternator had a wound field, I could vary the output by changing the field current.  With a permanent magnet field, the only way to regulate the output is to use a pretty heavy duty circuit. This, I really don't want to do.

If all else fails, I could design a pulse width modulated driver and use it in the alternator output to drive the drive motor.  A smaller linear regulator could be used to operate the battery charging system, ignition, etc.
-----------------------------------------------------------------------------------------------
20 September 2014:
Okay.  Before I tried the disconnected phase winding idea, I made a "pre-regulation" circuit for everything but the motor drive and Voltage sensing to the governor to lower the raw DC output from the alternator to around 25 Volts.  This was done using an 8 Amp, 150 Volt transistor and a 24 Volt Zener diode.  I hung it, with it's heatsink, temporarily off of the edge connector for the control board.  

Pre-regulator for the electronics and battery charging circuit.
When I cranked-up the engine, the alternator Voltage was not appreciably lower.  However, the pre-regulator works fine and, typical of a Zener diode Voltage regulator, it's sloppy.  The output Voltage of the pre-regulator varies from about 22 Volts when the engine is idling to about 25 Volts at full roar.  This is much more than adequate regulation.

After scratching my head a bit, I came-up with a way to lower the Voltage without removing turns from the windings.  What I did was space the rotor out about an inch from the stator.  The experiment consisted of a stack of 3/8" flat washers between the end of the rotor shaft and the rotor (inside the hub).  This worked well, so I machined-up a permanent spacer to go between the face of the rotor hub and the step diameter on the rotor shaft.
 
         Spacer sitting on rotor hub face.                                               Rotor installed, spaced away from the stator.
Now, the peak Voltage during governor excursions isn't nearly as scary.  Instead of the Voltage rising to nearly 75 Volts at the peak, it only rises to about 50 Volts.  Sure, the efficiency of the alternator is compromised but it's the easiest fix and avoids serious surgery on the stator coils.

The road test confirms that this modification is definitely an improvement.  The engine can now run at about 900 RPM at maximum load.  At idle, it runs at about 450 RPM with the idle circuit engaged.  With no load and with the forced idle disengaged, the RPM is around 600.  The drive motor will pull The Washbuggy in high gear, high range but falls on it's face on any decent grade.

While testing, the engine started faltering.  By the time I got it in the driveway, I had a towel almost covering the air intake.  It finally quit about 10 feet from the garage.  It acts like a fuel system problem - probably a fleck of dirt in a passage.  I'll figure that out tomorrow.
-----------------------------------------------------------------------------------------------
21 September 2014:
Well, tomorrow is here and I figured out why the engine sorried-out yesterday.  The jet in the mixer must have had a boulder in it.  Wire wouldn't move the obstruction and I ended-up running a #65 drill through it to bust the clod loose.  While I had the mixer apart, I noticed that the flutter choke had broken.  Although it still worked, there were two pieces to the valve instead of one.  I fixed that by turning a new air valve out of brass bar stock all in one piece.

Then, I decided to go ahead and make a proper choke so I wouldn't have to mash a shop towel over the air intake to get it started.

The choke works fine and, after re-adjusting the mixture and flutter choke, the engine runs much better.  One more thing I think I'll do is to force the timing to retard with the foot feed is up.  This will give a good slow and steady idle.

Last night, I had a brainstorm that I might try when I get another washing machine motor.  Regulation of the output of a relatively high powered permanent magnet alternator is not easily done. As I did yesterday, I de-coupled the rotor (field) from the stator (armature) to lower the output.  What I'd eventually like to try is an arrangement where the stator can be slid back and forth, changing the coupling.  Fully engaged is full output. Pull the stator back far enough and the output will be very small.  Link the stator to the foot feed.  To go faster, move the stator to more tightly couple it to the rotor.  That idea should work but it still doesn't lend itself to easy automatic regulation because of the power required to provide the force to move the stator.

About an inch and a half of movement should work.
-----------------------------------------------------------------------------------------------
22 September 2014:
A little time was spent today updating the documenttion.

Latest revised schematic.
Note that the pre-regulator is now incorporated along with the forced idle circuit.  I haven't drawn how I think the spark retard should be wired into the forced idle circuit.
-----------------------------------------------------------------------------------------------
23 September 2014:
Today, I got the pre-regulator mounted on the circuit board.  I also added a resistor and diode to make the ignition go to full retard when idle is forced.  All of that seems to work well.  

While testing today, I made up a Voltmeter that I mounted on the control panel to monitor the output of the alternator.  First, I noticed that there was a pronounced Voltage droop between no load and full load.  This was partially fixed by lowering the value of  R43 from 220K (it's wrong on the schematic above) to 100K.  

The Voltage still droops quite a bit at full load and the meter shows that the peak Voltage is only about 45 Volts.  I believe I need to shave about 0.100" from the rotor spacer to increase the coupling.  This will give the alternator more muscle while not raising the Voltage to a dangerous level for the filter capacitor.
-----------------------------------------------------------------------------------------------
16 October 2014:
After a brief pause to make a throttle plate for The Upside Down Engine, I'm back to The Mighty Hoyt-Clagwell Washbugy project.

Since it looks like we have a winner with the drive, I'm now shortening the frame of the tractor to get rid of some unused space on the frame.  The smokestack has been removed and will be saved for use when the 30-60 engine is skid mounted.  The original radiator which was inside the stack will be retained.
   
Frame marked and ready to cut.                                                     Cuts made.                                                       Front section and join plates.
A sawzall isn't what I'd call a precision cutting tool but it gets the job done and is a heck of a lot easier than a hacksaw!  The total amount removed is right at 23 inches, give or take, depending on how much material is lost in sqauring-up the two pieces that are to be re-joined.

The join plates are made of six inch lengths of 1/4" X 2" cold rolled steel.  The pieces will be drilled and tapped for four 1/4-28 machine bolts and will be inserted in the tops and bottoms of the frame rails.  The rails will be drilled to 3/8" for the bolts.  This will allow some wiggle room to get the alignment correct.
-----------------------------------------------------------------------------------------------
17 October 2014:
It's stuck back together and aligned.  Just for funzies, I put the radiator on to see what it looks like now.

The Mighty Hoyt-Clagwell Washbuggy.
Nice and compact!  As you will notice, the front end is a little high.  It was like that before the shortening but didn't show-up so much.  I do have another mounting point for the swing front axle that is about an inch lower and I might try that to see if it's worth the loss of axle swing angle.

I still have to mount the radiator and plumb it up to the engine and work out the exhaust routing.  After that, I can proceed to re-install the horn.
-----------------------------------------------------------------------------------------------
22 October 2014:
Yesterday, I changed the front axle height and it now looks a bit more level.  

I got it put back together and did a test run.  All was well until I looked down and found my shoe covered with oil.  Ever since I changed the crankshaft around, there has been a slight leak around the flywheel end main bearing.  I think the reason is that I took out a shielded ball bearing and replaced it with a sealed bearing.  This causes the oil that gets past the seal and the slip fit of the bearing in the crankcase to accumulate faster than it can run back.  Once the space between the bearing and the thrust plate is filled, there's no place for it to go but out where the shaft goes through the thrust plate.  I also think the problem is made worse by unrelieved crankcase presure.  The breather is working fine but, under load, you can really feel it "puffing".  Then, there's the fact that I ran the engine almost at high idle (1,000 RPM) for most of the time yesterday.

Today, I took off the alternator, flywheel, etc. and removed the thrust plate.  There wasn't but a drop of oil in the space between the bearing and where the shaft goes through but it did have all night to drain back.

What I'm doing now is fitting a lip seal to the thrust plate to seal against the flywheel hub.

Seal mounting plate on thrust plate.
I made an aluminum plate into which to press the seal.  The plate bolts to the bearing thrust plate.  Tomorrow, I'll put the flywheel on the rotary table on the mill and mill down the hub to fit the seal.  I have to do it that way because the flywheel is much too large to fit in the lathe and turn.

I'm also thinking of removing the one remaining seal from the bearing to allow the oil to drain back.
-----------------------------------------------------------------------------------------------
23 October 2014:
It took a good part of the day but I've now got the flywheel hub machined to take the seal.

Turning with an end mill.
Even though it's slow, turning a diameter using and end mill gives a very nice finish.  When the seal gets here (tomorrow), I can put it back together and see if it leaks any more.  Oh, yes.  I removed the outer seal of the main bearing so oil can drain back quickly.  I'd already removed the inner seal when I put the engine back together.  That allowed the bearing to get plenty of oil.

With nothing better to do to finish off the day, I put the horn back on.  I find that, with a much shorter air line from the pump to the horn, it sure blows louder.
-----------------------------------------------------------------------------------------------
24 October 2014:
Okey dokey.  Yesterday afternoon, I got the seal and today I put the engine back together.  I did some tweaks to the timing and, after running it for about three hours, there is no more oil leak so I can consider the fix a success.  

I'm about to declare the project finished.  Once I do the cosmetics, I'll pose The Mighty Washbuggy for a final photo and, if I feel like it, make a movie.
-----------------------------------------------------------------------------------------------
27 October 2014:
Today, I consider The Mighty Hoyt-Clagwell Washbuggy to be done enough for a formal portrait.
 
Ready to go.
I've got sore fingers from polishing the copper and brass.  I really should consider regular polishing of the brightwork because, when it is left for a few years, it gets really grungy.  

Still to be done is a cover for the control panel, some kind of linkage so the starter can be worked from the driver's seat and a few more minor tweaks.  

As far as performance is concerned, it will go at a reasonably fast walk on the level in high gear/high range.  In top gear, the motor bogs down on any more than a minor grade but the engine isn't really breathing hard.  On a really steep grade, it will pull with ease in low gear/low range but it takes a while to get anywhere.

Maybe tomorrow I'll put the new memory card in the camera and make a movie of it.
-----------------------------------------------------------------------------------------------
1 November 2014:
Tomorrow?  You betcha!  Anyhoo - here's the flick of The Mighty Hoyt-Clagwell Washbuggy.
And here's the schematic as it is today.  Note that I've added a bleeder resistor across the alternator filter and I've added a load dump circuit to keep the Voltage from going through the roof on governor excursions and when qwuickly removing the load.

Schematic as of 1 November 2014
-----------------------------------------------------------------------------------------------
16 November 2014:
The last couple of weeks have been occupied with making minor tweaks.  

One of the things I've been doing is experimenting with the coupling of the rotor and stator of the alternator, trying to find the sweeet spot between too tight coupling and the engine not able to make enough power to drive the motor at full load and too loose coupling with the engine hitting the RPM limit before the Voltage limit is reached.  Right now, the engine can just pick-up to reach high idle at maximum load.  Another thing that hurts when the coupling is tight is that the engine can't increase RPM enough to get higher on the torque curve.  It gets stuck at a point where it is maxed-out and can't climb the curve to enough torque to speed-up. I think this is going to be an inherent trait of the drive system.

As it is, the tractor goes at a fast walk in top gear and will pull very well in all but the highest gear.  In low range/low gear, I think it will pull stumps but it will take a long time to do it!

Another thing I haven't been satisfied with is the throttle plate in the carburetor.  It's very difficult to get a small plate to fit tightly in the bore.  As a result, idle speed is too high.  What I did today after making two brass throttle plates was to make a third one out of PTFE Teflon.  This is a high temperature "slippery" plastic that I figured would act a little like rubber and would contour itself to fit the throttle bore.

The two metal butterflies and the new Teflon butterfly.
Because Teflon is not noted for it's strength or mechanical stability, I made the new butterfly 0.200" thick.  The brass ones are made from 0.035 material.  It was a real challenge to get the butterfly fitted onto the throttle shaft in the bore but, after some cussing, it got there.

Tests this afternoon show that the new butterfly works a lot better than the metal ones and I can get the engine to idle down to nothing (which is just a tad too slow!).  Now, I will have to adjust the spark advance setting so it advances at a higher speed because, when the pedal is up and the spark is retarded with the engine idling around 300 RPM, when I push the pedal down, the spark advances too much for that low speed and the engine stops.

Once I get the timing shift point changed to about 400 RPM, the engine should pick up just fine.

Then, I have to work on the Voltage governor damping.  Now that the throttle closes properly, the governor hunts at just about any speed.  That, in itself, isn't a deal breaker but I'd prefer for it to behave a little better.  I'll probably fiddle with the resistor in the collector circuit of the Voltage sensing transistor.  I hope all I'll have to do is increase the value some in order to introduce a little more negative feedback.  Knowing circuits, that simplistic approach probably won't work and I'll get involved in a tail-chasing exercise.  Oh, well.  Because I don't have a client breathing "deadline" down my neck, the tail-chasing should be semi-fun.

We'll see how it works out.
-----------------------------------------------------------------------------------------------
28 November 2014:
This is one of those "why didn't I think of that!!" entries.  Armand Hoffstetter emailed me with the suggestion that I change the Wye configuration of the alternator to Delta.  This would lower the output Voltage and make the output a bit more controllable.    

Today, I removed the stator and changed a bunch of coil connections in order to make it a Delta.  That took a while to do but, almost as time consuming was re-indicating the stator when re-mounting it.  This has to be done in order to make the stator and rotor concentric.  Here's the latest schematic.

 

Revised schematic (Click for higher resolution image).
I've drawn the alternator with all the coil connections showing.  Also note that I've changed a few things in the Voltage controlled throttle portion of the circuit to give a little more stability.

When I reassembled the rotor, I left out the spacers that had decoupled it from the stator.  

Tests this afternoon proved that the Delta connection gives lower Voltage even with full coupling and now the load dump circuit rarely kicks-in.  It's now reached a point where engine power is the limiting factor.  When pulling a higher gear, if the foot feed isn't carefully pressed, the engine will bog down at a low RPM and is incapable of climbing the power curve against the increased power requirement of speeding the alternator.  The solution when this happens is to back-off the foot feed, causing the motor controller power draw to lessen.  This will allow the engine RPM to rise.  Once the Voltage rises to where the throttle closes, the feed can be increased incrementally until best speed is reached.

Gearing of the tractor is such that high gear in high range is only good for more or less level ground.  Second gear in high range gives reasonable grade pulling capability.  The lower gears are for greater grades or softer ground.  Low gear in low range allows the tractor to pull the steepest grades, albeit slowly.
-----------------------------------------------------------------------------------------------
29 November 2014:
This morning I tried an idea that occurred to me.  I added a 5K potentiometer in series with the top (+8 Volts) terminal of the RPM LIMIT pot.  This allows the engine speed to be controlled from the dash.  I then adjusted the maximum Voltage pot for about 35 Volts.  Doing this allows the engine speed to be controlled independent of the Voltage control. All the Voltage control does is to limit the RPM to keep the Voltage within reason.  The RPM pot sets the engine speed from about 600 RPM to about 950 RPM.  

When going slow on flat ground with the RPM set to it's lowest point, even in high gear, high range, the accelerator can be pressed until the engine starts losing speed then backed off a little.  The tractor will run along happily at about half of it's full RPM speed in the same gear.  The engine still can't quite make enough power at full RPM to allow full pedal operation in high gear, high range on flat ground but it will almost do it, running at almost a slow jog.

Next is to find the pesky glitch that causes the tachometer to jump occasionally.  I'm sure it has something to do with the ignition circuit introducing some pulse garbage into the +8V line.  Adding bypass caps doesn't seem to affect it.  I might try putting a small power inductor in series with the +12V going to the coil driver.  Something to fiddle with.
-----------------------------------------------------------------------------------------------
4 December 2014:
Well, I FINALLY figured out what was causing the circuit instability.  While adding bypass caps, I happened to touch the clock input to the 4013 D-type flip flop.  The ignition went bonkers! Long story short, the CD4013B chip (RCA brand) was flakey.  The Clock input is supposed to only trigger on the positive-going transitions.  This one was triggering on both the positive-going and negative-going pulses.  THEN, I remembered that I'd had trouble with that particular brand of 4013's.  I swapped it with the last Motorola 4013 in the drawer and it straightened out.

Also, while fiddling around chasing the glitch, I changed the ignition pickup from the magnet and reed switch to an optical interruptor operated off of a slotted disk.  Then I changed the input transistor inverter circuit to a Schmitt trigger, giving it about a Volt of hysteresis.  That cleaned up the interruptor input.  The ignition is more stable now and the throttle is virtually smooth.  At low loads, it nicely moves the throttle.  The ignition timing is now firing the plug at TDC on retard and at -20 degrees on advance.  It seems to be the best setting for smooth running.  The spark automatically goes from retard to advance at about 600 RPM.  I may change that as I get more time on the Washbuggy.
 
And that's the "fallin' off a log" method of engineering.

Oh, yes - D.W. thinks a better name for the buggy is "The Cabrio".  This is the name of the washing machine the motor/alternator came out of.  I'll think about it.
   BOY! This is fun!

Comments? [email protected]

Hits since 29 June 2014