The Earth Explorer Project

Auxiliary Power Unit - Mark 2

I decided my original design for the Auxiliary Power Unit (APU) was too large. It would give me all the power that I need. But it was too big to cram into the current frame. So I wanted to experiment with a smaller solution.

I have a "weed whacker" (gas powered string trimmer) that I bought at an estate sale for $5. I went to start it one day and the starter rope got munged up inside. So I took it apart and thought I fixed it. After the third time of fixing it, with super glue the third time, I decided it was a lost cause. So it went in a box in my garage. I decided to get it out and do APU experiments with it.

I had been toying with the idea of using a permanent magnet DC motor as a generator. They are easily obtainable and I thought I might be able to use it to start my gas engine as well as using it for a generator. But I had no idea how big the motor would have to be to start a weed whacker (WW) engine. My WW engine is a 2-cycle engine that is either 18 or 20 cc's. I think engines this size are rated at either 1 or 1-1/2 horse power. I think the manual (yes, I got a manual with it) says it is rated at a maximum of 7000RPM.

Here is a series of messages I posted to the Twin Cities Roboitcs Group list server about this project:

Subject: Measuring Starting Torque
Date: Fri, 16 Feb 2001 03:32:51 -0600
From: jsampson@pobox.com
To: tcrobots@orbis.net

I have been thinking about my gas engine generator project again. I was wondering how I can determine the amount of power it will take to start a gasoline engine.

Specifically I am thinking about using my weed whacker engine to experimnet with. First I thought I could put a torque wrench on the shaft and turn it through a compression stroke. However I don't expect the torque to be the same when slowly turning the shaft as when the shaft is turning fast enough to start the thing. (IE. Turning slowing lets the compression bleed off.) And they also brag about "Easy Spin Starting" on the box on the new ones. I don't know what that implies. Maybe the way the ports are cut so there is less compression at slow speed?

So I figured it might be easier to connect the engine to an adequate size DC motor, like the ones in my Earth Explorer. Apply power and see if it turns over. If it does I can record the instantaneous current (digital scope) and estimate how much power it takes to turn it over. I could then apply the voltage I think represents the correct amount of power and see if it turns over.

I'm still toying with the idea that I can use a DC motor as both a starter and a generator. I found a flat motor at Axman that claimed it was rated at 24V and 55W. It really screamed when connected to what they claimed was 24V. So I may be able to gear it 2:1. This means my engine could run slower (quieter) and I would also have more starting torque.

Maybe I need to take my engine into Axman and hook it to various DC motors and see what will turn it over. I just better make sure it won't start. I'm sure they would kick me out if it actually started up...

Subject: Re: Measuring Starting Torque
Date: Sun, 18 Feb 2001 00:59:28 -0600
From: jsampson@pobox.com
To: tcrobots@orbis.net

Well I still don't know how much torque it will require. But I am closer...

I looked at motors at Axman today to experiment with. But I had no idea what I needed and nothing seemed to jump out at me. So I was sitting around trying to think of the next approach. I decided I need a way to mechanically connect an electric motor to the gas engine.

The output shaft has a square hole that is .2" across and accepts a flexible cable to run the "whacker". Somewhere along the line I realized I could put a toothed pulley or sprocket on the outside of the shaft. But it is a strange size. (.0424") So I want to turn a couple pulleys on Alan's lathe to fit.

I kept looking at it and decided to cheat. I have one of those "57 bits" screwdrivers that you can buy for $9.99. I was hoping I could find a square bit that would fit the shaft. But instead I found a flat scewdriver bit that would fit across the square hole. So I chucked it into my variable speed drill, grabbed everything real tight and squeezed the trigger. (I pulled the sparkplug wire first, since I was doing this was in my kitchen. :-) I thought it turned over okay, but it was hard to tell since everything went flying. So I pressed it up against the counter top and tried it again. And it cranks it right over. So I still don't know how much torque it takes, but I know my drill will do it.

So I looked around for a motor of appropriate size but didn't find anything. I have big motors or small motors or motors with gearboxes. But no medium size motors. Then I remembered the motor that Rob T. got rid of about a year ago at a meeting.

I dug it out. It is a Pittman 14203 and says 30.3 volts.

http://www.pittmannet.com/pdf/lcm_bulletin.pdf

I played around until I had a mechanical adapter to couple it 1:1 to the engine shaft. I held everything down and connected it to 24V. It turned over, sort of. Probably not fast enough to start. But it gives me a good idea of what I need.

I need to make a better mechanical coupling to do some more experiments.

So how well would this motor work for a generator? I have decided to chuck the motor shaft in my electric drill. It has the RPM listed on the side of the drill to give me an idea of the speed. Then I can measure the voltage and current.

So I will play with it some more.

Subject: Generator experiments
Date: Sun, 18 Feb 2001 07:40:06 -0600
From: jsampson@pobox.com
To: tcrobots@orbis.net

I connected my drill to my Pittman 14203 motor. I just tighened the motor shaft into the drill chuck. The drill claims to have a speed of 0-1200RPM.

Running wide open with no load on the motor I read about 10.5 volts at approx. 1200RPM.

Then I connected a 1 ohm load. I read about 3.3 volts at 1200RPM. Since I had a 1 ohm load I also had 3.3 amps of current.

So I found toothed pulleys to give me a 3:1 gear ratio. I put the big pulley in the drill and the small one on the motor. I used the 1 ohm load and started it up. My belt kept slipping, but I got around 6 volts (and 6 amps).

So I created a 5 ohm load resistor. Now I ran it and got about 16 volts which should have been around 3600RPM. Actually the drill sounded slower since it had a bigger load. So at 16 volts with a 5 ohm load I had about 3.2 amps. (16*3.2=51.2 watts)

So I need to do some more experiments. (Maybe with a little more control. :-) But I may need a bigger motor for a generator. I probably need a bigger motor to be used as a starter anyway. I could just spin this one faster. But I don't know how fast I can spin it before it comes apart. (The windings tend to fly off the armature when over-reved.)

Subject: More money down the drain???
Date: Sun, 18 Feb 2001 18:11:42 -0600
From: jsampson@pobox.com
To: tcrobots@orbis.net

I just bought another motor at Axman, $25. It is a Dynetic Systems http://www.dynetic.com model 312003A. It is 3" diameter and about 4-3/4" long. It has a tachometer generator on the end of that. It says on the label, "Rated 35 Volts DC 8.4 Amps", "Generator volt grad. 14 VDC/KRPM". I don't plan to use the generator and I would like to take it off. But it looks like it is the same shaft as the armature. So it may require special tools, like a hacksaw. ;-)

I dug through my pile of adapters that I have gathered for this project. I used the 1/2" to 1/4" hex adapter from my Earth Explorer. I used the same screwdriver bit that I used before. I held everything down and applied 24V. It cranked over and sat there purring away. Sounded like about a fast idle. So that should be fast enough to start. If it was at the top of the compression cycle it would hesitate when power was applied, but took off spinning after it passed top-dead-center. So it looks like getting this thing started will be no problem.

Next I connected the motor to my old drill. It has a replacement chuck that is 1/2" instead of 3/8". I hooked it to the 5 ohm load and started the drill. This old drill seems to rum much slower than my "real" drill. The label on the drill is worn off so I don't know how fast they claim it is. It was putting out around 5 volts which would have been 1 amp.

So I dug out some pulleys (and used bigger pitch pulleys this time) and put one on the motor and one on the drill. It was still slipping at full speed. But I was getting around 10 volts and 2 amps. So I need to get a better mechanical coupling and run some more experiments. Maybe I will chuck it into my drill press. It has various pulleys for adjusting the speed.

So I will get back with more results. Once I get the electrical part figured out I will have to tackle the mechanical part...

Subject: More numbers (Was: Re: More money down the drain???)
Date: Mon, 19 Feb 2001 20:43:09 -0600
From: jsampson@pobox.com
To: tcrobots@orbis.net

I chucked the new motor into my drill press. The drill press has 5 speeds. The fastest speed is claimed to be 3000 RPM. I hooked up the load and ran some tests. Too bad I didn't know the "real" speed it was running...

DOH!!! The motor has a tachometer generator on it!!!

So I wrote down the tach voltages and calculated the actual RPM.

So with the drill press set to 3000 RPM:

No Load, motor = 22.49V, tach = 47V / .014 = 3357 RPM, power out = 0 Watts
10 Ohm, motor = 20.4V, tach = 45.7V / .014 = 3264 RPM, power out = 41.6 Watts
5 Ohm, motor = 18.58V, tach = 44.3V / .014 = 3164 RPM, power out = 69 Watts

Then I wondered how fast the DC motor is supposed to run when used as a motor. So I hooked it up to different battery voltages and measured the tach output voltage.

@12V, tach = 24.6V / .014 = 1760 RPM
@24V, tach = 50.2V / .014 = 3600 RPM
@36V, tach = 74.5V / .014 = 5321 RPM

So it will turn at 5300 RPM without flying apart... So I was thinking about using a 1.5:1 pulley ratio so the motor turns faster than the engine. The pulleys I have will give me a 1.33:1 ratio. So 4000 RPM on the engine will give me 5320 RPM on the motor. I think the manual for the weed whacker says it has a 7000 RPM max speed. I'm also starting to think the rule of thumb that Alan and William pointed out "a DC motor puts out max power at 50% of its rated speed" may apply to using it as a generator also. So maybe I should expect to get usable current at 50% voltage when turning the motor at max RPM????

I could make some predictions based on these numbers. I'm hoping to get 3 amps at 30V when I get the faster speed, which is 90 watts.

So I will try to get the engine and motor hooked together this week. Then I can see if it really works. Then I'll try some more experiments.

I guess then I will have to make the "TCRG - Core Technology - Gasoline Engine Controller Module". ;-) I will probably need an R/C servo to run the throttle. (I also need to build a mechanical governor) I need a big MOSFET or relay to start the engine. I can measure the voltages and currents it produces. I should measure the engine temp to know if the choke needs to be set or not. So let's see, what's the "Kilian Command" to say start engine, charge battery then shut down??? ;-) Oh, I need a "you are low on gas" status message too.

Alan bored out some pulleys for me at "Kilian's Robot Shop of Horrors". You can see pictures of the engine and pulleys here.

Now I need to get the engine and motor bolted down so I can get them to run together...

Previous | Index | Next

Visit Twin Cities Robotics Group
Back to my Home Page http://www.pobox.com/~jsampson

This page is currently maintained by Jeff Sampson