Coil Tests

[phil b]

Since I still have my genny down for repairs, I thought I'd try to answer some of my own questions about differences in basic coil configurations. My first two good machines have been wired for 12 volts. This time I'm shooting for a 24 volt machine . I made a contraption to fit into my new drill press that will spin the rotors at 300 rpm. Unfortunately, that's the lowest setting.

I have one test coil of AWG #18 wire, 60 winds. It's mounted on 1/4" plywood and held on with red duck tape.

The first test was conducted with 2) 12" rotors with 12 2"X1"X0.5" magnets. The coil legs did not match the magnets exactly.

In the second test, I used 24 wedge shaped magnets, single rotor, mounted on top of the rotor as the 12 rectangular magnets. Also, I held some laminated steel from vacume motors I had on hand over the center of the coil, then doubled it to see what would occur. That's where "no steel, double steel is from in the chart. Here are the results.

rpm     winds   volts          

300.     60    5.7        no steel    double rotor    2x1x.5 rectangle
     
300.     60    5.4        double steel    single rotor    3/16" wedge
     
300.     60    5.3        single steel    single rotor    3/16" wedge
     
300.     60    3.3        no steel    single rotor    3/16" wedge
     
     

The results are probably slightly scewed because I put the wedges on the back of the rotor with the rectangles. It still tends to say the steel laminates are making a big difference in voltage. More tests are needed. While it's down, is there any tests that you would like to see?

[RP]

Can we get a closeup of the laminates and can you clarify the dimensions of the laminate pieces compared to the coil and magnets?  The picture seems to show the laminates not fully covering the coil leg but if they do fully cover the magnets then I suppose it might not matter.

Thanks

[phil b]

Hi RP,

The dimensions of the laminates are 1"X3". There are 45 pieces that make up the laminate. When laid across the coil, they cover both legs from side to side and only 1/2 the length of the leg. Two of the laminates do completely cover the coil's length and width. I would have thought that the voltage would be higher with complete coverage. It seems to have little additional effect.

The coils are 1/4" thick. They are wound using a rectangular 2x1x0.5 magnet. Preey much standard.

[RP]

Phil,

Thanks for the new picture and dimensions.  Huh, I too would've expected it to make a more substantial difference having the laminate cover more of the coil.

I suppose one other test would be to see if the one or two laminates makes a difference in current into a resistive load.  I'm not sure why it would though since if the volts are there, the only other parameter should be the coil resistance which is constant regardless of the laminates.

BTW:  I really like the CV joint as a self aligning shaft.

[phil b]

I knew that old CV joint would come in handy someday!

I intend to do more work as time permits.

[Tom in NH]

Hi Phil,

Nice tests and photos. Are you able to vary the distance between your dual rotors? I'd like to see how changing the distance affects voltage. -tom

[phil b]

Hi Tom, I can vary the distance. I'll try that this weekend and post the data.

[Gary D]

Phil, with the 1x2x.5 magnet duel rotor setup, most use a half inch thick coil. If you did, you could have twice the turns and twice the voltage on the test coil (or two in hand with the same voltage) if you have the air gap set up to the 3/4 or inch most use. You aren't using the available power of the thicker magnets to their potential(even for a test coil). It would be interesting to see how far away you could jack the test coil from the 24 wedge magnets and still keep the voltage close to what you have now. That would give you an idea of what guage wire and turns each setup would require, and be able to produce. Nice tests as far as they went though!. Many do no test coil at all even when changing proven designs significantly.

 Windstuff Ed had the best results with coils imbedded in slotted/drilled laminates, although it was very labor intensive. That was a few years ago. I could look for a link to some of his tests, if you'd like. Just a thought...  Gary D

[phil b]

Hi Gary,

This is an experiment to change the configuration as many times as I can think of and record the outcome. Who knows what the results will be. I'm looking to the folks here for more ideas I may not have thought of.

You are right, 3/4" or even 1" air gap may not unreasonable with 1/2" magnets. If I double the number of winds, will the heat in the stator go up? by how much? How much resistance will hold the stator's temp down to reasonable levels?  etc...etc.  That's some of the questions I'm interested in answering. I plan to publish the results here as time permits.

Thanks for your input!

Phil

[Flux]

Phil, I can't work out what you are trying to do.

There are so many variables that you are creating that I can't see how you hope to sort out an answer.

Finding voltages for different configurations is one thing. Your different configurations will let you use different sizes of wire to fill the space in each case and the resistances will be very different. Resistance has as much bearing on this as voltage.

If you double the number of turns, what will happen depends on many things. If the gap is big enough to just double the number of turns, then the first coil is far from optimum.  You would double the voltage and double the resistance. If you compared like with like, your coil with half the turns could have been wound with twice csa an 1/4 resistance.

Single rotor tests will be of little use if the coil is very thick as the flux will not penetrate many of the turns, so conclusions based on a single rotor would give no idea of what it may do as a dual rotor.

The aim should be to find the best compromise between induced volts and resistance.

The requirements are affected a lot by how you intend to load it, so conclusions for stall limited battery charging will be different from heating or battery charging with mppt.

What you will eventually prove if you can sort all the variables is the maximum power you can expect from a given volume of magnet and a given speed. The configuration to do it will depend on the shape of magnet. Thin magnets will need short air gaps and thick ones will need longer gaps. You should find that the best gap is between 1 & 2 magnet thickness.

I think you can learn a lot if you keep the goals in mind, otherwise you are going to have a lot of data with no practical application.

Flux

[finnsawyer]

Flux, when I looked at this yesterday all I saw was confusion.  I think he would have done everybody a greater favor if he would have provided the following information about the actual alternator:

    A.  Complete specifications for the magnets.

    B.  Wire size, number of turns and dimensions of the coils.

    C.  Complete physical specifications for the construction of the alternator.

    D.  The resistance per phase.

    E.  The voltage out at one or more values of RPM.

This information would be of more use for someone looking for a design to build.  While it would be useful to know the behavior of the alternator for different parameters, such as a different stator, air gap, or dual rotor versus single rotor, that would probably be too much to ask from an individual.  On the other hand, if members of the community that build alternators would provide this information for their alternators, a data base could be built up that would aid someone seeking to build an alternator for their own use.  

Recently our hosts asked for suggestions for a new poll.  Well, here's one.  Invite site users to provide the above information for their creations.  Tabulate the results and put it in a FAQs section.  More work for the editors, but I don't mind giving other people more work.  He He.

[Tom in NH]

Phil,

I can't help but feel that a big wet blanket has been thrown on your tests. I hope you won't let that discourage you. I am one who will be interested in seeing how the distance between rotors affects voltage in a coil. I hope you can take half a dozen or so different readings, keeping all your variables the same except the distance between the magnets. BTW how thick is your test coil? --tom

[phil b]

No wet blanket for me.

Flux and all, the reason that I'm doing different tests is simply to create variables to that might work for a new unit.

What type of unit? One that is more than 50% efficient, thereby avoiding heat buildup in the stator and possible one that that can have multiple voltage outputs but mostly, just thinking out of the box.

All the data is being entered into a spreadsheet to determine what's the best trade-offs. All of the variables will probably change with the final machine.

I had the electric motor laminations laying around and thought... I wonder if... That's why I published this tidbit. I was surprised at seeing almost the same voltage from a single rotor versus a double rotor. Maybe this is leading to more confusion because the data is not being presented correctly.

 I do have several items that are constant. The wire is 18 AWG, 300 RPM, 2 magnet sizes because that's what's on hand. If I go by the charts for amps and resistance, I hope to reach a good compromise, which will probably call for more tests(of course).

The tests have been a lot of fun. I thought I would give other interested folks a chance to see results from their questions.

[Flux]

Please don't think I was discouraging you, the thing to do is to keep a logical approach and consider each step in its own context.

Using the same coil with different air gaps on a dual rotor is very useful, it shows how the flux density changes with gap.

Trying to compare the same thing with a single rotor and a single rotor with laminates would add confusion. Using the same coil at different distances on a single rotor would be useful. Similarly coils with the same numbers of turns but different thickness with the same clearance from a single rotor would again be useful.

As long as each step is an investigation into one thing and the results are not mixed with something different then the results will be of value.

At best all you can do is present the information in a logical way. The conclusions you draw from the tests may not be the same as someone else.For your spreadsheet, try to keep the data for each type of configuration separate.

Flux