Dual Rotor Toroid Core PMA build

[CraigM]

Found this on Temco Industrial web site concerning magnet wire insulation types. I feel better about the Polyurethane molding I'm using since it's also used as magnet wire insulation.

https://temcoindustrial.com/product-guides/wire-cable-and-accessories/magnet-wire/magnet-wire-faq

Here are some common types of insulation and their specifications:

Insulation Type            Thermal Class                    AWG Wire Size Range
Polyurethane            120°C, 130°C, or 155°C    20-18
Polyester                    >155°C                            20-14
Polyester-imide            180°C                               38-18
Polyamideimide            220°C                               38-14

I've seen photos of burned out coils in axial flux type stators. Does the burn out happen when the coil temperature exceeds the insulation temp rating, which in turn breaks down the insulation and causes a short within the coil?

Thanks
CM

[SparWeb]

Basically, yes, that is the reason the lacquer on "magnet" wire has a temperature rating.
That's not the only possible cause of burn-out, but certainly a common one.

(...wonder how he knows...?)

[MagnetJuice]

I've seen photos of burned out coils in axial flux type stators. Does the burn out happen when the coil temperature exceeds the insulation temp rating, which in turn breaks down the insulation and causes a short within the coil?

Insulation breakdown and failure is always the cause of stators burnouts. When the insulation fails, the wires shorts out. Overheating of the wire because of excessive current flow through the coil is one of the main causes of this breakdown. However, many other factors contribute to insulation breakdown.

Other factors that contribute to the deterioration and breakdown of the wire insulation are moisture, impurities on the wire, wild temperature swings and vibration within the coil.

Turbines that are powered by the wind are going to have drastic temperature swings, from very cold to very warm. There is nothing that we can do to eliminate those temperature swings. However, there are good practices that can be implemented during the winding of the coils that can minimize the possibilities of stator burnout.

Make sure that the wire is clean and dry before winding the coils. Wearing clean gloves during winding eliminates the possibility of oil from the hands getting on the coils. Superglue (Cyanoacrylate) should not be used on the coils. That type of glue breaks down over time because it is not designed to resist high temperatures. Superglue will become chemical impurities after the high temperatures breaks it down and stays within the coil. Regular superglue breaks down at around 100C, and the high temperature superglue breaks down around 135C.

If the coils are not wound tight, moisture can get inside the coils. If those coils are to be enclosed in resin, the moisture remains trapped inside the coils and over time will contribute to the deterioration of the insulation.

The best way to avoid these problems is to make sure that the coils are tightly wound. If high temperature epoxy can be applied during the winding, that would be the best protection against moisture sipping inside the coil as well as minimizing damage from vibrations. There are a few high temperature epoxies suitable for this application. MG Chemicals 832HT is one of them. Tightly wound coils also contributes to the efficiency of the alternator.

If the coils are not going to be enclosed in resin, the exposed part of the coils should be coated with a thin layer of high temperature epoxy to keep moisture out and impurities from accumulating on the wire.

The alternating current flowing in the wire causes vibration within the coils. During every cycle, the AC current reverses the flow of electrons. That happens many times per second. If the coils are not wound tight, that causes a jerk in any loose wire and it will hit and rub the adjacent wire, weakening the insulation and with time contributes to its breakdown.

Some builders get in a hurry to wind their coils and don't pay much attention to these details. Following these good practices during winding of the coils can help to insure that the alternator lasts a long time.

Another cause of stator burnouts is the magnet rotor hitting the stator. Worn out or poor quality bearings with excessive play can cause the magnet rotor to rub the stator.

The magnet rotor can also rub the stator if the coils expand and increase in thickness because of high temperatures. If that happens and there is not sufficient gap between the magnet and the coils, the stator could get hit by the magnets.

For stators that are to be enclosed in resin, one way to help with the cooling is to drill through the center hole of the coils after the stator is cast. That will allow air to flow and remove some of the heat.

I hope this is helpful.

Ed

[CraigM]

Hello Ed (MagnetJuice),

Yes, that was extremely helpful, thank you for providing the information.

I researched MG Chemicals 832HT and see how this epoxy is purposely manufactured for high heat applications and commonly used for potting electronic boards.

Once I have my single coil test complete and I have a better idea of wire size and number of turns required I'll be sure to use this during final coil winding. Probably will need to apply with a brush as each coil is wound.

I ordered a laser tachometer and currently working out how to spin the PMA at a desired RPM using a pulley and belt system on my cheapo' drill press.

Looking back I can see I should have used the 832HT when building the steel core. The core was tacked together with CA glue when winding and then I applied several layers of thinned epoxy to the surface of the core. The epoxy wicked into the core and after several applications if no longer wicked and started to pool on the surface. The epoxy I used has a service temperature of 300F but will that be high enough? Time will tell, experimental stamp on this.

Thank you,
CM

[CraigM]

SparWeb,

Thank you also for your reply. You have consistently provided valuable information in my effort to make something of this build.

Much appreciation!
CM

[CraigM]

Performed an open voltage test today. Please review and let me know if I did this correctly. Not really sure what this is telling me... if I did this correctly am I seeing close to cut in voltage for 24 volt system using 20 AWG and close to cut in voltage for a 12 volt system using 15 AWG? Also not sure what I want/need for RPM cut in, is 190 RPM a bit low?

What other information do I need? Assuming I need to know wire resistance (Ohms).

Single coil test – 190 RPM

20 AWG, 22 turns;
1.225 VAC X 8 coils in series = 9.8 VAC
9.8 VAC X 1.73 (star connected 1 WYE stator) = 16.954 VAC
16.954 VAC X 1.4 (AC RMS to DC conversion) = 23.74 VDC

17 AWG, 16 turns;
.895 VAC X 8 coils in series = 7.16 VAC
7.16 VAC X 1.73 (star connected 1 WYE stator) = 12.387 VAC
12.387 VAC X 1.4 (AC RMS to DC conversion) = 17.34 VDC

15 AWG, 12 turns;
.685 VAC X 8 coils in series = 5.48 VAC
5.48 VAC X 1.73 (star connected 1 WYE stator) = 9.48 VAC
9.48 VAC X 1.4 (AC RMS to DC conversion) = 13.27 VDC

Here's the test setup.



and useful chart I need to learn more about.



Thank you,
CM

[MagnetJuice]

I have to admit the electrical side of things is my weak point and what I need most to study and learn about.

If your weak point is the electrical side you are learning fast.

Your calculations are good with the exception of one small figure.

I am going to use the calculation with the 16 turns of 17 AWG wire.

Your calculations are good up to the 12.387 VAC.

The error you made is that the 12.387 VAC is not RMS voltage. It is Average voltage; therefore, it has to be multiplied by 1.57 to obtain your DC voltage. Then you have to subtract .7 volts per each rectifying diode. That will be a total of 1.4 volts for the two diodes on the bridge rectifier.

12.387 x 1.57 = 19.45 – 1.4 = 18 VDC

As far as deciding the cut-in RPM, first you have to decide what you are going to be using to power your alternator. VAWT or HAWT. I think SparWeb or someone with more experience in that area could help you with that better that I can.

By the way, that is a nice test setup.

Ed

[CraigM]

Ed,

Thank you for the reply. If I used a multimeter that reads true RMS then I'm okay with using 1.4 as the AC RMS to DC conversion? Then subtract 1.4 volts from DC volts for rectifier drop?



I repeated the test using a cheapo meter that doesn't read to as many decimal places and received a VAC reading of 1.0 to .9 on the 20 AWG coil, compared this with the VAC reading of 1.225 when using the true RMS meter. Believe this confirms the multimeter used in the first test does read true RMS.

I tried reading resistance (Ohms) with the good multimeter and with a Simpson analog meter I picked up cheap several years ago. I was getting readings that were not consistent and with not fully understanding what I was doing I didn't use this data.

I did research Ohms for magnet wire and found the following on several different websites;
15 AWG 3.18 ohms per 1000 feet. 3.18 / 1000 = .00318 ohms per foot
15 AWG .0104 ohms per meter
15 AWG using online calculator shows 6' of AWG 15 = .019 ohms. (each coil uses 6' of wire)

Circumference of coil is 6” (2” top, 2” bottom, .3125” radius half circle both ends is 1.96”.
15 AWG coil uses 12 turns x 6” per turn = 72 inches = 6 feet. .00318 x 6 = .019 ohms.
If using meters, 6 feet = 1.829 meters. 1.829 x .0104 = .019 ohms.
Online calculator showed .019 ohms per 6' of AWG 15.

So I feel pretty confident that each coil has a resistance of .019 ohms. Multiply this by 24 coils and I get .456 ohms. Checked this with online calculator for 144 feet (6' per coil x 24 coils) of 15 AWG and it gave a result of .458 ohms... close enough.

So now I know VAC, VDC and ohms per coil and for all 24 coils. A little extra wire will be needed for connections but I'm still in close enough mode.

Found a calculator online that will now give me amperage and watts. This looks really high to me... I was expecting watts to be much lower at cut in voltage.

Here's the calculator using 13.27 volts and .458 ohms... oops forgot, need to subtract 1.4 volts for rectifier loss... that gives me 11.87 volts. Only thing I'm not sure of is if I should enter VAC or VDC into the calculator.... thoughts on this?



My thoughts and/or questions on this testing.

• I'm seeing nearly 26 amps at 190 rpm, believe max amps for 15 AWG is around 28. Wire gauge seems to be too small.
• Using data I have can I calculate voltage if adding a second layer of winding? If 12 turns of 15 AWG = 11.87 VDC will 24 turns give me 23.74 VDC? This will also lower cut in speed for 12 volt system or will approach cut in speed for 24 volt system.
• For a HAWT how do I match blade size & tip speed ratio to this alternator?
• For a VAWT it appears I can bring cut in to a lower RPM and still use direct drive.

I'm open to anyone's thoughts. As I mentioned earlier I'm not in a good wind location and I'm not interested in spending the time and resources to put a HAWT 100 feet in the air. If I can feed a bit of wattage into my shop and run a smallish 24 volt system with a cheap inverter I'll enjoy the journey along the way and feel a bit of pride in the accomplishment.

Thank you much,
CM

[MagnetJuice]

I'll attempt to answer some of your questions.

The voltages that you measured with the True RMS meter are probably correct. It was after midnight last night when I posted and I was just thinking of arriving at the voltages using an equation. So go with what your meter tells you.

A regular multimeter, even the good quality ones, cannot accurately measure very low resistances. For that you would need a milliohmeter. Average price for one is about $1000. I have an accurate milliohmeter that is good to measure from .1 to about 3 ohms. I built it myself using a small wallwart and a small constant voltage, constant current module. If you don't have a milliohmeter, then using those tables where they give the ohms per 1000 feet is good enough.

Now, concerning the wire size for the coils. With 3-phase alternators current flows through each phase for only 66% of the time. If total current is 30 amps, current through one phase is 30/1.73 = 17.3 Amps. That means that 15 AWG wire would be OK if the total Amps output of the alternator is about 50 Amps. You could see over 50 Amps if a hurricane hit Pensacola. 

If you don't have good winds in your area, it would be pointless to go with the big expense of a tower for a HAWT. And a VAWT would turn slow if it is a Savonius so you would have to use a lot of turns on your coils to produce good voltage at low RPM if it is going to be direct drive. A good thing is that if you increase your voltage, you can use a thinner wire, for example 17 AWG.

By seeing how meticulous you are in your construction and design of this system, I am confident that you will end up with a nice working windmill.

Ed

[CraigM]

Hello Ed,

Thanks again for the help and direction you provide. It's greatly appreciated.

I spent the past few days researching HAWT blade design and tip speed ratio. Found that Hugh Piggott and if I remember correctly the “Dans” like a TSR of around 7. Also remembered seeing quite a bit of discussion about GOE 222, constant chord, no twist no taper blades. TSR for GOE 222 appears to be best in the 5-6 range. Chris Olson used these on most of his machines and strongly favored them... but then Chris had a strong opinion about many things... I miss Chris, Flux and many others, and the lively discussion from years ago, but nothing stays the same. Solar took over on price and the interest in wind power has waned.

So I wanted to relate TSR to wind speed and to alternator cut in speed. On Hugh's first wind turbine recipe book there is this formula for calculating TSR;
~ Rpm = windspeed (in m/s) x tsr x 60 / circumference.
My best guess at blade diameter to match my alternator is 2.4 meters or 8 feet. Using Hugh's formula to determine cut in speed at first useful wind speed of 7 mph (or 3 m/s) I get the following.
~ 3 (m/s) x 7 (tsr) x 60 / 7.539 (2.4 x 3.14 = 7.539) = 167 rpm
To get to 190 rpm... this is where I reach 24 volt cut in with two layers, 24 turns of 15 AWG I need to bump the m/s wind speed to 3.5 or 8 mph. So I may lose ever so slightly in the lowest wind by needing 190 rpm for cut in. And just for comparison in a 10 m/s (22 mph) wind the formula gives me 557 rpm.

I ran the same formula for a GOE 222 blade and used 6 for the TSR. They may be able to run at 5 without stalling but I have no hands on experience to validate this.
~ 3 (m/s) x 6 (tsr) x 60 / 7.539 (2.4 x 3.14 = 7.539) = 143 rpm
~ 10 (m/s) x 6 (tsr) x 60 / 7.539 (2.4 x 3.14 = 7.539) = 477 rpm
My understanding of GOE 222 is that they deliver greater torque and work well with a stiff alternator. With no experience I can't say if my alternator will be stiff or not. Will there be magnetic drag on the core? I don't know. The ferrite magnets don't feel strong enough to produce much drag. Anyone have an opinion on this? Chris? 

So now I'm at a spot of what gauge wire to use, how many winds and even if I should be moving forward with a 24 volt or 12 volt alternator.

When using 15 AWG wire, 24 turns (two full layers) I get 25.10 volts including 1.4 volt rectifier drop at 190 rpm. Using the same wire size and turns I get 12.6 volts at 90 rpm. So I'm a little high on rpm at 24 volts and will not cut in until 8 mph, 3.5 m/s winds and quite low rpm for 12 volts.

When using 17 AWG wire, 32 turns (two full layers) I reach 26 volts at 142 rpm which is just right for 8 foot GOE 222 blades and 12.8 volts at 70 rpm. Almost feels like I could use the alternator to run a HAWT at 24 volts or a VAWT at 12 volts.

So with regards to wire gauge, number of turns and tsr I just need to make a decision... $hit or get off the pot is what I hear my father saying.


Now for something that is frustrating me and can't seem to understand. How do I determine amperage from my voltage testing? Do I need to test with a load? Sparweb mentioned doing this. Sparweb? Using a formula from the chart posted earlier I can determine a number but what does it represent? Amperage = volts / ohms, looks simple enough. I have a voltage reading at a certain rpm and I know resistance, ohms for the total length of wire that will be used.
~ Voltage is 25.1 at 190 rpm, resistance is .917 ohms at 288 feet of wire = 27.37 amps which is 687 watts.

Where are my wheels falling off the bus? Using another handy formula in Hugh Piggotts recipe book he says blade power in the wind equals .15 x rotor diameter (squared 2) x windspeed (squared 3). This gives me .15 x 6 (2.4 meter squared 2) x 42.9 (3.5 m/s squared 3) = 39 watts.

If I only have 39 watts available with 2.4 meter blades at 3.5 m/s wind speed but I can get 190 rpm at 3.5 m/s at tsr 7 how is it the calculation above shows 687 watts?

I know a battery clamps alternator voltage down to battery voltage so is 687 watts the practical maximum available at .917 ohms resistance?

Voltage x amps = watts so at cut in with 25.1 volts I should see amps of around 1.5 to equal the 39 watts available to the blades.

Can I measure / plot volts and amps along a rpm range using only volts and resistance I now have?

Many thanks to all. It's now 5:11 am and I've been up for awhile... couldn't sleep from thinking of this.... way too much at times.
CM

[SparWeb]

Chris,
Very busy at work trying to win a bid for the coolest project I've ever done - and may ever do in my career.
I have made you wait for a response. 
Maybe in the meantime you've figured it out - or just done your business and got off the pot, as they say.

The balance between blades and alternator comes from mechanical power.  The only connection they make is the steel shaft between them, which transmits only torque at some speed or other.  As you have gathered, this pinch point in the variables makes it possible to solve the problem of matching blades to alternators, but it forces you to solve BOTH systems before you can see where the curve for one crosses the curve for the other.

So whether the blades are GOE222 or NACA0012, the power in the blade disk is generally about the same.  The airfoil choice alters subtle things, like the angle as-carved to get the needed angle of attack, and the torque to get started when they aren't turning.  There are some big things affected by airfoil, such as the L/D ratio, but when talking about un-twisted blades that doesn't matter because there is only a small portion of the blade running at max L/D.  The rest of the blade span is way off the ideal angle and the root is nearly stalled.

So to get rough matching, just use (power in) = (power out)
Power In = aerodynamic power that the blades can sustain a given RPM
Power Out = Mechanical power required to drive the alternator at a given RPM

You've got what you need to work out the alternator side.  The blades power can be worked out approximately enough now, too.

Number crunching will take some time for me to get into.  I might prefer to just open up the calculations I did years ago and run them with your test numbers.
I've got errands to do this morning but I'll try to get back soon. 

[CraigM]

Sparweb,

Thank you for your input, as always, GREATLY appreciated!

Good luck with your bid, sounds like a very exciting time for you. Focus yourself on the task at hand.

I'll review your original calculations and see what I can come up with.

Thanks again, don't over concern yourself with getting back right away. No hurry on this end.

Take care and win that bid!
CM

[CraigM]

WOW! It finally sunk in, the answer was right in front of my face and I just couldn't see it. If it was a snake it would have bit me.

I knew battery voltage clamped or limited alternator output voltage but it wasn't until I thought of voltage as pressure that it made sense. A 24 volt battery system wants to push electrons out at around 26ish volts. The alternator is trying to push more electrons into the battery and has its own pressure / voltage. Not until I exceed the push back pressure of the battery with a greater pressure / voltage from the alternator will a single electron move into that battery.

So with my test I'm getting 25 volts at 190 rpm using two rows (24 turns) of 15 AWG wire. Just reaching enough voltage to get that first electron pushed into a 24 volt battery. As rpm increases I get higher voltage but at this point I can now use ohms law to say Amperes = Volts / Ohms.

I did a voltage test at 340 rpm and reached 47 volts including 1.4 volt subtracted for rectifier. I subtract 26 volts (battery pressure ) from 47 and get 21 volts. 21 Volts / .917 Ohms = 22.9 Amperes. 22.9 Amperes x 21 Volts = 480 Watts.

Now I have something where I can determine wattage across the entire rpm range. I get .139 volts per one rpm. I can use wind speed and TSR to determine blade rpm and plot watts out per  wind speed.

Sparweb – thanks so much for your help! I went back to your first calculations and this is what turned on the light bulb in my head... I was feeling pretty dim witted up until last night when it finally flickered. Just for fun I'm going to go back into the math so I can determine Tesla and Weber units as well.

I learned something new and it feels good. Much appreciation to the board and all who contribute.

Regards,
CM

[SparWeb]

That great feeling with the "coin drops"

Don't forget to be careful with the phase-to-phase resistance and the phase-to-phase voltage as you work up from 1 phase to 3.
The 3 legs are out of phase with each other, and the resistances don't care about phase.
So voltage across 2 legs in Star is 1.73 x Vac in one leg, but resistance is 2xR in one leg.

[CraigM]

I've been working on the math for my project in an effort to see what potential this PMA holds. The following slides are screen shots from my spreadsheet calculations.

I'm not one to spend much time with math but this exercise really interested me. I can now see how calculating the given information can help fine tune the design before you build.

Through a bit of research and reading of older posts in this forum I've found the DC voltage when ran through a full wave bridge rectifier still has a bit of a bumpy wave form. Peak DC voltage of that wave form is considered 1.4 times the RMS AC voltage and the average is closer to 1.35 times RMS AC voltage. I used 1.4 to determine initial cut in as the battery will see a bit of a pulsed current and 1.35 to determine anything above cut in.

I ran my calculations for a 24 volt system as well as a 12 volt system.

I guess I'm surprised at the total watts available in the 500 rpm range and the difference in total watts from a 24 volt system to a 12 volt system.

I understand there are other losses that I'm not accounting for such as heat and core losses but I don't know how to measure these at this time.

Cut in for 15 AWG at 24V seems a bit too high at nearly 8mph wind and tsr of 7. If my blades only run at 6 tsr it would make cut in rpm even higher. At 12V there is no problem with low wind cut in speed.

Cut in for 17 AWG at 24V works well in low 6mph winds but I'm concerned about it not being able to carry as much current and heating up quickly. At 12V, again, there is no problem with low wind cut in speed.

Feeling like 16 AWG would be “just right”.

Initial impression is I'm happy with the output and feel this PMA would be comfortable at 500 watts continuous but is it really capable of producing 1000 watts?

Another part of this build I'm impressed with is the low cost for magnets and magnet wire. Compare the cost of 24 neo magnets at $150 range to 16 ceramic magnets at $32 is a win. I'm also seeing less wire will be used. Believe the traditional 24 pole, 9 coil, air gap axial flux uses around 6 pounds of wire. This PMA will use around 3 pounds of wire. Of course there is the problem of finding magnet steel that can be wound as a core. Maybe I got lucky with this.

I also backed into the math to find the tesla strength in the core. Hugh Piggot talks about Proven turbines of this same design being a flux multiplier. I believe I'm seeing this after calculating 2.91 T in the core. I'm seeing nearly 10 times the tesla strength of ferrite magnets. For comparison neodymium magnets are around 1.4 T.

Believe next step is to tear down the current test coils and do one more test coil with 16 AWG wire at around 26 turns. Hope this is the sweet spot I'm looking for.

Hmmm, suppose I could do the math and get a good idea as well.

15 AWG - 24 volt


15 AWG - 12 volt


17 AWG - 24 volt


17 AWG - 12 volt


Single coil test at 190 rpm.


Tesla calculation at core.


Thanks much!
CM

[SparWeb]

You may find this interesting:
https://www.fieldlines.com/index.php/topic,149585.0.html

I did carve my blades for TSR=7 but they don't operate at TSR=7.  Their TSR varies under different conditions.  Below cut-in speed, they run much higher TSR.  At cut-in, the slope of the line changes and levels off.
As the discussion in that thread explains, my blades are also under-sized for the generator, so they don't get to run at optimal TSR.

Just something to consider when you write: "...Cut in for 15 AWG at 24V seems a bit too high at nearly 8mph wind and tsr of 7. If my blades only run at 6 tsr it would make cut in rpm even higher..."

[CraigM]

Thanks Sparweb!

Good real world information in the link you shared.

Doing this sometimes feels like I'm trying to put a puzzle together without having the box to see what the picture looks like... and no one else knows what the picture is as well... it's created as you fit each piece.

You try to fit a piece in here and there... some fit, some don't. But when you find a piece that fits the picture becomes clearer and you have those Ah Ha! moments.

Blade (power) matching to alternator (load) matching will take some practice.

But just like a puzzle I keep coming back, rethinking it, working it and enjoying the process.

Thanks again,
CM

[SparWeb]

Blade (power) matching to alternator (load) matching will take some practice.

Indeed it does.  I still haven't got it right, yet.  This is the Nth change to my machine and I'm still closing in on the perfect match.

[MagnetJuice]

Craig, nice winds in your area. It would be nice if you had your turbine up and running.

Ed

[CraigM]

MagnetJuice,

Yes, wind was good when the tropical depression was moving through but that's about the only time around here. If I'm honest with myself a wind turbine just doesn't make any sense where I live... so my progress with this project has slowed considerably.

I also get distracted with other interests and bounce around a lot. I decided I want to understand electronic circuits much better so I dug out my old copy of Forrest Mims book and may try my hand with an Arduino project.

Not sure if Ghurd is still selling his LVD / Controller kits but would like to sharpen up my soldering skills too.

If anything worthwhile comes up I'll post a “just for fun” excerpt.

Thanks CM

[SparWeb]

You can ask ghurd:   http://ghurd.info/

Tell him I said "hi".