48V Direct Drive MPPT Turbine

[ChrisOlson]

Not many people building wind turbines anymore it seems.  And I'm too busy to post every one of my projects.  But I'm continually improving my designs and coming up with ways to get more power and better efficiency from 3.5 meter class machines.

When we recently switched our system to 48 volts the solar panels were no problem.  I was already running the panels at 90 volts.  So I just had to reconfigure the Classic for 48 volt battery and it "just works".

However, the wind turbines were another matter.  They are ferrite magnet geared 3.5 meter machines and they cut in at 30 volts.  Initially I just let them spin harder and worked up a new power curve in the Classic and they worked.  But not very efficient in the lower wind speeds because the turbines are running too fast.  So I "fixed" that by pulling the ferrite magnet rotors off and replacing them with neo magnet rotors with the same stators.  That instantly switched them 60 volt cut-in and they work fine in lower winds.

But one of the problems I ran into with 48 volts is that I can't run the turbine fast enough at the top end to get decent efficiency from it.  With a 24 volt turbine I can cut it in at 100 rpm @ 6 mph (TSR 7.0) and 30 volts.  At the top end I can spin it up to 370 rpm @ 25 mph wind speed (TSR 6.0) and it makes 111 open volts.  I can "clamp" that to whatever is necessary to get 90+% generator efficiency with the Classic's power curve for that turbine.

With 48 volt and the same rotor, if I cut in at 60 volts my open voltage at 370 rpm is 222 volts.  There's no way I can "clamp" that down to ~145 to keep it below the Classic's 150 volt input limit.  So the result is that the turbine has to be run in "stall" at about TSR 4-4.5 and it's very inefficient in higher wind speeds - approaching the horrendous inefficiency of most of the homebrew book designs.

So I set out on a mission to design a new direct drive turbine designed for the Classic 150 controller and a 48 volt system.  I always start out designing a new turbine with CAD (Cardboard Assisted Drawing):



This is a 20 pole, 15 coil three-phase and the generator diameter is only 13".  It will use 2 x 1 x .5" thick N42 neo magnets.  It will also use wedge coils like the rest of my generator designs, and will also incorporate "pole cramming", i.e. jamming as many generator poles on that size rotor as is practical and still be able to fit the required copper in the stator in the size and space available.

It will cut in at 60 volts and 100 rpm.  However, at the top end I need around 130 open volts instead of 222.  If you do the math on this, 130 x 1.732 = 225.  So this turbine will start out in wye configuration and when the input voltage gets to where the rotor has to start slowing down because the voltage is going too high, I will use the Classic's AUX 1 output with the PV V High option to activate the AUX 1 port and switch the stator to delta configuration.  I'll hold it activated by adjusting the "Hold" time to give the turbine a few seconds to adjust its speed.  This will let the rotor speed back up to ideal, drop the input voltage so it stays below 150, and I'll program a "bump" in the controller's power curve to make the transition smooth.  When the input voltage drops below the PV V High set point it will transition back to wye configuration.

According to all my wild scheming and figures this should produce a generator capable of 3.5 kW continuous output (58 amps into a 48 volt battery @ 60 volts) at >90% power efficiency running at full bore.
--
Chris

[tanner0441]

Hi Chris

Looks good. Do you have to program a hysteresis loop for the star/delta switching or does the controller stop it hunting at the set point itself?

I will follow the build with interest. I live in a valley with hills on two sides, one covered in large trees, and my house on a third side, the only time my turbine does anything near useful is when the wind comes straight down the ridge of the house, so for the last two months it has been shut down. I keep threatening to take it down and sell it.

I followed your gearbox build and was impressed that everything was considered right down to oil drag. Perhaps you should publish a wind power book based on your designs.  All the other books I have read cover pages of theory. You just fill it with "Designs That Work" Give you something to do on dark winter nights when you have done combining.

Brian.

[ChrisOlson]

No wind power books coming from here.  I like to design and build turbines, I don't like to write books.

The wye/delta switch can be made very reliably and smoothly with the Classic controller.  The main problem folks have had with wye/delta switching in the past was no reliable way to control it so the turbine ends up running in the wrong mode too much of the time. and the speed increase is quite radical when the switch is made.  All those problems are nicely taken care of in the Classic, since the high voltage input to the controller from the turbine is constantly monitored and precisely adjusted to make the turbine run at the right speed.

With the programmable delay and hold timers available on the Classic's AUX 1 output, it will be easy to program it so as to prevent nuisance switching.  Plus provide enough "hold" time when the switch is made so the turbine has the second or two that it takes to react to the rotor suddenly being allowed to spin up to get the voltage back up to the proper output level to match the power curve.

With the Classic it can all be adjusted simply by pushing some buttons.  And if you're really lazy it can be done from your PC over your home wireless network too.  Just yesterday I was in the office in the house looking at how the turbines and solar were doing, and my wife was in the kitchen.  I hollered out to her to look out the kitchen window at the one turbine there and said "Honey - watch this!".  Then I activated the voltage clipper from my laptop and it stopped the turbine.  My wife hollered back "What'd you do that for?  Turn it back on.  I got laundry to do!"  LOL!
--
Chris

[bob g]

Chris

i have a few questions, if i may

1. what is the airgap distance you usually use?  with neo's and with ceramic magnets

2. have you setup a test stand to determine the actual efficiency of your machines? or is the efficiency calculated?

3. if measured how did you do it? what apparatus did you use?

4. if calculated,  did you include the losses of the transmission gearing? the rectifier losses? and the losses in the mppt controller?  even if small they do add up?

trying to learn what i can about your processes and design philosophy.

thanks
bob  g

[ChrisOlson]

I usually run an air gap of around 20 mm with the type of magnets I'm using.  I do have a dynamometer that I run generators on sometimes, but haven't used it in a couple years.  I've found I can predict efficiency within a percent or two from measured on a dyno, so it doesn't pay to go thru all the setup and calibration to use it.

For generator efficiency, only the power efficiency of the generator itself is considered.  That's what determines if a generator is going to produce a rated output continuously, or burn up.  They'll only survive short duty cycles at the rated ampacity of the winding wire if it's only 65-70% efficient.  Wind turbines are different than anything else because the wind don't care if your turbine generator gets hot.  It's either built to survive or it will burn up because it's being driven by an engine that can be relentless.

When you look at overall efficiency, wind turbines are like automobiles.  At every step there's losses.  With one of my 3.5 meter machines you start out with about 11.1 kW of kinetic energy flowing thru the swept area of the rotor.  The rotor can capture 36% of that running at TSR 6, so you got about 4 kW at the shaft.  The generator can turn that 4 kW of shaft power into 3.6 kW of electricity.

Now you transmit 135 volt power down a 90 foot tower, then thru a rectifier with a 1.4V forward drop, then thru 260 feet of underground wire run to get to the utility room.  You end up with 143 watts in transmission losses.  So now you feed ~3,457 watts into a MPPT controller that's about 91% efficient and you end up with 50 amps @ 62 volts going to the battery, or about 3.1 kW.  So from the shaft to the battery you only get 77% efficiency.  And from the wind to the battery you only turn 28% of the total power available into useable power at the battery.

But that's still more efficient than a solar panel which can only convert 15-20% of the available energy to electricity.

However, unlike combustion engines, no matter what type of fuel you burn in them, the fuel for wind and solar power is totally free.  Absolutely zero energy input required to make it.  It's "just there" so the efficiency of a machine or device designed to capture it and convert it to electricity is non important if the machine can be built economically and produces enough electricity to meet your needs.
--
Chris

[tanner0441]

So power factor doesn't come into a wind turbine with MPPT control. Like a 1Kva generator with the PF of 0.8 is really only an 800W generator in real terms. As long as your windings are rated for 1Kw and with MPPT you get 1Kw, as long as there is a 1Kw or more in your wind conditions.

I used to have to explain to people with a couple of the boats I worked on that the cheaper 10Kva generator wouldn't cook a meal while someone stood under a 5-7Kw electric shower.

Brian

[Flux]

Power factor is not a big factor with air gap alternators, they have little inherent reactance so your power factor will be about 0.95 from the rectifier.

Star delta or series parallel changing should be fine with the classic as you can track each curve optimally, it is a neat way of overcoming the converter input voltage limit.

From the switching point of view star/ Jerry (IPR) is the easiest and can be done with a single contact, it is basically root 3 to 1 as for star delta. I look forward to the results.

Flux

[ChrisOlson]

Flux - yes, IRP is a much simpler switch.  I just happen to have three wye/delta switching boxes from back in the days when I used to use this on some turbines.  They are all pre-wired with 100 amp SquareD contactors and rectifiers already in them and enclosed in nice weatherproof boxes.  I just need to bolt them to the tower power pedestal, run the six turbine leads and 24V control signal into them to make them work.  That's the only reason I'm using wye/delta.  If I had to build from scratch I'd use wye/IRP for simplicity.
--
Chris

[ChrisOlson]

This is an explanation of the problem when using MPPT on wind power systems.  Both of these examples are real world tested, and both use the identical turbine with two different different generators on it:

This is the turbine power curve for one of my 3.5 meter machines with a 12 pole 1.14 ohm stator, as loaded into the controller:



This is the actual output numbers from this turbine:



This is the same turbine with a 16 pole .55 ohm stator - the Classic Power Curve for it



This is actual output numbers from it:



As you can see, the high resistance 12 pole stator is less efficient.  I get less output to the batteries because of the fact that it has to deliver 28 amps with 15 AWG windings to get 3.2 kW at the batteries.  The low resistance stator delivers the same 28 amps, but at higher voltage and with 13 AWG windings, so I get 3.6 kW at the batteries.

The generator efficiency is down to about 90% with the 16 pole with about 5.5% gearing loss.

Both of these power curves run the turbine rotor too slow on a 48 volt system.  When I initially tested the 12 pole it was designed to let the turbine run faster because it put out less volts/rpm than the 16 pole units.  I figured I could run it faster with the high resistance 12 pole generator.  But when you turn the rotor loose and let 'er really RUN, the shaft power increases faster than what the generator can handle and it gets VERY hot - hot enough to start to be able to see smoke coming off the stator up there on the tower in good wind.  I just about set that turbine on fire on a 90 foot tower about 5 days ago.  When the smoke started trailing behind it, the big red shutdown handle saved it before I lost the stator in it.

The 16 pole geared generators make the right number of volts for cut-in, but too many at the top end to get the turbine rotor up to at least TSR 6.  I can't switch those to delta because they only got three-lead stators.

So that's how the new direct drive MPPT turbine concept was born.  I don't need the gearing to get high voltage and low resistance.  I'm going to put more poles in and turn it slower to get the voltage.  And swap the stator to delta configuration at higher power outputs to get the low resistance and at the same time cut the voltage back to let the rotor spin.  I will easily get 3.5 kW output continuous, and I'm betting I can get intermittent spikes to 4-5 kW from a 3.5 meter turbine.
--
Chris

[ChrisOlson]

Haven't had a lot of time to work on this turbine this week.  I'm short on winding wire - have enough to wind one coil group on hand and have to get another spool.  But I did determine that I can fit a 20 pole 15 coil 150 volt direct drive generator under 13" diameter rotors.

This shows the relationship of a dummy mag to a coil.  It's going to take a little scrunchin and pokin and stuffin to get it to fit.  But it will.  Resistance will be just over an ohm in wye and about .36 in delta:



--
Chris

[gww]

You are not wasting any space.
gww

[just-doug]

it seems to me you have already out grown the classic 150.it is now a limiting factor.a classic in the 250 volt model would be a better match.55 amps gets you 13750 watts.you probley wont out grow it with your 12 footers .then the next limiting factor will probably be rotor rpm and longevity.

[ChrisOlson]

The Classic 150 can go to about 5 kW on a 48 volt system, which is fine.
--
Chris

[boB]

it seems to me you have already out grown the classic 150.it is now a limiting factor.a classic in the 250 volt model would be a better match.55 amps gets you 13750 watts.you probley wont out grow it with your 12 footers .then the next limiting factor will probably be rotor rpm and longevity.

13,750 watts at 55 amps is at 250 volts.   55 amps at that input voltage would translate down to
around 275 amps at 50 volts.  The Classic 250 is only good for around 55 to 60 amps at that voltage.

boB

[just-doug]

my bad.the amp limit is applyed to the out put .i forgot.it seems to me that the principal reason for this new design is to keep the rotor on a optimized tsr.it also occured to me that a different managment to the calassic might be helpful,but i dont know how to work the electronic end.if you had another electronic brain that could be fed wind speed and rpm,then enter rotor diameter,the box could calculate real tsr.this tsr then could be fed to the calssic and it would vary load to meet the tsr goal.watts out put would run wild .

[ChrisOlson]

I haven't posted progress on this machine because I don't have it perfected yet.  Turns out I could not use wye/delta switching the way I thought.  So I'm winding a different stator for it that will operate in wye only.

I have a few photos of the machine that I took when I was building it, and before I put it on the tower the first time to test it.  It is undergoing some revisions at present, although the first prototype is flying as we speak and has produced 14.4 kWh since midnight today.













--
Chris

[jlt]

That is a very impressive build.Love it and it is direct drive.

Not that the geared turbines were not good but they were too complicated for me to try.

[ChrisOlson]

Thanks jlt - the turbine I'm test flying right now tries to go to <200 volts if the rotor was running at the right speed.  I need to wind a different stator for it with less turns that runs the rotor faster than it should be running at cut-in so I can achieve ideal or near ideal blade tip speed at full rated power @ 25 mph.

The present iteration of the machine is producing 1 kW @ 20 mph with a 3.2 meter rotor on it, and it has hit spikes around 2.3 - 2.4 kW.  But the rotor is running too slow at wind speeds over 20 mph and until I get that fixed I hesitate to post too many details on it because it's not "right" for MPPT.

The challenge here is that if you cut in at ideal TSR you need to increase the rotor speed by 3.8x to maintain ideal TSR at full rated power.  That also increases the open voltage by 3.8x with a permanent magnet generator.  There is not enough room from a 48 volt cut-in to 150 volts max input to the controller.  On 24 volt it is no problem because the turbine can cut in at 30 volts and run 3.8x that at full rated power, which is well under 150V.

The bottom line is that on 48V the turbines don't perform anywhere near as well as they do on 24V at wind speeds over 20 mph.
--
Chris

[CraigM]

Chris,

You may have covered these questions on a previous post but locating them would probably be a challenge, so here goes.

What do you use to mount your rotors? Is it a quick disconnect hub/bushing that is welded directly to the rotor plate? If so is it difficult to get the hub exactly centered on the rotor plate to avoid an oscillating rotor?

Something like this?


Also what do you use to set the distance between the rotors? Will a spacer placed over the shaft work? Would shaft collars placed on each side of the rotor/hub assembly hold the rotors in place?

I'd also expect the use of a keyed shaft and what diameter does the shaft need to be for strength and reliability?

Last question, I promise… for now.  Do the bearing housings hold tapered roller bearings or double row angular ball bearings?

The reason I ask is because I recently started a new job working as a maintenance, repair and overhaul purchaser for a large chemical company. I purchase hundred of drive and motion components daily and can pick these items up at a good price. I'm building a small personal inventory of components so if all goes as planned I'll be able to start my wind turbine in the near future.

Thanks for your insight,
CM

[ChrisOlson]

The generator rotor hubs are 1" keyed bore Weasler Engineering weld-in type.  They are not split taper-lok hubs.  It is not hard for me to get them centered because I have a CNC Plasma table and I cut the rotors with it, including the hub bore.  I still machine them on the lathe for radial and axial runout accuracy because welding the rotor disc to the hub warps it some.  The lightening holes being cut in the rotor AFTER welding to the hub relieves stresses in the steel and makes it easy to straighten and true it on the lathe.

The rotors are shimmed to set air gap:



The shaft is machined from 4130 forged (heat formed) chrome-moly billet - same material high-performance crankshafts are made from for racing engines.  It is machined with a 1:10 taper on the nose for the hub, 1.125" diameter for the front bearing and 1.000" diameter for the PTO.  4130 CRM does not machine well, so it is finish ground and polished on my crankshaft grinder, then heat treated (hardened and tempered).  If you use regular old cold-rolled steel I would increase the shaft diameter to 1.250" but still use a 1.000" PTO.  I buy 4130 chrome-moly by the 20 foot length to make machine shafts, so I get a fairly good price on it.  If you try to buy it by the drop length you'll pay dearly for it.

The bearings are Swedish-made (Göteborg, Sweden) SKF, rated for 50% axial of max radial load, relube type, self-aligning, single row radial ball with extended inner race.  The shoulder of the PTO shaft seats against the rear main bearing so both bearings absorb axial loading.  Identical setup to the input shaft bearings in my geared turbines, except those use sealed non-relube type with pressed flanges.

Bearing life is determined by rotational speed and loading, and in this application the main bearings have a service life of 50,000 hours, or about 7 years on a good wind site for either the input shaft on my geared machines, or the mainshaft in this direct drive.  The PTO bearings in the geared machines have a service life of 90,000 hours due to the very light loading on those bearings, even though they rotate at much higher speed than the input/mainshaft bearings.


--
Chris

[CraigM]

Chris,

Thanks for sharing, always interesting to see sound engineering practice in use.

Your build is beyond what I can do at home but I now understand the technique and with the help of a local machine shop shouldn't be overly expensive to duplicate.

Great information if looking for an alternative to the trailer hub.

Thanks again,
Craig

[ChrisOlson]

The design of it should be able to be duplicated fairly easily by anybody with reasonable skills.  It's simple enough that extreme accuracy in making parts for it, and building it, is not required.  For instance, if you don't have a lathe to make a shaft just use a straight piece of steel and weld a thrust washer on it that seats against the rear bearing instead of machining a shoulder, and use hubs with bigger bores for the generator rotors that fits the bigger PTO.
--
Chris

[fabricator]

The only problem for mere mortals is the 20 ton press it takes to get that stator thin enough. 

[ChrisOlson]

That's not necessary either.  I got the air gap at 20mm and a different stator wound for it now that makes less voltage.  Still don't know if I have the ideal for 48 volt - have to test it.
--
Chris

[ChrisOlson]

The build for this turbine is up on my website.  There's way too many details to post here.
http://dairylandwindpower.us/projects/

For people posting projects here, like the one I put in Controls on heating water with a Classic 150, when all the images mysteriously disappear it doesn't help anybody in the future when they want to see how it's done.  So my projects are going on my own website from now on where the images won't disappear into thin air after awhile.
--
Chris

[kitestrings]

Chris,

Really, really nice build. A thing of beauty really.

While you may not be writing books, many of us are pleased you take the time for these writes ups & clips. Thanks for sharing.

~ks

[ChrisOlson]

I've had so many emails and phone calls on this machine that I'm going to be providing AutoCAD drawings of some of the parts on the build website.  People can download the AutoCAD drawings and put them on a memory stick and take them to a CNC machine shop where they can plug them into their CAD/CAM system and cut the parts.

Most people are worried about the hole placement on the generator rotors and stator support.  And while there is a fair amount of leeway there, attention has to be paid to accurate cutting with generator rotors for proper balance and such.  When the turbine rotor is on one end of a shaft and the generator on the other you don't want a couple generated due to improper dynamic balance, even though you have perfect static balance on both ends, that will cause a severe "wagging" on the tower.

Providing the drawings so the parts can be CNC cut will help insure that things are accurately made for the homebuilders out there that want to build one.

I've seen this new machine running at 3,100 watts and 52.6 amps to the battery @ 144 volts controller input with a 3.2 meter rotor on it.  It doesn't seem to have much advantage over the geared ones at lower wind speeds.  But once the wind gets over about 12 mph it makes slightly more power than the geared machines that I've built.  Mainly due to slightly better generator efficiency in this one and running at higher voltages than the geared ones are designed to run at, so less loss in the wiring and rectifier.  It's on its second stator revision right now.  For a 3.2 meter I could make the stator just a tiny bit more efficient by reducing the turns in it.  But the stator that's in it will be a good match for a 3.5 meter rotor.  As soon as I get my 3.5 meter rotor that I got in the shop rebuild and repainted it will be going on this machine for testing.  I expect to get close to 3.5 kW peak from it with the 3.5 meter rotor on it.

The generator in this turbine is a bit more noisy, though, compared to the geared machines.  The geared machines run pretty much totally silent except for blade noise.  This thing hums a tune on the tower all the time, due to operating at much lower current frequency than the geared generators for any given output.  Once this thing gets above about 200 watts it's fairly quiet except for a distinct hum from the generator.  But below 200 watts there's a couple places in the speed range where the harmonics in the winding reach a natural frequency that amplifies it and it's loud enough to be easily heard from 200 feet away.  When I get it down I have to look at installing a damper on the stator to stop that.
--
Chris

[fabricator]

That's just the sweet sound of harvesting your own power, although after an hour or two it prolly gets old.

[ChrisOlson]

That's just the sweet sound of harvesting your own power, although after an hour or two it prolly gets old.

There's one spot right between 90-115 watts where it gets pretty loud.  You'd swear there's an alien space ship landing because it's a wavering type pretty loud, ghostly sort of sound.

I'm pretty sure a ring of steel around the perimeter of the stator, damped with rubber, will reduce that noise considerably.  I have a picture of the damper I need for it in my head.  Just have to build it and put it on there.
--
Chris

[fabricator]

I know how that picture in the head thing works, sooner or later you're gonna be able to stick an electrode on your skull and download that right into solid works or autocad.

[ChrisOlson]

I prefer to leave it where it is.  Then I can use my built-in CAD/CAM system to build it direct from the software design.  No need to download it into a vastly inferior computer.
--
Chris

[Boondocker]

Hello Chris,

Notice the rotor and magnets appear black in the photos.  Are the magnets and rotor treated with POR-15?

Also how much of an offset is the tail pin from the outside of the yaw shaft?  In your write-up you provide, 18 degrees from vertical at a 45 degree to the main shaft, however the distance is not mentioned. 


Thanks

Boondocker

[ChrisOlson]

The gen rotors are sealed with POR-15.

The drawing for the tail hinge brackets is attached - they are made of 1/4" steel - two of them, one on each side of the tail hinge pin.
--
Chris